Intraoperative Neurophysiological Monitoring For Deep Brain Stim

INTRAOPERATIVE
NEUROPHYSIOLOGICAL MONITORING
FOR DEEP BRAIN STIMULATION
INTRAOPERATIVE
NEUROPHYSIOLOGICAL
MONITORING FOR
DEEP BRAIN
STIMULATION
Principles, Practice, and Cases
Erwin B. Montgomery Jr., MD

Medical Director
Greenville Neuromodulation Center
The Greenville Neuromodulation Scholar in Neuroscience and Philosophy
Thiel College
Greenville, PA
1
1
Oxford University Press is a department of the University of
Oxford. It furthers the University’s objective of excellence in research,
scholarship, and education by publishing worldwide.
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam
Oxford is a registered trademark of Oxford University Press
in the UK and certain other countries.
Published in the United States of America by
Oxford University Press
198 Madison Avenue, New York, NY 10016
© Oxford University Press 2014
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form or by any means, without the prior
permission in writing of Oxford University Press, or as expressly permitted by law,
by license, or under terms agreed with the appropriate reproduction rights organization.
Inquiries concerning reproduction outside the scope of the above should be sent to the
Rights Department, Oxford University Press, at the address above.
You must not circulate this work in any other form
and you must impose this same condition on any acquirer.
Library of Congress Cataloging-in-Publication Data
Montgomery, Erwin B., author.
Intraoperative neurophysiological monitoring for deep brain stimulation : principles, practice, and cases / Erwin B.
Montgomery Jr.
p. ; cm.
Includes bibliographical references.
ISBN 978–0–19–935100–8 (alk. paper)
I. Title.
[DNLM: 1. Deep Brain Stimulation. 2. Intraoperative Neurophysiological Monitoring—methods.
3. Nervous System Diseases. WL 368]
RC386.6.N48
616.8′0475—dc23
2014003253
The science of medicine is a rapidly changing field. As new research and clinical experience broaden our knowledge,
changes in treatment and drug therapy occur. The author and publisher of this work have checked with sources
believed to be reliable in their efforts to provide information that is accurate and complete, and in accordance with
the standards accepted at the time of publication. However, in light of the possibility of human error or changes
in the practice of medicine, neither the author, nor the publisher, nor any other party who has been involved in the
preparation or publication of this work warrants that the information contained herein is in every respect accurate
or complete. Readers are encouraged to confirm the information contained herein with other reliable sources, and
are strongly advised to check the product information sheet provided by the pharmaceutical company for each drug
they plan to administer.
9 8 7 6 5 4 3 2 1
Printed in the United States of America
on acid-free paper
To Lyn Turkstra, my love and partner in all things, and
Gary I. Allen, who taught me to listen to neurons.
CONTENTS
Preface ix
1. Importance of intraoperative neurophysiological monitoring 1
2. Preparations for intraoperative neurophysiological monitoring 30
3. Basic concepts of electricity and electronics 48
4. Electrode recordings: Neurophysiology 63
5. Microelectrode and semi-microelectrode recordings: Electronics 86
6. Noise and artifact 106
7. Microelectrode recordings: Neuronal characteristics and

behavioral correlations 125
8. Microstimulation and macrostimulation 142
9. The subthalamic nucleus 159
10. The globus pallidus interna nucleus 172
11. The ventral intermediate nucleus of the thalamus 187
12. Clinical assessments during intraoperative

neurophysiological monitoring 201
13. Cases 227
14. Future intraoperative neurophysiological monitoring 317
vii
viii / / C ontents
Appendix A. Subthalamic nucleus deep brain stimulation algorithm 331
Appendix B. Ventral intermediate thalamic deep brain stimulation

algorithm 337
Appendix C. Globus pallidus interna deep brain stimulation algorithm 359
Appendix D. Microelectrode recording form for subthalamic nucleus

deep brain stimulation 370
Appendix E. Microelectrode recording form for globus pallidus interna 373
Appendix F. Microelectrode recording form for ventral intermediate

thalamus 376
Appendix G. Intraoperative macrostimulation for clinical effect

in Parkinson’s disease 379
Appendix H. Intraoperative macrostimulation for clinical effect in

tremor disorders 381
Appendix I. Intraoperative macrostimulation for clinical effect

on dystonia 383
Appendix J. Intraoperative macrostimulation for clinical effect on tics 386
Appendix K. Intraoperative macrostimulation for clinical effect on

dyskinesia 388
Index 391
PREFACE
A remarkable therapy, deep brain stimulation (DBS) has helped many thousands
of patients who, other therapies having failed them, otherwise faced great hardship.
Numerous neurological and psychiatric disorders are amenable to DBS, and many
more promise to become so. Reasons exist as to why this amenability is increasing.
The brain is essentially an electrical device. Within it resides neurotransmitters
that convey information between neurons. Neurotransmitters are the messenger and
not the message. The conveyance of information is accomplished by virtue of their
use rather than any inherent property. Specifically, spatial and temporal patterns of
neurotransmitter release, which are ultimately determined by the neurons’ electrical
activities, convey information. Neurotransmitters are like electrons flowing through
the computer: Nothing inherent to an individual electron perforce implies information.
Like electrons in action in computer circuits, patterns and pulses determine neuronal
operations. If a computer fails, one does not simply lift the lid and dump a bunch of
electrons onto the computer motherboard. Deep brain stimulation targets neurons’
electrical activities.
The spatial and temporal specificity of DBS admits of few pharmacological equiv-
alents, its accuracy and precision in some regions of application currently measured
in sub-millimeter units. Thus, the spatial and temporal resolution of information
processed in the brain is on the order of sub-millimeters and milliseconds. Whereas
pharmacological agents act over the whole brain, at least over wide areas with similar
neurotransmitter receptors, and act on the order of hours.
None of this is meant to denigrate neuropharmacology. Indeed, further research
in neuropharmacology and its foundational sciences is sorely needed. Because it has
its basis in neurotransmitter physiology, neuropharmacology will continue to prove
quite effective in treating a wide range of neurological and psychiatric disorders.
Though medications may produce many more side effects than does DBS, and do so
more frequently, their reversibility recommends them over surgical therapies. Yet, the
ix
x / / P reface
fact remains that an increasing number of patients will need DBS as pharmacological
approaches fail to produce desired benefits. The number of centers offering DBS will
therefore increase to meet this need.
Clinical success of DBS depends on accurate placement of leads, which house the
electrical contacts for stimulation. The ability to change the patterns of electrodes used
for stimulation and the various properties of the stimulation—frequency of stimula-
tion pulses, pulse width, current, and voltage—offer means of tailoring DBS to each
patient’s unique anatomy in the vicinity of the DBS leads. Misplacement by even a mil-
limeter, however, spells the difference between success and failure.
Thanks to remarkable advances in image-based surgical navigation, DBS surgery
has become safer and easier. Targets never actually seen directly by the surgeon may
be reached by aid of current technology, which is capable of placing DBS leads with
precision and accuracy on the order of a millimeter. The critical question becomes how
the target is “seen,” if not by the unaided eye, then by some other method, such as mag-
netic resonance imaging (MRI) or other imaging techniques. However, effectiveness
depends on whether the target “contrasts” with its neighbors in terms of the physics
underlying the imaging.
In a way, the very terms used in DBS confuse the issue. For example, one speaks of
DBS of the subthalamic nucleus (STN), which presumes or implies that it is stimula-
tion of the STN that actually provides the therapeutic benefit. Consequently, one only
has to be able to see the STN, relative to its neighbors, to successfully direct the DBS
lead to the target. However, this is false on several accounts. First, it is not the STN
that is the target of STN DBS. Rather it is the sensorimotor region of the STN with the
limbic and associative regions of the STN avoided. Current MRI and other imaging
methods cannot visualize these other regions in the STN. Consequently, there is no
contrast detectable by these methods that differentiate the sensorimotor region from
the limbic and associative regions. When the descriptive term is DBS of the sensorimo-
tor region of the STN, or the sensorimotor region of the globus pallidus interna or the
arm region of the ventral intermediate nucleus, the issues regarding targeting become
more realistic.
Even with the clarifying specification of DBS of the sensorimotor STN, misunder-
standing can be conveyed. For example, it is not clear that actually stimulating neu-
rons in the sensorimotor region of the STN is responsible for the therapeutic benefit.
Rather, there is evidence that stimulation of the cortical projections to or in the vicin-
ity of the sensorimotor neurons of the STN is critical to the therapeutic mechanisms.
Thus, one might better speak of DBS of the axons in the vicinity of the STN. At the
least, this is a more honest expression of the current state of knowledge and does not
Preface / / xi
have the effect of reinforcing misplaced presumptions. Paraphrasing Claude Bernard,

father of modern physiology, “we are more often fooled by things we think we know
than things we do not.”
The problem becomes, then, one of context; a target deemed optimal for radiologi-
cal purposes may not be optimal for clinical purposes. For this reason, surgeons rely on
additional means of determining the best target. The earliest days of stereotactic func-
tional neurosurgery involving electrocautery or radio-frequency lesioning saw employ-
ment of stimulation through the lesioning electrode and other electrophysiological
means of assuring best targeting. Cases in which test stimulation through the lesioning
electrode confer observable benefit serve to bolster a surgeon’s confidence that future
lesioning will meet with success and reduced risk.
Microelectrode and semi-microelectrode recordings of extracellular action poten-
tials generated by neurons, local field potentials, test stimulation through the DBS
leads—the range of electophysiological means of identifying the optimal DBS target
are numerous and varied. Nearly every surgeon uses test stimulation through the DBS
leads. Thus, whatever opinion one has of other forms of electrophysiological studies,
she must understand fully test stimulation through DBS leads. Many, if not most,
surgeons and neurologists appreciate the importance of supplementing excellent
image-guidance and DBS test stimulation with additional neurophysiological meth-
ods. In one study of 144 STN DBS surgeries, for example, 30% of cases required more
than one trajectory of microelectrode recordings, because the initial image-guided
trajectories failed to meet the criteria for a physiologically defined optimal target
(Montgomery 2012). This observation suggests that imaging alone proved insufficient.
Failure to encounter the physiologically defined optimal target in the image-guided
trajectory, for example determined by microelectrode, semi-microelectrode record-
ings, or DBS test stimulation, leaves the intraoperative neurophysiologist won-
dering in which direction and distance she ought to move the microelectrode,
semi-microelectrode or DBS lead. Image guidance brings her no closer to that deci-
sion. The development of MRI and CT and other intraoperative imaging may be of
aid, but whether they will supplant electrophysiology based studies remains uncer-
tain. Only electrophysiological means employed interoperatively provide an answer.
It thus behooves those involved in DBS lead-placement surgery to gain expertise in
electrophysiology based methods even if just for intraoperative DBS test stimulation.
All of these methods are based on fundamental properties of biophysics, electricity,
and electronics.
The development of turnkey commercial systems for intraoperative neurophysi-
ological monitoring has been both a boon and bane; these systems may leave one with
xii / / P reface
the impression that intraoperative neurophysiological monitoring is routine, when it is

anything but. In the abovementioned study 30% of participants needed two or more
microelectrode recording trajectories to locate the target. One must therefore acquire
a measure of expertise in order to determine whether the microelectrode trajectory is
acceptable. If she deems it unacceptable, then she must know the fundamental prin-
ciples of neurophysiology and neuroanatomy to help her to decide her next course of
action. This presupposes that she will be able to identify whether intraoperative neuro-
physiological monitoring systems work properly, without artifact and electrical noise,
to guide her. Distinguishing artifact and noise, as a preliminary to removing them,
requires advanced knowledge of biophysics, electricity, and electronics.
The present writing endeavors to maximize the probability of excellent outcomes in
DBS. Yet one seldom gains expertise solely by reading, no matter how expert the text
in question may be. Those already engaged in intraoperative neurophysiological moni-
toring for DBS will find in this book ample material for ongoing discussions. Those
planning to engage in intraoperative neurophysiological monitoring—or in any other
diagnostic or therapeutic method based on electrophysiology, for that matter—may
regard this book as a primer.
Many chapters repeat certain material (the chapters related to specific DBS targets,
for example). This repetition is wholly intentional; certain chapters are intended to
serve as stand-alone references.
Algorithms also appear in this book. These help the reader to interpret the infor-
mation obtained from microelectrode recordings and other intraoperative neuro-
physiological monitoring. The reader should regard these algorithms as assisting rather
than dictating the intraoperative neurophysiologist and surgeon’s respective actions.
Because no algorithm anticipates every individual patient’s circumstances, intraopera-
tive neurophysiologists and surgeons bear responsibility for how they apply any infor-
mation herein to their patients’ care.
Every attempt has been made to provide evidence and reason for the various algo-
rithms, procedures, and claims made. However, this is not to claim that every claim or
suggestion has been subjected to randomized controlled prospective investigation, nor
could they. One could take this lack of evidence-based medicine level 1 evidence as an
excuse to dismiss the claims and recommendations or to hold that in the absence of
level 1 evidence one has license to do as one chooses. This would be an exercise in solip-
sism (see http://ReasonBasedMedicineAndScience.com) and would not be in the best
interests of patients or advancing DBS therapies. The current state of affairs is that no
two surgeons perform DBS implantation surgeries exactly the same. The high degree
of variability leads to the question whether all of the different techniques cannot be the
Preface / / xiii
best or whether the differences matter little in the outcome. These important questions
cannot be resolved if each surgical team isolates themselves by avoiding discussion
and debate. While many readers may disagree with the claims and recommendations
offered here, it is hoped that this will be an invitation to discussion and debate that is
widely accepted.
The book presents a number of actual cases. Each case’s intraoperative microelec-
trode recordings appear as they did in the operating room at the time they were made.
Interpretive commentary follows the presentation of microelectrode recordings for
each trajectory. Finally, there appear postoperative imaging studies that demonstrate
the monitoring results, some of which may contain errors or complications. The author
included these errors and complications because he believes that one often learns more
from failure than success. The reader should not gather from these inclusions that com-
plications happen frequently. Provided one follows proper surgical techniques, taking
care especially to prevent brain shift from air entering the skull, complications occur
but rarely.
Indeed, DBS has met with such remarkable success that it risks leaving one with the
impression that the procedure is easy. Most of the time, the procedure goes smoothly
and brings the patient the desired benefit. The value in extended experience-based
training of intraoperative neurophysiologists owes to those rare moments when the
procedure meets with complication and the neurophysiologist’s mettle is tested. For
this reason, the author recommends that a training period comprise ten surgeries of
each type of DBS; it affords the trainee opportunity to gain experience in difficult cases.
This author’s own training would have been impossible were it not for the many
others whose fellowship inspired, enlightened, and encouraged him. Among them is
Dr. Gary I. Allen, under whose guidance this author first eavesdropped, via micro-
electrode recordings, on the incredible “conversations” neurons have with each other.
Dr. Allen sacrificed many evenings to mentoring the author, including the evening of
his tenth wedding anniversary. The kindness Dr. Allen showed is difficult to repay, but
is just as difficult to forget.
This author owes a debt to the following individuals: Dr. Lyn Turkstra, whose
love and support through years of marriage have sustained him; Steven Buchholz,
who collaborated with him in early studies of neuronal activities in the basal ganglia–
thalamic-cortical system in nonhuman primates; Drs. Doug Stuart and Thomas Hixon,
who advised and protected his research; Dr. John Gale, former technician and now col-
league and friend, whose uncommon enthusiasm recharges this author’s intellectual
“batteries” whenever they run low; He Huang, who began his association with this
author as a computer programmer but over the years has become an excellent fellow
xiv / / P reface
neuroscientist; and Drs. Thomas Mortimer, Dominic Durand, and Warren Grill, whose
course on neurostimulation at Case Western Reserve University influenced this book’s
content and direction.
This author also thanks FHC, Inc., which extended its generosity by funding a
writer and editor to give the manuscript some much-needed polish and left the project
under complete authorial control, and his students and patients, who have made pos-
sible the great privilege of his being a teacher and physician.
Erwin B. Montgomery Jr., MD
May 8, 2014
REFERENCE
Montgomery EB Jr.: Microelectrode targeting of the subthalamic nucleus for deep brain stimulation sur-
gery. Mov Disord. 2012 Sep 15;27(11):1387–1391. doi: 10.1002/mds.25000. Epub 2012 Apr 16.
/ / / 1 / / / IMPORTANCE OF
INTRAOPERATIVE
NEUROPHYSIOLOGICAL
MONITORING
INTRODUCTION
An evolving therapy for neurological and psychiatric disorders, Deep Brain Stimulation
(DBS) succeeds where all manner of medications and brain transplants fail. Though
currently in an early stage of development, gene therapy has yet to match DBS’s ben-
efits, which outstrip those of the best medical therapy. In addition to the risks attend-
ing DBS surgery, the difficult task of placing the stimulation electrodes accurately and
precisely means a premium attaches to getting it right the first time.
Not at all a foregone conclusion, precise and accurate placement of stimulating elec-
trodes remains the subject of ongoing debate. The diverse methods employed by phy-
sicians share a common element, namely, use of some manner of neurophysiological
monitoring in the operating room, if for no other reason than to confirm the absence
of adverse effects.
Experience and the present state of technology have led this author to favor the
use of microelectrode recordings for optimal placement of DBS leads, which house
the stimulating electrodes for therapy. Yet this preference does not diminish the
continued importance of macrostimulation through the DBS lead, a form of intra-
operative neurophysiological monitoring that all physicians utilize to some extent.
By discussing both microelectrode recordings and macrostimulation through the
DBS lead, the author hopes to make his book useful to physicians partial to either
method.
1
2 / / I ntraoperative N europhysiological M onitoring for D B S
THE GROWING IMPORTANCE OF DEEP BRAIN STIMULATION AND

INTRAOPERATIVE NEUROPHYSIOLOGICAL MONITORING
The best outcome possible for patients guides the purpose of intraoperative neurophys-
iological monitoring, and the best outcome depends on proper placement of DBS leads,
which neurophysiological monitoring helps to ensure. Intraoperative neurophysiologi-
cal monitoring admits of several techniques: microelectrode recordings for purposes
of identifying and analyzing extracellular action potentials; electrical stimulation via
the microelectrode (microstimulation); local field-potential recordings via macroelec-
trodes (DBS leads, for example); and macrostimulation via DBS leads or the indifferent
electrode in some bipolar microelectrodes. This book discusses each technique.
Though these techniques differ in terms of the roles they play in identifying the
clinically optimal stimulation target, their effectiveness depends on a single factor: an
understanding of the biophysical properties, physiological characteristics, electronics,
and regional anatomies of structures surrounding DBS targets. This book focuses on
providing such an understanding.
Intraoperative microelectrode recordings of extracellular action potentials gener-
ated by neurons, which have assisted surgical procedures for decades, were initially
made before lesioning (ablating) the target structure. The diminution of benefit over
time and the significant risk to speech and swallowing consigned these surgical proce-
dures to rare use. Early in its development, DBS reduced these risks, and surgical proce-
dures came to enjoy a revival as various movement disorders and psychiatric conditions
recommended themselves for such intervention. As the number of eligible patients has
increased, the need for persons trained in intraoperative neurophysiological monitor-
ing has also increased.
That for many disorders DBS surpasses the best medical therapies owes to its remark-
able effectiveness. Early in its development, DBS brought relief to patients who faced
surgery because all other reasonable medical and psychological therapies had failed. So
impressively has DBS succeeded, in fact, that the US Food and Drug Administration
(FDA) has approved it for the following notable indications: Parkinson’s disease (uni-
lateral thalamic); essential tremor (unilateral thalamic); primary dystonia in patients
aged seven years or more; and obsessive-compulsive disorder (via a humanitarian
device exemption [HDE]). A number of “off-label” uses have come to be considered
standard and accepted therapy for the following notable conditions: Parkinson’s dis-
ease (bilateral thalamic); essential tremor (bilateral thalamic); Tourette’s syndrome;
secondary dystonia owing to perinatal injury; tardive dystonia; tardive dyskinesia;
Huntington’s disease; and multiple sclerosis. Also underway are clinical trials of DBS in
1. Importance of Intraoperative Neurophysiological Monitoring / / 3
the treatment of epilepsy, depression, Alzheimer’s disease, tinnitus, and stroke, among
others. The vast number of therapeutic applications—actual and potential—augurs a
substantial increase in the need for intraoperative neurophysiological monitoring.
Randomized controlled trials pitting DBS against the best medical therapies in
Parkinson’s disease have shown that patients undergoing the former experienced
greater relief than did patients undergoing the latter, peculiarities in definitions of
acute adverse effects related to the former notwithstanding (Weaver, Follett, Stern
et al. 2009; Weaver, Follett, Stern et al. 2012). Physicians who treat Parkinson’s disease
recognize that, with the number of available medications, the number of possible drug
combinations has increased exponentially. They therefore confront a question similar
to the one epileptologists ask themselves, which is how many anticonvulsants they
must prescribe before these drugs’ benefits grow fewer than those to be gained from
surgery. Epileptologists realize that, whether tested alone or in combination with other
drugs, a new medication faces a long trial period—a period, one hastens to add, dur-
ing which patients continue to face uncontrolled seizures, mounting side effects from
currently prescribed medications, and other risks. Compounding the epileptologist’s
dilemma is the fact that, though a degree of risk attends many surgical procedures, they
can cure certain patients. A similar dilemma confronts neurologists and psychiatrists
in treating disorders amenable to DBS.
DEEP BRAIN STIMULATION’S FUNDAMENTAL DIFFERENCE FROM

PHARMACOLOGICAL THERAPIES
Fundamentally different from, and therefore an important supplement to, available

pharmacological treatment, DBS represents a sea change in therapeutics for neurologi-
cal and psychiatric disorders. The difference lies in the specificity of action at particular
spatial and temporal scales. Though pharmacological treatments, whose risks relative
to surgery are modest, ought to endure as the initial treatment option, DBS does pos-
sess unique virtues enough to win it appreciation in its own right. As this apprecia-
tion grows, so grows the need for intraoperative neurophysiological monitoring and for
future basic, translational and clinical DBS-related research.
Increasingly, neurological and psychiatric disorders are being appreciated as the
consequence of misinformation rather than excess or insufficient neuronal activity. This
means that most neurological and psychiatric disorders result from worsening of the
signal-to-noise ratio, misinformation, or other reduced or degraded information. The
question then becomes how potential therapies affect changes in the misinformation.
The basis for most pharmacological treatments, neurotransmitters do not constitute
messages; they are the messengers whose message originates by other means. Patterns
of intraneuronal and interneuronal electrical activity mediate information in the brain.
Though neurotransmitters form the basis for most interneuronal communication (gap
junctions are the exception), viewing such communication as owing solely to the flux
of neurotransmitters represents an instance of the commission of a particular error in
reasoning, essentially reductionist in nature, that is characterized by one’s deeming a
quality or function of a part as applying also to the whole. This error, known as the
mereological fallacy, can lead to a loss of knowledge. Reducing brain function to neu-
rotransmitter levels and neuropathophysiology to the relative excess or deficiency in
the quantity of neurotransmitters (the so-called neurohumoral paradigm) risks miss-
ing the complexity and thus, the most relevant level of analysis and intervention. For
the same reason that one cannot infer from an electron the function of a computer or a
telephone, one cannot infer from the chemical nature of dopamine its function in the
brain or the effects of its diminishment, as happens with Parkinson’s disease.
Electrical activity occurs in neurons and between them, and this activity from
other neurons converges in the next neurons’ dendrites, cell bodies, and, in some
cases, axo-axonal connections. From the pattern of convergence within a neuron then
results a pattern of extracellular action potentials that are transmitted down axons to
make contact on subsequent neurons at the synapse. The pattern of electrical activ-
ity that reaches the synapse contains the information and determines the pattern of
neurotransmitter release. The information is not inherent in the neurotransmitter but
in the pattern of neurotransmitter release. The applications of neurotransmitters (or
agents that block neurotransmitters) without regard to the temporal dynamics at the
time scales of electrical activities within neurons is not likely to replicate and restore
normal physiology. Deep brain stimulation acts at the level of the electrical patterns
within and between neurons.
Deep brain stimulation enjoys over most pharmacological approaches a signifi-
cant advantage in terms of spatial or anatomical specificity of action. Because most
pharmacological agents depend on specific types of neurotransmitter receptors, the
selectivity of action for the pharmacological agent depends on the receptors’ spatial or
anatomical distribution, which, in the case of most brain neurotransmitters, is wide.
Side effects occur when neurotransmitter replicants (agonists) or blockers (antago-
nists) reach receptors occupying areas outside the desired targets. Dopamine receptors
in the motor areas of the caudate nucleus and striatum, for example, probably mediate
the therapeutic effect of dopaminergic agents in the treatment of the motor symp-
toms of Parkinson’s disease. Thus, it is likely that many psychological and cognitive
DBS side effects proceed from two phenomena: (1) stimulation of dopamine receptors
in areas that interact with the limbic and cognitive systems and (2) inadvertent acti-
vation of dopamine receptors in the cortical and limbic systems. The vastly smaller
volume of tissue affected by DBS (approximately 2.5 mm in radius) may account for
the fewer long-term adverse events observed in patients with Parkinson’s disease who
underwent DBS than in those who received best medical therapy (Weaver, Follett,
Stern et al. 2009).
Remarkable advances in pharmacology made since the mid-twentieth century have
benefited countless patients. Remarkable advances in pharmacological treatments
for Parkinson’s disease, particularly, have benefited countless neuroscientists. These
advances led to the development of theories of pathophysiology that assign critical
significance to neurotransmitters—the theory of cholinergic/dopaminergic imbal-
ance popular in the 1970s, for example. The current theory of globus pallidus interna
overactivity as causal to Parkinson’s disease rests on a fundamentally pharmacological
notion. The fact that neurotransmitter function–based pharmacological therapies met
with considerable success conceals the post hoc nature of the undergirding reasoning,
which proceeds from an incorrect inference, drawn from improvements observed in a
variety of neurological and psychiatric disorders.
The ad hoc reasoning undergirding pharmacological theories of pathology has sub-
tending it a second error in reasoning that instances what is known as the fallacy of
pseudotransitivity. In the 1920s scientists applied acetylcholine to an isolated heart
preparation, which slowed the heart rate just as electrical stimulation of the vagus nerve
did. From these phenomena scientists drew the inference, which subsequent research
validated, that acetylcholine must mediate the effects of the vagus nerve (Valenstein
2005). Again, such validation obscures the fact that the inference as to synonym-
ity between neurotransmitters and neuronal activity rests on a fallacy—in this case,
the fallacy of pseudotransitivity—which assumes the following formal expression: If
a implies c and b implies c, then a implies b. Though stimulation of the vagal nerve (a)
slows the heartbeat (c) and the application of acetylcholine (b) slows the heartbeat (c),
it does not necessarily follow that electrical stimulation of the vagus nerve (a) implies
acetylcholine (b). Happily, subsequent research demonstrated acetylcholine as the neu-
rotransmitter of the vagus nerve. Thus, the fallacy employed served a constructive role
in providing the hypothesis, subsequently demonstrated, but it would have been a dis-
service had the fallacy been taken as evidence or fact.
Though fallacious, the inference linking heart rate to the mediating effects of
acetylcholine led to the development of a reasonable hypothesis that, once vali-
dated, advanced knowledge of acetylcholine and the vagus nerve. Yet there fol-
lowed also an adverse consequence: From phenomena observed in the specific vagus
nerve–acetylcholine instance scientists derived the notion that one can explain
all brain function in terms of neurotransmitters. Whatever success realized in that
instance fails to translate to other situations. Dopamine replenishment, by medication
or cell transplants, in brains of patients with Parkinson’s disease often brings about
no improvement in symptoms (Olanow, Goetz, Kordower et al. 2003). A nonpharma-
cological therapy, DBS succeeds in this respect, thus putting paid to the notion that
neurotransmitters govern all brain function.
Understanding how DBS differs from pharmacological therapies depends on under-
standing the former’s mechanisms of action. One notes two peculiar qualities of DBS’s
therapeutic effects: (1) they bear no relation to dopamine levels in the brain (Hilker,
Voges, Ghaemi et al. 2003) and (2) they bear no relation to injected electrical charge
alone but to the latter in combination with the timing of its pulse (Montgomery 2005).
Attesting to the importance of timing is the fact that in some patients with Parkinson’s
disease DBS at 130 pulses per second (pps) proves effective, while DBS at 100 pps
does not. A mere three milliseconds (3/1000 of a second) difference in the duration
between electrical pulses, this difference makes all the difference and it offers a sense
of the time scales at which DBS operates. Coupled with the importance of timing is
that of dynamics (changes in state over time). Failure to account for dynamics explains
the inadequacy of the neurohumoral paradigm informing current pharmacological
and neurohumoral approaches as described above. On the order of 100 ms, the time
course of dopamine release in the basal ganglia as represented by the discharge patterns
of dopamine neurons in the substantia nigra pars compacta (Figure 1.1), for example,
contrasts dramatically with the time course of the action of dopaminergic agents in the
pharmacological treatment of patients with Parkinson’s disease.
Reward predicted
Reward occurs
CS
R
FIGURE 1.1 Raster and histogram, recorded over the time course of a behavior, of a dopamine
neuron residing in the substantia nigra pars compacta of a nonhuman primate. The raster
in the bottom of the figure shows a row for each trial of the task. Each dot represents the
discharge of the dopamine neuron. The top graph shows the number of spikes at each bin of
time summed across trials. CS is the condition stimulus that predicts a reward (R). One notes
a large increase in the activity of the dopamine neuron related to the CS. The time course of
the dopamine release is on the order of 100 ms (Schultz 1998).
IMPORTANCE OF TARGET LOCALIZATION AND THE ROLE OF

INTRAOPERATIVE NEUROPHYSIOLOGICAL MONITORING
Though it would seem trivially obvious that the success of DBS depends on stimulating
the correct target, the demands of the accurate placement of the DBS lead make the
matter anything but trivial. Indeed, the complexity of stimulation of the desired targets
and the avoidance of undesired targets places marked constraints on the methods avail-
able to achieve the necessary accuracy.
Some 13 mm in length and 6 mm in width, and roughly the shape of an American
football, the subthalamic nucleus presents a small target (Yelnik and Percheron 1979).
The challenge of reaching such a small target is compounded by the fact that a particu-
lar region of the subthalamic nucleus must be reached (Figure 1.2). This true target,
the sensorimotor region, occupies approximately half of the subthalamus (Figure 1.3)
and thus effectively doubles the accuracy requirements. Stimulation of other portions
of the subthalamic nucleus having connections to the limbic system, the prefrontal
cortex, and the orbital frontal cortex can result in significant mood and cognitive
problems. Knowing the location of the sensorimotor region within the subthalamic
nucleus is therefore imperative. Similarly, the posterior limb of the internal capsule,
which occupies a lateral, anterior, and ventral position vis-à-vis the subthalamic
nucleus, can experience tonic muscular contraction if stimulated. Posterior to the
subthalamic nucleus lies the ascending medial lemniscus, the inadvertent stimulation
Midbrain (Mesencephalon)
Medial lemniscus
Posterior
Lateral
Medial
STN Anterior
Corticobulbar
corticospinal
tracts
Oculomotor complex
and fascicules
Substantia nigra pars
Reticulata
Compacta
FIGURE 1.2 Schematic representation of the regional anatomy of the subthalamic nucleus

through the midbrain (mesencephalon). The subthalamic nucleus (STN) lies above the sub-
stantia nigra pars reticulata and compacta with the reticulata lying more laterally. The corti-
cobulbar and corticospinal tracts run lateral, anterior, and ventral to the STN, and the medial
lemniscus fibers run posteriorly. The oculomotor complex with its exiting fascicles is medial.
FIGURE 1.3 Schematic representation of the spatial accuracy and precision required for DBS
of the subthalamic nucleus. Shown is a sagittal section showing the subthalamic nucleus
(indicated by tip of the ballpoint pen). The actual target for stimulation, the sensorimotor
region, lies in the subthalamic nucleus. The size of the sensorimotor region relative to the tip
of a ballpoint pen offers an idea of the accuracy required.
of which by suboptimal currents may result in intolerable side effects. Medial to the
subthalamic nucleus lie nerve fascicules (roots) of the oculomotor nerve. Stimulation
of these can produce double vision. The region of the brain physicians must navigate to
place DBS leads is not only exceedingly small but also fraught with difficulties.
The methods one can use to identify the actual target depend on whether one can
distinguish the target from other regions that, as a consequence of their proximity,
DBS may inadvertently affect. One approach rests on the fact that such white-matter
structures as the posterior limb of the internal capsule and the medial lemniscus dif-
fer from such gray matter structures as the subthalamic nucleus in terms of proton
density, radiodensity, and electrical resistivity. Thus, one can differentiate these struc-
tures by use of magnetic resonance imaging (MRI), computerized tomography (CT
scan), and measurements of electrical impedance, respectively. In their present form,
however, none of these methods can differentiate areas within a gray-matter structure.
Whether employed preoperatively or intraoperatively, these methods also cannot dif-
ferentiate the sensorimotor region within the subthalamic nucleus from those parts
within the subthalamic nucleus where inadvertent stimulation could produce seri-
ous side effects, such as regions projecting to the limbic and frontal cortical systems.
At best, currently, these methods only identify the neighbors to the target and not
the actual target. Microelectrode and semi-microelectrodes specifically identify the
sensorimotor regions and consequently define the optimal target. The question that
must be addressed by those who do not use microelectrode or semi-microelectrode
recordings is whether identifying the neighbor is sufficient.
The same issues apply to DBS surgery for the thalamus and the globus pallidus
interna. The methods described above do not differentiate the ventral intermediate
nucleus of the thalamus (the DBS target) from any of the nuclei of the ventral and lat-
eral region of the thalamus. Similarly, these methods do not differentiate the senso-
rimotor region of the globus pallidus interna from nonmotor regions.
Compounding the demands on accuracy is that for some targets merely identifying
the sensorimotor region is insufficient. Rather, one must identify specific homuncular
(body) representations within the sensorimotor regions (Figure 1.4). Thalamic DBS
CORONAL SOMATOTOPIC ARRANGEMENT AXIAL SOMATOTOPIC ARRANGEMENT
FIGURE 1.4 Sensorimotor anatomy (homunculus) of the ventral intermediate nucleus of the

thalamus. The view of the coronal plane, to the left, shows the homuncular representation as
layers where the lower extremity is lateral and then sweeps ventrally. Just medial and supe-
rior is the upper extremity with the head representation the highest. The medial-to-lateral
organization also is seen in the axial view. Source: Reproduced with permission from Hassler R
in Schaltenbrand and Wahren (1977).
presents a useful example. It involves a homuncular representation consisting of the

following correspondences: the medial region, the homuncular head; the lateral region,
the lower extremity; and the region between, the upper extremity. A DBS lead placed
too close to the head region can, once stimulation occurs, increase risk of impaired
speech, language, and swallowing ability. This makes identifying the head representa-
tion important.
Targeting the upper extremity representation in the thalamus in patients who experi-
ence predominantly proximal or distal extremity tremor requires that one distinguish
the proximal extremity representation from the distal upper extremity representation.
Targeting the distal upper extremity representation brings the DBS lead even closer to
the head representation and thus increases the required degree of accuracy. Some patients
with predominately lower extremity tremor, such as in primary orthostatic tremor (a vari-
ant of Essential tremor), require targeting of the lateral region of the ventral intermediate
nucleus. This targeting brings the DBS lead close to the posterior limb of the internal
capsule and thus risks causing tonic muscle contraction, which limits therapeutic benefit.
When attempting to assess the clinical response to micro- or macrostimulation,
one cannot use the distribution of the reported paresthesias for localization within
the homunculus as means of so doing. Studies combining microelectrode recordings
with subsequent microstimulation demonstrate the difficulty. Discordant responses
are those where the distribution of the paresthesias from microstimulation is not the
same as the regions of the body that drive neuronal responses; concordant responses
are when there is an overlap. Discordant responses were found in approximately 50%
of the stimulated and recorded sites (Grill, Simmons, Cooper et al. 2005). It is possible
that the discordant stimulation occurred in axons that, though they pass in the vicinity
of the microelectrode, project to a different homuncular representation. The paresthe-
sias would be referred to the homuncular representation being stimulated distant from
the site of the stimulation (see Figure 1.5).
A similar situation involves the globus pallidus interna, a structure whose homun-
cular region covers a larger spatial and anatomical area relative to the effective radius of
the volume of tissue activated by DBS. The lower extremity representation is anterior,
medial, and dorsal to the head representation, which resides in the posterior, lateral,
and inferior region of the globus pallidus interna. The upper extremity representation
lies between the two (Figure 1.6). Targeting within the sensorimotor region of the glo-
bus pallidus interna thus varies according to the region of the body affected—cervi-
cal region from upper or lower extremity segmental dystonia, for example. Figure 1.6
shows the homuncular representation whose spatial extent, which is approximately
2.5 mm, exceeds the usual radius of effective DBS (Butson, Cooper, Henderson et al.
Leg
Amp
Concordant
Arm
Leg
Amp
Arm
Discordant
FIGURE 1.5 Possible mechanism of discordant and concordant paresthesias in response to

stimulation. In the case of concordant responses, microelectrodes recorded activity changes
correlated with movement of the arm but not with movement of the leg. Thus, there is
relative certainty that the microelectrode is within the arm homuncular representation.
Microstimulation (represented by the spark images) at the same site activates local neurons
(white cartoons of neurons). This activation patients experience as paresthesias of the arm. In
the discordant response, the microelectrode continues to record changes in neuronal activity
with movement of the arm but not with movement of the leg. Hence, the microelectrode is
within the arm homuncular representation. However, microstimulation activates axons that
pass through the site as they project to the leg representation. Stimulation of these axons as
they pass through the arm representation causes paresthesias referred to the leg.
GPe
IC
OT
UPPER LIMB
L-20 AXIL OR FACE L-22
LOWER LIMB
TREMOR CELLS
FIGURE 1.6 Sensorimotor anatomy (homunculus) of the globus pallidus interna. One notes
a greater number of facial units (filled diamonds) situated posteriorally and laterally (L-22
indicating 22 m lateral to the AC-PC line) relative to 20 mm lateral to the AC-PC line (L-20), and
a greater number of lower limb–related units (filled circles) situated medially and anteriorly.
Upper extremity–related units (open circles) are interposed between. Source: Reproduced with
permission from Guridi, Gorospe, Ramos et al. (1999).
2007). Simply by placing the DBS lead into the sensorimotor region of the globus palli-
dus interna one risks missing the most optimal region in patients with focal or segmen-
tal dystonia. The large spatial distribution of the sensorimotor homunculus relative to
the volume of tissue activation with DBS presents a problem for patients with hemidys-
tonia or generalized dystonia; a significant percentage of them require placement of
multiple DBS leads in order to obtain satisfactory response.
Important also is the trajectory angle of the DBS lead, particularly with respect to
the ventral intermediate nucleus of the thalamus (see chapter 11), because if it hap-
pens to be tangential rather than parallel to the long axis of the ventral intermediate
nucleus of the thalamus, relatively few of the lead’s stimulating electrodes may actu-
ally enter the ventral intermediate nucleus of the thalamus. Failure to enter this area
reduces efficacy. The angle of the long axis of the ventral intermediate nucleus of the
thalamus, moreover, may vary considerably relative to the Cartesian coordinate sys-
tem of the most precise image-guided surgical navigation. As of this writing, no moni-
toring technique to determine the actual trajectory exists besides microelectrode and
semi-microelectrode recording.
Neuroimaging targeting, which is performed prior to opening the patient’s skull,
complicates the accuracy issue, because the accompanying brain shift increases the for-
mer’s inherent variability (inaccuracy). That is, it affects the degree of precision required
reliably to pinpoint the appropriate target. Microelectrode and semi-microelectrode
recordings can compensate for brain shift to some extent; and intraoperative MRI and
CT scans, along with other methods, are being developed. But these techniques will not
obviate the above-mentioned concerns about identifying sensorimotor regions or their
corresponding homunculi. The magnetic field generated by MRI makes its adjunctive
use in intraoperative neurophysiological monitoring difficult. Improved surgical tech-
niques that reduce brain shift would also reduce the attractiveness of intraoperative
MRI, because it would obviate the need for aided targeting and would therefore only
interfere with microelectrode recordings.
THE EPISTEMIC STATUS OF MICROELECTRODE OR

SEMI-MICROELECTRODE RECORDINGS
The different methods of intraoperative neurophysiological monitoring being

explored—local field recordings made through macroelectrodes, for example, or elec-
troencephalographic evoked potentials from test stimulation produced through DBS
leads—are judged in light of the current use and concerns of microelectrode record-
ings. Many leading centers for DBS surgery make use of microelectrode recordings,
but this hardly means the latter have gained universal acceptance. Indeed, some centers
may only make selective use of them, employing them for other targets but not for the
ventral intermediate nucleus of the thalamus, for example. Some use a single microelec-
trode during monitoring and others an array of electrodes.
Most approaches owe to habit—habit acquired by one’s own experience or in
apprenticeship to a mentor, himself beholden to habit. The understanding underlying
these approaches is thus little more than the result of various attempts at post hoc justi-
fication. This is apparent in the lack of prospective randomized controlled studies com-
paring the different forms of intraoperative neurophysiology.
This author maintains that nearly all discussion of intraoperative neurophysiol-
ogy methodology consists of comparing one method’s outcomes to published out-
comes of previous studies performed typically by different surgeons and their staff.
These studies contain many significant, indeed fatal, flaws. In several studies of DBS
implantation surgery, for example, comparison is made between outcomes of surgery
eschewing intraoperative neurophysiological microelectrode recordings and out-
comes of surgery utilizing them. To the outcomes of either set these studies impute
equal success, thus implying that ceteris paribus surgery eschewing intraoperative
neurophysiological microelectrode recordings ought to be preferred on the basis of its
reduced cost and risk.
Whatever evidence of equivalence discovered by these studies rests on a failure to
find a statistically significant difference, a result that may owe less to absence than error.
Specifically, the failure may owe to commission of a type II error (oversight of a truly
existing difference), an insufficiently small sample size, or a highly variable outcome
measure. More appropriate means exist for demonstrating noninferiority (Wellek
2010), but these see little if any use.
At present, no level 1 Evidence-Based Medicine data exist to determine which set
of intraoperative neurophysiological monitoring methods necessarily and sufficiently
optimizes outcomes. And the foreseeable future holds no promise of any such data
coming to light. The question, then, is how to proceed. Physicians should not consider
the lack of level 1 Evidence-Based Medicine data as license for nihilism or adventur-
ism; doing so would be to commit the logical error of argumentum ad ignorantiam, or
arguing from ignorance. The physician must continue to make responsible decisions
according to the best available information and knowledge. A misconception prevails
that Evidence-Based Medicine is only synonymous with randomized, controlled trials.
Yet in its original formulation, Evidence-Based Medicine included expert consensus.
Missing from this formulation is, of course, any sense of how best rationally to proceed,
because the rationale behind expert consensus goes unstated.
By way of providing a rationale, this author proposes that, in the absence of level
1 Evidence-Based Medicine, one can proceed in a reasonable manner by appealing
to fundamental anatomical and physiological principles. These principles, if logically
applied, can aid rational consideration and decision-making. For example, a physi-
ologically demonstrated DBS-effect radius of approximately 2.5 mm (discounted for
the moment are the electrical inhomogeneities of tissues surrounding the DBS elec-
trodes, which affect the shape of the volume of tissue activation) means that any target-
ing method must be able to place the DBS electrodes within some distance less than
2.5 mm of the appropriate target or more than 2.5 mm from any structure that when
stimulated produces an adverse effect.
If one hews to the principle that the optimal target must be identifiable, his task
becomes that of distinguishing optimal from nonoptimal targets. For example, CT
scans differentiate structures according to each structure’s respective radiodensity.
Given this, one must establish whether the optimal target’s radiodensity differs from
those of nonoptimal, possibly contraindicated structures. Alternatively, establishing
this difference may depend on the optimal target’s location vis-à-vis other structures
or neighbors—its distance from, say, the midpoint of the line connecting the ante-
rior commissure to the posterior commissure. One must also expect that opening the
patient’s skull can cause the brain to shift and thus cause the optimal target to devi-
ate from the position established preoperatively. The stability of the anterior-posterior
intercommissural line’s midpoint relative to external landmarks, in other words,
becomes the issue.
Unfortunately, there exist no controlled direct comparisons of outcomes for
patients who were randomized according to the use or nonuse of intraoperative micro-
electrode recordings. Difficult for a number of reasons, such studies also pose an ethical
problem concerning participation, as well as a practical problem of achieving equipoise
sufficient to enable randomization (Fins 2008). Reports have been limited, rather, to
those made by physicians who, having forgone use of intraoperative microelectrode
recordings, claim to have produced results no different from those presented in pub-
lished reports by physicians who do use them (Zrinzo, Zrinzo, Tisch et al. 2008). Such
claims typically rest on the failure to demonstrate statistically significant differences
via standard hypothesis testing. Yet the truth is that the sample size was too small, and
for this reason the reports containing them do not permit interpretation (Montgomery
2012). Patients having the same neurosurgeon were never randomized according to
their microelectrode recording status. As a result, unfair becomes any comparison
made between the results produced by neurosurgeons who make such recordings and
the results of the neurosurgeons who do not, because the second category does not
exclude the possibility of better outcomes had the surgeon occupying that category
made use of intraoperative microelectrode recordings.
Though imperfect and requiring acceptance of significant assumptions, one study
attempts to shed light on the necessity of intraoperative microelectrode recordings by
examining the spatial variability of the physiologically defined optimal location in the
subthalamic nucleus. The physiologically defined optimal location possesses the fol-
lowing characteristics: it covers at least 5 mm of sensorimotor representation; micro-
stimulation produces in it no adverse effects; and it is attended by improving symptoms
in patients with Parkinson’s disease. In order to translate the observations of the spatial
variability of the physiologically optimal target to clinical meaningfulness, one must
assume that the physiologically defined optimal target is a reasonable surrogate for the
clinically optimal one. The results indicated that much larger than the volume of tissue
activation by DBS was the 99%-confidence volume, that is, the volume of tissue DBS
would have to activate in order to include 99% of the physiologically defined optimal
target (Montgomery 2012). One therefore one needs to look within the 99%-confidence
volume to find the physiologically optimal target for the individual patient (Figure 1.7)
and that is best accomplished by microelectrode recordings.
Volume of Optional Sites
24
22
n
20
Depth dimensio
18
16
18
n
sio
14 16
en
14
m
12
di
2 12
l
ra
0
te
Ante –2 10
-la
rior-p –4
oste –6
ial
rior d –8 8
imen
ed
sion
M
FIGURE 1.7 The volume and distribution of the 99%-confidence volume of the physiologically
defined optimal target relative to the midpoint of the line connecting the anterior commissure
(AC) and the posterior commissure (PC) in the anterior-posterior and medial-lateral dimen-
sions. The radius was 4.5 mm. The sphere shows the approximate volume of typical tissue
activation. Source: Modified from (Montgomery 2012).
Some argue that brain shift introduces the greatest variability in the vertical axis, and
that one could compensate for this by adjusting the location of the volume of tissue acti-
vated by DBS via selection among the electrical contact. Though this assertion does hold
some truth, it overlooks the fact that the variability in the plane orthogonal to the DBS lead
remains greater than the radius of the volume of tissue activation (Figure 1.8). One thus
finds himself obliged to search in the plane orthogonal to the long axis of the DBS lead.
Future developments in functional neuroimaging may allow neurosurgeons to iden-
tify the sensorimotor regions and their homuncular representations. These images they
would obtain preoperatively and merge with intraoperative MRI and or CT, thereby
obviating the need for some forms of intraoperative neurophysiological monitoring.
Until such time, however, microelectrode recordings, risks posed notwithstanding,
will remain necessary to optimal outcomes.
As this chapter seeks to establish, understanding the principles governing optimi-
zation of target localization depends on defining the latter’s requirements according to
anatomical and physiological principles. For example, one principle rests on the prem-
ise that treatment of cervical dystonia, or other dystonia affecting the head, requires
activation of the region of the globus pallidus interna specific to the head’s function.
Confidence Area in the Horizontal Plane

vs. Confidence Length in the Depth
3.2 mm
24
22
n
20
Depth dimensio
7.6 mm
18
16
18
n
sio
14 16
en
14
im
12
ld
12
2
ra
0
te
–2 10
Ante
-la
rior-p –4
ial
oste –6 8
rior d
ed
imen –8
M
sion
FIGURE 1.8 The area and distribution of the 99%-confidence volume of the physiologically
defined optimal target ion viewed in the plane orthogonal to the long axis of the DBS lead
and along the long axis of a trajectory for the DBS lead. The sphere shows the approximate
volume of typical tissue activation. Source: Modified from (Montgomery 2012).
A second premise, which is derived from physiological experience and computational

modeling, holds that the volume of tissue activation surrounding a single cathode (neg-
ative electrode) is approximately 2.5 mm. A third premise holds that, since the volume
of the globus pallidus interna is extremely large relative to the volume of tissue activa-
tion by DBS, one must devise some method for identifying the homuncular representa-
tion of the head.
Indeed, even level 1 Evidence-Based Medicine data, if they existed, would not
relieve one of having to reason from principles. Because they refer to populations and
not individual patients, inferences from randomized controlled trials require that
one extrapolate to the individual patient (Montgomery and Turkstra 2003), and this
requires that she invoke physiological, anatomical, and other principles that exist inde-
pendently of or do not derive from specific randomized controlled trials. Commitment
to individual patient benefits therefore demands that physicians know these fundamen-
tal physiological and anatomical principles.
Issues will arise for which exist no data from randomized controlled trials or funda-
mental principles. Similarly, situations will arise in which principles oppose each other.
Such instances and situations nonetheless require that one decide on a way to resolve
them. The recommendations made here have evolved with many years’ experience and
have come to inform the author’s practice. The experience of this author alone is not
for or against alternatives not addressed. All recommendations in this book the author
makes for educational purposes only. He does not intend that they should direct the
care of any particular patient. Physicians and healthcare professionals should always
base their care decisions on their individual assessment and judgment.
MICROELECTRODE AND SEMI-MICROELECTRODE RECORDINGS
Identifying and analyzing extracellular action potentials arising from individual

neurons among a collection of neurons constitutes the primary purpose of micro-
electrode recordings. Pursuit of this purpose requires use of microelectrodes, such
as those made with tungsten or a platinum-iridium alloy metal, whose fine-tip expo-
sures are on the order of 20 microns (μm) and impedance—some 0.6 to 1 megaohms
(semi-microelectrodes have impedances less than 0.1 megaohm). The difference
between microelectrode recordings and semi-microelectrode recordings turns on the
number of neurons whose extracellular action potentials are contained within the elec-
trode recording (Garonzik, Hua, Ohara et al. 2002). The relatively few neurons involved
in microelectrode recording allow one to identify the extracellular action potentials
from distinct individual neurons and to recognize their individual behaviors, whereas
semi-microelectrode recordings generally do not allow such recognition, because the

density of other neurons being recorded causes interference of a sort resembling that
which one sees on electromyographic recordings.
The experience of standing in a sports stadium presents a useful example. The roar
of the crowd one hears while thus situated is analogous to what one hears with use
a semi-microelectrode, and the ability to isolate individual conversations amid the
roar is analogous to what use of a microelectrode allows one accomplish. Each use has
its advantage. On one hand, the semi-microelectrode’s larger tip exposure and lower
impedance allow for a larger volume of recording, albeit at a resolution lower than the
resolution of which a microelectrode is capable. A semi-microelectrode, therefore,
may not allow identification of specific thalamic nuclei or of neurons within a specific
homuncular representation, and this places it at a disadvantage (Garonzik, Hua, Ohara
et al. 2002). On the other hand, the semi-microelectrode’s lower impedance makes it
less prone to electromagnetic artifact. (Improvement in the quality of modern ampli-
fiers, however, has reduced this advantage.)
LOCAL FIELD-POTENTIAL RECORDINGS
Because local field-potential recordings are not made for the purpose of identify-
ing extracellular action potentials, their electrical recording characteristics are less
demanding. Much larger electrodes—electrical contacts on DBS leads, for example—
typically have much lower impedances, and they make contact with, and record from, a
larger volume of tissue. This means they cannot record extracellular action potentials.
Rather, local field potentials tend to function much like a filter by summing (average)
activities over a wider volume of tissue. As such, they require that some phase syn-
chronization occur among the sources of electrical currents being recorded. Two sine
waves of equal frequency but opposite phases (Figure 1.9A and B), for instance, begin
at zero volts. One initially increases (phase equal to 0 degrees), while the other ini-
tially decreases (phase equal to 180 degrees). The sum or average of these two signals
would be zero everywhere. If such were the case in local field-potential recording, no
signal would result. Two sine waves of equal frequency and phase (Figure 1.9C and D),
however, produce a sum that is not zero everywhere. If such were the case in a local
field-potential recording, a definite signal would result.
Local field potentials depend on synchronization and positive interactions, that is,
interactions whose sum is greater than the constituents (as in Figure 1.9E). Interactions
depend on the duration of the signals, which for postsynaptic dendritic potentials are
longer—on the order of 10 ms—while extracellular action potentials are shorter: on
the order of 1 ms. Multiple and spatially distributed, inputs onto dendrites from the
C
E
FIGURE 1.9 Schematic representation of the effects of synchronization on the local field poten-
tial. A and B show two anti-phase sine waves (phase difference of 180 degrees) that, when added
together, render a sum of zero everywhere. Should sine wave A shift into phase (C) (phase dif-
ference of 0), the two sine waves (C and D) would combine to produce a larger amplitude sine
wave (E). Synchronization of underlying oscillators thus produces positive resonance.
same source tend to result in increased synchronization and increased magnitude of

the summed or averaged responses. Local field potentials consequently tend to empha-
size presynaptic dendritic inputs over extracellular action potentials representing neu-
ronal outputs.
Local field potentials share with the semi-microelectrode the advantage of recording
from a larger volume, which presents few problems with regard to electronics. Yet, local
field-potential recording encounters problems regarding spatial resolution. Multipolar
recordings (simultaneously recording from multiple electrodes) and other techniques
exist that allow one to increase the spatial resolution in such instances where phase rever-
sals in the local field potentials help to localize the source. What remains in question,
however, is whether these techniques achieve the needed degree of spatial resolution.
The uniqueness of the local field-potential signal relative to the structures
encountered in the electrode trajectory also remains in question. Undoubtedly
unique is the pattern of extracellular action potentials in the various structures
encountered in the trajectories to the specific DBS targets (chapters 9–11 contain
detailed discussion of this uniqueness). It has been suggested that increased beta
band–frequency power in the local field potentials correlates with Parkinson’s dis-
ease. This correlation recommends itself as a useful marker for the clinically optimal
target. It remains unclear, however, whether the sensorimotor region emits a unique
signal vis-à-vis the nonsensorimotor region. The ability to differentiate the senso-
rimotor from nonsensorimotor regions of DBS targets perhaps depends on evoked
potentials that derive from local field-potential recordings and that are time-locked
to behavioral activations.
Those who do not use microelectrode or local field-potential recordings must rely
on effects of stimulation—effects of macrostimulation through the DBS lead, typi-
cally—to guide their inferences as to optimal location. Situations arise, however, in
which the effects, particularly improvements in symptoms indicative of proper place-
ment, are unavailable. Improvement in symptoms with the passage of a recording or
stimulating electron through the target is known as a “micro-otomy” effect, such as
a micro-subthalamotomy effect in the case of subthalamic DBS, and it occurs prior
to or absent any electrical stimulation. Simply passing electrodes through the sub-
thalamic nucleus, for example, can result in remarkable yet temporary improvement
of symptoms. Similarly, a micro-subthalamotomy effect can impose a “ceiling” above
which subsequent stimulation fails to result in further improvement. When such an
effect results, the intraoperative neurophysiologist or neurologist can no longer infer
the proper location of the DBS electrode based on symptom improvement; all she can
observe are adverse effects. Absent the use of microelectrode or local field potentials,
the neurophysiologist or neurologist must relieve any adverse motor effects—tonic
contraction owing to the electrical current’s spreading to the corticospinal fibers in
the posterior limb of the internal capsule, for example—because patients are typically
under anesthesia. From occurrence of tonic contraction the neurophysiologist or neu-
rologist can infer a subthalamic DBS lead’s too-anterior placement. As to a too-medial
or too-posterior placement, however, she remains in the dark.
MICROSTIMULATION
One can accomplish microstimulation by passing an electrical current through a micro-

electrode. A microelectrode, however, must be sturdy enough to withstand the stimu-
lation effects. Platinum-iridium microelectrodes tend better to withstand stimulation
effects than do tungsten ones. Current practice involves constant stimulation with a
current of less than 100 microamps. Microstimulation admits of a variety of uses. Some
use it to predict the clinical effects of subsequent DBS by using stimulation parameters
similar to those used in clinical DBS. Others use it for purpose of identifying the physi-
ological location of the microelectrode.
Extremely small by several orders of magnitude, the volume of tissue activa-
tion may render ineffective the use of microstimulation in predicting subsequent
clinical DBS effect. Experience with microstimulation in the vicinity of the optic
tract in globus pallidus interna DBS surgery suggests that 100 microamps activates
a volume of tissue approximately 500 μm in radius (effects of inhomogeneities in
the regional tissue resistivity aside), and that subsequent clinical DBS produced
Indifferent or reference
Electrode connector
Indifferent or reference electrode
Active electrode
Active electrode
connector
FIGURE 1.10 Schematic representation of a bipolar microelectrode. The active electrode is the

tip, and the indifferent electrode is the band of conductive material (typically metal) appearing
approximately several mm from the tip.
gratifying benefit, even if microstimulation had failed to do so. This author therefore
views with skepticism the use of microstimulation in predicting clinically optimal
target sites.
In cases involving use of bipolar microelectrodes (Figure 1.10), stimulation via the
indifferent contact becomes possible. Because the contact is larger than other types
of contact, it resembles macrostimulation more than it does microstimulation. The
large contact area could mean activation of a large tissue volume. Yet, such activation
remains unclear as to its prediction of subsequent clinical response.
Microstimulation is also used to drive physiological responses that indicate
structures in the vicinity of the microelectrode. Production of paresthesias with
microstimulation during subthalamic DBS microelectrode recordings, for example,
suggests a trajectory too close to the ascending medial lemniscus and too posterior.
When used for this purpose the stimulation parameters characteristic of clinical
DBS are not optimally effective. Offering more effective stimulation parameters,
high frequencies—on the order of 300 pps, typically—take advantage of tempo-
ral summation, a phenomenon in which the effects of each subsequent stimulation
pulse build on the lingering effect of the previous stimulation (chapter 8 contains
detailed discussion of temporal summation). False localization, which results from
stimulation of axons in the vicinity of the target as the microelectrode passes them,
confounds microstimulation employed for the purpose of identifying the sensorim-
otor homunculus (Figure 1.5).
Three issues limit one’s using the larger indifferent electrical contact, for example
on a bipolar microelectrode, Figure 1.10, to drive physiological responses: (1) the
achievement of large-volume tissue activation, though it increases the probabil-
ity of a physiological response, sacrifices spatial resolution; (2) the position of the
indifferent electrode, which is typically several millimeters from the recording tip,
requires that one advance the indifferent electrode to the microelectrode record-
ing site in order to correlate with the latter the effects of microelectrode and
semi-microelectrode recordings; and (3) indifferent electrodes vary in size and
structure as a result of different manufacturer’s specifications, and therefore make
necessary some calibration of the effective radius of the volume of tissue activation
(chapter 8 contains discussion of this need for calibration).
MACROSTIMULATION
In the context of intraoperative neurophysiological monitoring, macrostimulation

typically is performed through the implanted DBS lead by use of stimulation con-
figurations, that is, the arrangement of active cathodes (negative electrical contacts)
and anodes (positive electrical contacts). The purpose is to mimic, for purposes of
prediction, the results of subsequent clinical DBS. In the course of so doing there
arise two challenges: (1) demonstration of efficacy and (2) occurrence of adverse
effects.
Important considerations challenge the use of such intraoperative macrostimula-
tion. First, time pressures in the operating room severely constrain the range of elec-
trode configurations and stimulation parameters (the combination of stimulation
voltage or current, pulse width, and frequency) that are available for use in testing and
relevant to postoperative care. Though the stimulation parameters used in the oper-
ating room may prove ineffective, others discovered during the course of subsequent
outpatient DBS may produce a satisfactory benefit.
Second, quite a different electrical context may develop during subsequent out-
patient clinical DBS than attended DBS in the acute phase. Marked changes in DBS
electrode impedances can have a marked effect on the amount of electrical current
introduced into the brain via a constant voltage stimulator (Montgomery 2010).
For example, changes in interstitial fluids associated with the acute microtrauma—
cytotoxic or vasogenic edema, for example—can significantly affect tissue impedance
(chapter 8 contains discussion of these effects on tissue impedance).
Capacitance at the electrode-brain interface can also affect the electrical current in
the event that an altered stimulation pulse waveform is introduced (chapter 8 contains
discussion of these effects). Indeed, the acute changes in impedance described above
may affect capacitance as readily as it may affect resistivity (the converse of permittiv-
ity) or the dielectric constant determining capacitance. These terms and concepts are
explained in detail chapter 3.
Use of constant current stimulation helps to mitigate the effects of changes in

impedance and capacitance (see Montgomery, 2010, and the discussion in chapter 3).
The primary factor in neuronal activation is the current delivered. Inhomogeneities in
the regional resistivity, however, complicate the situation. Specifically, significant dis-
tortions in the shape, size, and current densities—which acute changes associated with
microtrauma only make more severe—can appear as a consequence of the combina-
tions of gray and white matter present in the vicinity of the stimulation, thus making
difficult prognostication from DBS macrostimulation, even in such instances where
constant current stimulation is applied. Chapter 12 discusses inferences and clinical
judgment concerning particularly the issue of where to move a DBS lead if the original
placement produces side effects.
ETHICAL CONCERNS
When making any medical decision, one must weigh the benefits against the risks in
the context of alternatives. And, unfortunately, no rubric exists that provides means
of deciding each individual patient’s case in those terms. Indeed, an already complex
medical decision faces further complication at the hands of the physician, who might
favor her own ethical convictions over the patient’s wishes or the wishes of the patient’s
family members. These convictions might rest on implicit—perhaps even unacknowl-
edged—biases. One common implicit bias, Omission bias, involves deeming errors
of omission worse than errors of commission. The well-known “Runaway Trolley
Car dilemma” illustrates this bias in action (Thomson 1976). It involves a situation in
which a trolley has gotten free of its driver’s control and is speeding downhill toward a
group of five unsuspecting pedestrians. There is a switch that can divert the trolley, but
there is a single pedestrian on that track who would be killed. A bystander observes
that, if she acts quickly, she can save five of the endangered pedestrians; the other one
will have to perish. Most persons would pull the lever to throw the switch. Consider
another scenario where the agent is standing on a bridge over the tracks. Standing
next to her is a large man with a large backpack. The agent knows that if she pushed the
large person over the bridge, the large person would land on the tracks, derailing the
trolley, saving the five pedestrians but at the death of the large person on the bridge.
Most persons would not push the large person off the bridge. The net result, in terms
of lives lost, is the same and the agent cannot argue that her actions did not determine
the results. Somehow there is something very different about throwing a switch and
pushing a large person off the bridge. In some ways, the dilemma described above
plays a role in a physician’s approach to DBS. Chapter 14 contains detailed discussion

of ethical issues related to intraoperative neurophysiological monitoring.
Whereas the microelectrode’s sharp tip cuts tissue in its path, the blunt end of the
DBS lead tends to dissect any tissue and push it away. One might conclude from this
that microelectrode recording increases the risk of hemorrhagic complications. Though
such an assumption is consistent with the author’s observation that most hemorrhages
occur during the microelectrode’s final 25-mm traversal, the length through which the
microelectrode moves, this risk must be weighed against the risks of DBS leads’ poor
placement and need for subsequent surgical revisions.
Subtler risks await patients, physicians, and healthcare professionals after surgery.
Patients, physicians, and healthcare professionals experience what is known as the
“tyranny of partial improvement,” that is, a less-than-expected benefit in cases where
the DBS lead was otherwise optimally placed. As with most symptomatic therapies,
the potential benefit relates directly to the severity of the symptoms weighed against
the probability of success. An improperly placed DBS lead that slightly improves
symptoms reduces the potential benefit of a revised and properly placed lead and thus
reduces also the likelihood of a misplaced DBS lead’s replacement. Compounding
this problem is a patient’s reluctance to undergo any subsequent surgery. Her confi-
dence shaken, the patient may simply choose to abide the poor results and continue
a regimen of medications whose failure led her to seek surgical treatment in the first
place.
Physicians and healthcare professionals providing postoperative care may also
have their confidence shaken when confronted with less-than-satisfactory responses in
patients. While puzzling over these responses, physicians and healthcare professionals
might wonder whether they simply did not find the right combination of electrode con-
figurations and stimulation parameters (these combinations number in the thousands)
and pursue further programming or did not place the DBS lead in the optimal location
(this second error makes future effort and expense incidental to pursuing the desired
result unjustified). Physicians and healthcare professionals’ confidence in proper DBS
lead placement remains highly important. Every effort must be made intraoperatively
to ensure this confidence.
Lack of strenuous effort made to ensure such confidence smacks of complacency.
Complacency is an ethical issue, and physicians and healthcare professionals bear
the responsibility of recognizing and forfending against any risk of slipping into it.
Decades of experience with intraoperative neurophysiological monitoring (micro-
electrode recordings included) and the excellence of many of the devices and sys-
tems for intraoperative neurophysiological monitoring together constitute a social
hazard. In the past, most DBS surgeries and their attendant intraoperative neuro-
physiological monitoring were performed at university-affiliated medical centers
primarily by academic clinician-scientists whose already substantial interest in the
subject motivated a deeper understanding of methods and the science. Today, an
increase in the number of centers offering such surgery and monitoring has led to an
increase in the number of patients who stand to benefit, as well as an increase in the
responsibility borne by the physicians and healthcare professionals involved. Where
scientific and academic interest is lacking ethical obligation must suffice, which
means that intraoperative neurophysiologists must gain the expertise of their aca-
demic clinician-scientist colleagues. As Alexander Pope wrote in his famous poem
“An Essay on Criticism” (1709), “A little learning is a dang’rous thing.” Insufficient
understanding engenders risk. Preventable problems arise unforeseen. Habit passes
for knowledge, and some undesirable outcome goes unrecognized or is dismissed as
an anomaly for which the physician bears no responsibility. All of this culminates in
the defeat of any effort toward quality control.
Indeed, even when efforts meet with success and little complication, complacency
can result. Localization of the physiologically defined optimal target location based on
image-guided navigation, for example, is on the order of 70% accurate (Montgomery
2012). Of 144 cases of subthalamic DBS studied, 100 involved instances in which
the first trajectory led to the physiologically defined optimal target, as suggested by
the image-guided navigation. Though this finding leads one to conclude that more
complicated microelectrode recordings could be avoided for a majority of patients,
macrostimulation through the DBS lead and other types of intraoperative neuro-
physiological monitoring should nonetheless be done, because for 30% of patients
the image-guided navigation would fail to direct the DBS lead to the physiologically
defined optimal target. Admittedly, this analysis cannot ensure that one need neces-
sarily place the DBS lead at the physiologically defined optimal target in order to
produce an optimal clinical outcome. The absence of any other predictive measure,
however, recommends targeting for the physiologically defined optimal target as a
reasonable procedure.
Having established that the initial trajectory, as suggested by image-guided navi-
gation, does not reach the physiologically defined optimal target, and that micro-
stimulation does not produce adverse effects, one must choose her next move. The
intraoperative neurophysiological monitoring in the study described above is based on
two items: (1) intraoperative microelectrode recordings and (2) an algorithm devised
to determine, in cases of errant initial trajectory, the target’s probable location (Baker,
Boulis, Rezai et al. 2004). Of the 44 cases involving an errant initial trajectory, 30 cases
required a single additional pass of the microelectrode, 11 required two, and 2 required
three. (One patient was excluded because an intraoperative hematoma prevented fur-
ther recordings.)
Reasonably robust, the algorithm used to determine where to search next for the target
is based on two items: (1) an understanding of the physiological anatomy in the vicinity
of the subthalamic nucleus and (2) the use of microelectrode recordings. Understanding
of the regional physiological anatomy thus recommends itself, though it depends on the
interpretation of the microelectrode recordings. Only a deep understanding of the mak-
ing and interpretation of those recordings inspires one to have confidence in them.
Microelectrode recordings may also occasion complacency. Microelectrode
recordings must contend with the issue of whether the signal’s rising above the
background indicates a real extracellular action potential or an artifact. This
author’s experience has shown him that rare is the case in which this issue does
not arise. Indeed, one commonly must modify the electrical environment in order
to minimize artifact and noise. Otherwise, the surgery cannot proceed. One can
accomplish this rather easily (chapter 6 contains discussion of how this is done). Yet
more frequently than one might imagine there arise problems of greater complex-
ity, and these require investigation aided by a deep understanding of biophysics and
electronics.
Though efficient and robust, turnkey systems—systems consisting of seamlessly
integrated components (Figure 1.11)—conceal a certain danger. Currently available
F G
C D A to D Computer
E system
converter
B
FIGURE 1.11 Schematic of a typical microelectrode and semi-microelectrode recording sys-

tem. A represents the source of the neural signals, B is the electrode, C the high-impedance
probe (unity gain amplifier) used for impedance matching (discussed in chapter 5), D the
amplifier, E the filtering systems, F the analog to digital (A to D) converter (note in systems
using digital signal processing systems, the A to D conversion occurs before much of the fil-
tering, which is done digitally), G the computer system for analyses, H the visual display, and
I the audio presentation.
commercial FDA-approved systems improve considerably on the “home-built” sys-

tems in use during the early days of intraoperative neurophysiological monitoring
for DBS surgery. Yet, “home-built” systems possessed the distinct virtue of being
familiar to those individuals who made them and who could therefore identify and
fix their problems. Manufacturers’ response to problems with their turnkey systems,
as prompt and expert as this might be, is no substitute for the situation of having the
engineer who designed the system on hand during surgery to address any problems.
The intraoperative neurophysiologist must adopt a healthy but respectful criti-
cal attitude toward the neurosurgeon, whom she should consider simply another
instrument. As such, precision, accuracy, failure rate, and other concerns apply to
the neurosurgeon as much as to any other instrument. (For their part, the neurosur-
geon and the patient’s treating nonsurgical physician ought to adopt exactly the same
attitude toward the intraoperative neurophysiologist.) Correct image-guided naviga-
tion is something that the neurosurgeon cannot simply assume, because it greatly
affects the neurophysiologist’s ability to interpret the intraoperative neurophysi-
ological monitoring. As discussed in chapters 9 and 10, microelectrode recordings of
the globus pallidus interna are nearly identical to those of the subthalamic nucleus.
Incidents have happened in which image-guided surgical navigation determined an
initial trajectory situated on the wrong side of the posterior limb of the internal cap-
sule. This led to the subthalamic nucleus’s being mistaken for the globus pallidus
interna. Other incidents have occurred in which exactly the opposite mistake was
made. In those cases, microstimulation, macrostimulation, and other clues produced
results directly opposite to those one would expect had the electrode entered the
subthalamic nucleus instead of the globus pallidus interna. Beginning anew after
many hours of surgery would thus prove difficult. The intraoperative neurophysiolo-
gist must therefore understand all of the various aspects of DBS surgery and must be
able to recognize the signs of any potential mishap.
SUMMARY
So important is intraoperative monitoring that one finds the idea of performing DBS
without it difficult to conceive. Though this monitoring may only take the form of macro-
stimulation through the DBS lead, it nonetheless requires considerable knowledge in
order to ensure an optimal outcome. Fundamental scientific principles inform all meth-
ods of intraoperative neurophysiological monitoring, and an understanding of these prin-
ciples enables the intraoperative neurophysiologist to offer her patient the greatest hope
for benefit.
The question whether invasive methods of intraoperative neurophysiological

monitoring—microelectrode recordings in particular—merit application admits of
no easy answer. However, discussion on these questions should include every stake-
holder, including the nonsurgical referring physician and her patients, their family
members, and their caregivers. Indeed, the choice of to whom to refer for surgery
is a choice of the surgical techniques to be employed. After attending to this issue
with d iligence, this author has come to favor the use of intraoperative microelec-
trode recordings (although he remains open to alternatives). Any intelligent decision,
whether it involves intraoperative neurophysiological monitoring or some other issue,
requires that the decision maker appreciate the many nuanced and complex factors
bearing on it.
REFERENCES
Baker KB, Boulis NM, Rezai AR, et al.: Target selection using microelectrode recordings. Microelectrode
Recordings in Movement Disorders Surgery. Z. Israel and K. Burchiel. New York, Thieme Medical
Press, 2004: 138–151.
Butson CR, Cooper SE, Henderson JM, et al.: Patient-specific analysis of the volume of tissue activated
during Deep Brain Stimulation. Neuroimage 34(2): 661–670, 2007.
Fins JJ: Surgical innovation and ethical dilemmas: precautions and proximity. Cleveland Clinic Journal of
Medicine 75(Suppl 6): S7–S12, 2008.
Garonzik IM, Hua SE, Ohara S, et al.: Intraoperative microelectrode and semi-microelectrode record-
ing during the physiological localization of the thalamic nucleus ventral intermediate. Movement
Disorders 17(Suppl 3): S135–S144, 2002.
Grill WM, Simmons AM, Cooper SE, et al.: Temporal excitation properties of paresthesias evoked by
thalamic microstimulation. Clinical Neurophysiology 116: 1227–1234, 2005.
Guridi J, Gorospe A, Ramos E, et al.: Stereotactic targeting of the globus pallidus internus in Parkinson’s
disease: imaging versus electrophysiological mapping Neurosurgery 45: 278–289, 1999.
Hilker R, Voges J, Ghaemi M, et al.: Deep brain stimulation of the subthalamic nucleus does not increase
the striatal dopamine concentration in Parkinsonian humans. Movement Disorders 18: 41–48, 2003.
Montgomery EB, Jr.: Effect of subthalamic nucleus stimulation patterns on motor performance in
Parkinson’s disease. Parkinsonism and Related Disorders 11(3): 167–171, 2005.
Montgomery EB, Jr.: Microelectrode targeting of the subthalamic nucleus for Deep Brain Stimulation
surgery. Movement Disorders 27(11): 1387–1391, 2012.
Deep Brain Stimulation Montgomery EB Jr.: Microelectrode recordings in DBS—still in need of rea-
soned discussion. Movement Disorders 28(2): 255, 2013.
Montgomery EBJ: Deep Brain Stimulation Programming: Principles and Practice. Oxford, Oxford
University Press, 2010.
Montgomery EBJ and Turkstra LS: Evidenced based medicine: let’s be reasonable. Journal of Medical
Speech Language Pathology 11: ix–xii, 2003.
Olanow CW, Goetz CG, Kordower JH, et al.: A double-blind controlled trial of bilateral fetal nigral trans-
plantation in Parkinson’s disease. Annals of Neurology 54: 403–414, 2003.
Schaltenbrand G and Wahren W: Atlas for Stereotaxy of the Human Brain. Stuttgart, Thieme, 1977.
Schultz W: Predictive reward signal of dopamine neurons. Journal of Neurophysiology 80: 1–27, 1998.
Thomson JJ: Killing, letting die, and the trolley problem. Monist 59(2): 204–217, 1976.
Valenstein ES: The War of the Soups and Sparks: The Discovery of Neurotransmitters and the Dispute over
How Nerves Communicate. New York, Columbia University Press, 2005.
Weaver FM, Follett K, Stern MB, et al.: Bilateral Deep Brain Stimulation vs best medical therapy for
patients with advanced Parkinson disease: a randomized controlled trial. Journal of the American
Medical Association 301: 63–73, 2009.
Weaver FM, Follett KA, Stern M, et al.: Randomized trial of Deep Brain Stimulation for Parkinson dis-
ease: thirty-six-month outcomes. Neurology 79(1): 55–65, 2012.
Wellek S: Testing Statistical Hypotheses of Equivalence and Noninferiority. Boca Raton, CRC Press, 2010.
Yelnik J and Percheron G: Subthalamic neurons in primates: a quantitative and comparative analysis.
Neuroscience 4: 1717–1743, 1979.
Zrinzo L, Zrinzo LV, Tisch S, et al.: Stereotactic localization of the human pedunculopontine
nucleus: atlas-based coordinates and validation of a magnetic resonance imaging protocol for direct
localization. Brain 131(Pt 6): 1588–1598, 2008.
/ / / 2 / / / PREPARATIONS FOR
INTRAOPERATIVE
NEUROPHYSIOLOGICAL
MONITORING
INTRODUCTION
The intraoperative neurophysiologist must understand the methods involved in

stereotactic functional neurosurgery, with special attention given to the method
of determining an electrode’s trajectory. The first trajectory of the microelec-
trode, semi-microelectrode, or deep brain stimulation typically is based on radio-
logical images, most commonly magnetic resonance imaging (MRI). Among other
things, trajectory determines which particular structures the intraoperative neuro-
physiologist is likely to encounter, enabling the latter to prepare herself should she
encounter structures other than those expected. Important for microelectrode or
semi-microelectrodes use, these issues also carry implications for intraoperative
macrostimulation with DBS electrodes.
A number of neurosurgical issues concern the trajectory. An intraoperative neu-
rophysiologist wishes to avoid passing the electrodes through what are known as elo-
quent areas—areas whose injury could result in major disabilities. She thus determines
the trajectory accordingly. Seeking to avoid penetration of the motor cortical areas, she
finds it necessary to enter at a site anterior to the coronal suture. Entering at this site
often results in an angle which could have a significant impact on the ultimate place-
ment of the DBS lead in subthalamic nucleus, the globus pallidus interna, or the ventral
intermediate nucleus of the thalamus.
Many surgeons try to avoid penetrating the lateral ventricles in order to negate two
possibilities: (1) deflection of electrodes as they traverse the ependymal surface en
30
2. Preparations for Intraoperative Neurophysiological Monitoring / / 31
route to the brain parenchyma and (2) bleeding in any of the number of vascular struc-
tures present in the ventricles—the choroid plexus and thalamostriate veins residing
in the sulcus between the caudate and thalamus on the lateral wall of the lateral ven-
tricle, to name two. In cases of significant ventricular enlargement, avoiding the lateral
ventricles could result in an extremely shallow trajectory in the coronal plane. Such a
trajectory significantly affects regional anatomy relative to, and thus the ultimate clini-
cal utility of, DBS. (The particular effects are discussed in the chapters devoted to each
target.) Cases in which the subthalamic nucleus is the target risk limitation of post-
operative DBS management owing to tonic contraction, because a trajectory begun at
an extremely lateral entry point may traverse only the posterior limb of the internal
capsule before entering the subthalamic nucleus.
Trajectories planned by the intraoperative neurophysiologist bypass the sulci, whose
many blood vessels pose a serious bleeding risk. Upon creating a burr hole and opening
the dura, one may visually observe these sulci and the vessels intercalated therein. The
extremely convoluted structure of the cerebral cortex, however, permits one to traverse
more deeply residing sulci not visible on the surface. The capacities of most image-based
navigation systems are such that they can reveal the anatomy surrounding the trajectory,
including the portion sometimes referred to as the “probe’s view.” Some surgeons will
obtain contrast-administered MRI scans prior to establishing the surgical frame or fidu-
cial markers in order to visualize and thus minimize the risk of traversing a blood vessel.
TARGETING METHODS
Because the intraoperative neurophysiologist never visually observes the DBS target, tar-
geting requires the use of other landmarks and spatial coordinate systems to relate the
landmarks to the manipulators for the electrode system. (A generic term, “electrode sys-
tem” denotes the types of electrode used for recording or stimulation—everything from
single-tip microelectrodes to macroelectrodes housing multiple contacts). There exist two
types of landmark: internal and external. Registration of external landmarks occurs by
use of neuroimaging under two conditions: (1) with the stereotactic frame or (2) with a
frameless system, both of which utilize fiducials. These fiducials are markers that appear in
the MRI or computerized tomographic (CT) scans simultaneously with visualization of
the internal landmarks. Registration of internal landmarks occurs by use of neuroimaging
under either condition. Each imaging technique possesses certain advantages. The MRI
scan—T1 weighted images, particularly—can achieve high spatial resolutions, and T2
weighted images make best use of the paramagnetic effects in the red nucleus and substan-
tia nigra pars compacta, which lie in the vicinity of the subthalamic nucleus and the globus
pallidus. Indeed, some systems permit one continuously to shift between or merge T1
and T2 images. Inhomogeneities in the magnetic fields of the MRI, however, may distort
images. Less susceptible to distortion, the CT scan does not offer the spatial resolution or
contrast of the various nuclei that the MRI scan achieves. Some centers therefore use both
MRI and CT, the second acting as a check on any spatial distortions produced by the first.
The most typical internal landmarks, the anterior and posterior commissures,
AC and PC respectively, are easily found—the first a bit more readily than the sec-
ond, perhaps (Figure 2.1). The AC-PC line—a line connecting the anterior and pos-
terior commissures—serves as the reference plane for a Cartesian system, that is, a
three-dimensional system defined by distances along three orthogonal axes, X, Y
and Z, each situated at right angles to the others (Figure 2.2). The midpoint of the
AC-PC line forms the origin of the three axes in the Cartesian system. This mid-
point has a value or distance of zero on each of the three axes. The horizontal plane
orthogonal to the AC-PC line forms the medial-lateral–anterior-posterior, or X-Y
plane, with Y indicating the anterior-posterior direction and X the medial-lateral
position. (One should be aware, however, that conventions governing these indica-
tions may differ.) Orthogonal to the X-Y plane at the midpoint of the AC-PC line
lies the vertical, or Z, axis. Initial targeting is based on distances along the X, Y and
Z axes that subsequently undergo conversion to settings for the system, be it frame-
based or frameless, to which are attached manipulators carrying the electrode sys-
tems (Figure 2.3).
A
A
R
AC
PC
FIGURE 2.1 The image on the left shows an axial section of an MRI scan showing the anterior
and posterior commissure in the image on the left. The representation of the head shows
the orientation as though looking down at the top of the head though the right side of the
MRI scan corresponds to the left side of the brain. The anterior commissure (AC) and the
posterior commissure (PC) appear as strands of brain tissue that cross the midline. The white
spots around the circumference (A) are the fiducial markers that facilitate co-registration of
the external and internal landmarks. The fiducial markers are part of the frame attached to the
head, as shown schematically on the image to the right.
FIGURE 2.2 Axial section of an MRI scan showing the AC and PC. An expanded view shows
the three-dimensional Cartesian coordinate system with the y axis in the anterior-posterior
dimension, the x axis in the medial-lateral dimension, and the z axis in the dorsal-ventral
direction. The representation of the head shows the orientation of the head relative to the
images.
Alternatives to the frame system described above exist as shown in Figure 2.4. In
these systems, the there is no frame that is fixed to the head by pins that press through
the skin. Rather, a set of screw-like fiducials are screwed into the outer table of the
skull. A patient then has an MRI or CT scan with the fiducials in place. Just as the
internal landmarks (AC and PC) are found and related to calibration on the frame, as
described above, the internal landmarks are referenced to the fiducials in the frame-
less system. The frameless systems are much better tolerated by the patient and often
reduce operating room time, thereby rendering the surgery more cost efficient.
AC
PC
FIGURE 2.3 Schematic representation of transposing the position of the midpoint of the AC-PC
line (shown by the X in the figure to the left) seen on the MRI scan to a position relative to
the head and the external frame. As can be seen, the frame is calibrated and the positions of
the fiducials (A) are known relative to the frame. Now the midpoint of the AC-PC line can be
determined on the calibrated frame.
FIGURE 2.4 Representations of a frameless system. The CT scan on the left shows the recon-
structed image of the skull. Screw-like fiducials are placed through a small incision into the
outer table of the skull. The arrow points to a fiducial in the reconstructed image to the left.
The patient either has an MRI scan with the fiducials or an MRI scan previously obtained is
merged with the CT scan containing images of the fiducials. The internal landmarks, such as
the AC and PC, are then registered with the fiducials (arrow in the image to the right). In this
case, a custom fabricated frame is made that mates with the fiducial screws and the frame
contains mounting rings that allow manipulators to be attached which then direct the elec-
trodes to the target (courtesy of Greenville Neuromodulation Systems, Inc., Greenville PA).
Variables X, Y, and Z often denote different systems. For example, the X, Y, and
Z of the Cartesian system for the stereotactic space (oriented relative to the AC-PC
line) may differ from the horizontal and vertical planes in which may operate manipu-
lators carrying the electrode system. The convention is to designate these planes for
the manipulators with lowercase x, y, and z. (Differences between the X, Y, and Z and
x, y, and z systems are discussed below.) Many surgeons use millimeters or some other
absolute measurement. Some divide the length of the AC-PC line in twelfths. The
resulting unit length, which becomes some multiple of 12, allows one to account for
variability across subjects.
Many surgeons will modify the initial targeting originally determined according
to prescribed coordinates in the AC-PC line space. For example, they may make allow-
ances for a very wide third ventricle. Some among them will adjust initial targeting
according to the location of other structures rendered visible by T2 weighted MRI
scans. Slightly lateral and a bit anterior to the red nucleus, the subthalamic nucleus, for
example, often appears on T2 weighted images by virtue of a paramagnetic effect on
the MRI signal owing to the red nucleus’s greater iron content. A similar paramagnetic
effect foregrounds the globus pallidus against the internal capsule and putamen.
Some surgeons attempting direct visualization of targets—the subthalamic
nucleus, particularly—use special MRI sequences to do so. An external frame-based or
frameless system in place, however, often makes performing these sequences difficult.
Surgeons overcome this difficulty by obtaining MRI scans preoperatively. These they
can merge with those scans (MRI or CT) obtained with the frame-based or frameless
system in place.
FRAME-BASED AND FRAMELESS STEREOTACTIC SYSTEMS
These external systems serve two important purposes: (1) providing an external,
observable coordinate system capable of co-registering internal landmarks and the
internal coordinate system; and (2) providing stability needed to minimize movement
relative to the brain during implantation of micro-, semi-micro- and macroelectrodes
(a DBS lead, for example). The brain either moves with the frame in a frameless system,
or moves not at all in a frame-based system, other than movement possible in the sub-
arachnoid space.
The frame-based system consists of a cage-like device known as a halo that is affixed
to the head by sharp screws driven into the skull (skin at the screw sites receives local
anesthesia prior to the screws’ insertion). At some centers, patients are sedated with an
agent whose effects can be reversed quickly in order to avoid interfering with subse-
quent microelectrode or semi-microelectrode recordings. In order to avoid problems
associated with administering propofol and to eliminate the necessity of having an
anesthesiologist in attendance, physicians at some centers use midazolam, believing its
effects quickly reversible. Midazolam’s plasma half-life may be short, but its effects on
the brain can last for some time, as hours-long changes in the EEG have shown. (This
EEG evidence is further discussed below.)
Once in place, the frame has placed over it a box-like device bearing fiducials vis-
ible on MRI or CT scans (Figure 2.1). This device facilitates co-registration of fidu-
cial markers and internal landmarks. Upon completing scans, physicians remove the
box-like device, while in the operating room the surgeon puts in place the hardware
system used for mounting and co-registering the manipulators for electrode placement.
Small fiducial markers placed on the scalp and often anchored to the bone char-
acterize frameless systems (Figure 2.4). The fiducial markers in place, MRI and/or
CT scans are obtained. Then, internal landmarks and the fiducials are co-registered.
Fiducials in some systems also serve to anchor the electrode-implantation devices and
to direct them to the target without need for an intraoperative MRI scan.
Fixed to the burr hole rather than anchored to fiducials, the platform for manipu-
lating the electrodes in other systems alters the process of intraoperative registration,
which in this case occurs typically by use of an ultrasonic or luminescent locator system
to triangulate the fiducials’ position vis-à-vis the platform. Able to project the trajectory
onto MRI scans obtained with the fiducials in place, these systems typically include a
stage for translocating the electrode systems by use of spherical coordinates in tandem
with another system dedicated to translocating the electrode systems in the horizontal
plane. (This use of the two systems in tandem is discussed in greater detail below.)
Frame-based systems are attached to the operating table and are therefore rela-
tively fixed. Frameless systems, on the other hand, permit some movement, which a
hard cervical collar serves to limit. Thus, though patients better tolerate them, frame-
less systems produce more movement artifact than do frame-based systems. Yet greater
concern with emesis attends the latter, because patients’ inability to turn their head may
occasion aspirations.
METHODS OF TRANSLOCATION
The interoperative neurophysiologist must understand the mechanisms for manipu-

lating electrode systems within the brain, because responsibility for determining the
direction and magnitude of any movement of the electrode systems often falls to her.
The multiple devices for electrode-system translocation (linear translation and rota-
tion) featured by these systems will hereafter be termed primary, secondary, and (if
applicable) tertiary. The first to be determined and fixed on the target, primary translo-
cations address three-dimensional space. Secondary and tertiary translocations permit
adjustments made subsequent to the initial targeting entail translocations in a single
plane. Each device incorporates either translocations in the Cartesian x, y, and z coordi-
nates (use of lowercase letters serves to indicate positions on the system, and applies to
both frame-based and frameless system) or a rotation system based on spherical coor-
dinates. Manipulator position systems often include both types of coordinates.
Rotational devices employ spherical coordinates whose spatial positions are
determined by a length (or vector) and two angles, ϕ and φ, relative to a reference
line (Figure 2.5). First consider the more intuitive situation of a two-dimensional
system (Figure 2.5a but disregarding angle φ). In a Cartesian system, each and any
point in space can be specified by the position of the point on the x and y coordinates
or axes. Note that the two axes, x and y, are orthogonal, which means in Cartesian
two-d imensional space, the axes are perpendicular, or at right angles. In a two
p-dimensional polar coordinate system, each and every point in space can be described
by a length, L, and an angle, ϕ, relative to a reference axis. In polar coordinates the
notion of orthogonality is more problematic. Each of these coordinate systems can be
expanded to three dimensions. In the case of the Cartesian system, going from two to
(a) (b)
φ φ
L L
ϕ ϕ
FIGURE 2.5 An example of a spherical coordinate system in which a vector of length L and

two angles ϕ and φ, relative to a set of reference lines (A), may reference any point in space.
Typically, a restricted range of angles and lengths in use creates a cone of brain volume acces-
sible to electrode systems (B). Arranged in such manner as to situate the cone’s vertex over
the target, some systems require that the target sit at the center of rotation, that is, the point
about which the angles move (see Figure 2.6). Other systems whose vertex or center of rota-
tion sits above the skull delimit an accessible brain volume that occupies the center of rotation
above the skull (see Figure 2.6).
three dimensions involves adding another axis or plane (where lines become planes)
that is orthogonal to the two original planes to create the axis, z. Similarly, in three-
dimensional spherical coordinate systems, another axis of rotation, φ, is added that
rotates in a plane orthogonal to the plane sweep by the first angle, ϕ.
Rotational systems generate a volume that electrode systems can access. This vol-
ume appears as a cone whose apex (or vertex) can be positioned over the target as is the
case for frame-based systems (Figure 2.6). However, some systems invert the cone such
at the apex is above the skull (Figure 2.7).
Cone placement determines which of two general types of rotational device will be
used. Use of the first type depends on situating the cone’s apex at the target, the origin
of the spherical coordinate system. In such an instance, changes in the two angles do
affect radii trajectories, but all converge on the target occupying the apex (Figure 2.6).
From this first use comes a significant advantage: Errors in setting the two angles
(known as the ring and arc in some systems) have virtually no negative effect, because
electrode systems unfailingly meet the target.
The most common implementation involves mounting the above-mentioned rota-
tional system on another that can move in the three dimensions that correspond to a
Cartesian coordinate system. The intraoperative neurophysician moves, relative to the
head, the box housing the rotational system in order to situate the apex of the cone—the
origin of the spherical coordinate system—on the target (Figure 2.8). The lowercase let-
ter values thus specify the following items: x, y, and z, the box system; L, ϕ; and φ. Values
x, y, and z together establish the origin of the spherical coordinate system. Translocations
of the box-like device represent the primary method of translocation, while the rotations
of the arc and ring corresponding to angles ϕ and φ constitute the secondary method.
(a) (b)
ϕ φ ϕ φ
L ϕ L ϕ
PC
Center of φ AC
φ rotation
FIGURE 2.6 A schematic representation of systems that place the center of rotation at the
target (target-centered systems) typically by use of an arc (in A) occupying the coronal plane.
The arc is situated in such a way that a line connecting its two ends passes through the brain.
The situation of the electrode system manipulators is such that the electrode system pur-
sues a trajectory orthogonal to the tangent of the arc (the trajectory described by spokes in a
wheel, for example), whose position is denoted by φ. The arc also rotates at its ends, which
occupy the sagittal plane and are denoted by ϕ. Any constant length L of the electrode system
conduces to the same outcome: arrival of the electrode system’s tip at the center of rotation,
the particular values of ϕ and φ notwithstanding. This single outcome allows the surgeon to
adjust the point of entry according structures she has identified on the brain’s surface. One
observes in B the method, which consists of placing the frame system in such a way that the
center of rotation (the vertex of the cone of accessible brain) comes to sit a specific distance
from the midpoint of the line connecting the anterior commissure (AC) to posterior commis-
sure (PC) (see Figure 2.4). This distance corresponds to the spatial coordinates of the target.
Center of
rotation
ϕ
Y
L
Y
PC
Y
Y
AC
Fiducial
FIGURE 2.7 Schematic of systems in which the center of rotation sits above the brain. The
cone of accessible brain results from change in the spherical coordinate systems, angles ϕ
and φ. These systems make use of fiducial markers affixed to the skin or scalp. MRI or CT
scans obtained with the fiducials in place precede co-registration of the latter and internal
landmarks of the AC-PC line. Operators affix the rotational system carrying manipulators for
the electrode systems to a burr hole in the skull and then triangulate its position by use of the
fiducials. These systems typically employ an additional system that moves in the horizontal
plane and thus serves as a method of secondary translocation (see Figure 2.9).
ϕ ϕ φ
φ
L L
ϕ ϕ
z(
z(
do
do
rs
rs
al-
Β
al-
ve
ve
nt
nt
φ l) φ l)
ra
era era
ra
t t
l)
l-la
l)
ia ial-la
y(
ed
y(
med m
an
x(
an
x(
t.-
t.-
po
po
st.
st.
)
)
FIGURE 2.8 Schematic representation of the systems for targeting by use of a target-centered
rotational system (indicated by the star) in such manner that it brings the center to rest on
the target (see Figure 2.6). As indicated by the box (figure on right), the system includes a
rotational system mounted on a Cartesian system. One observes that the center of rotation
becomes a position in the x, y, and z coordinate system of the box. (This system is different
from the X, Y and Z coordinate system based on the midpoint of the AC-PC line.) Movement of
the center (figure on right) is accomplished by moving the entire rotational system within the
box. This has the effect of specifying different values of x, y, and z.
By moving the box-like device containing the manipulators one can effect translo-
cation in the horizontal. These translocations lie in the x-y plane of the box and not in
the X-Y plane of the AC-PC spatial system. Movement in the x-y plane measuring 1 mm
in distance and direction therefore does not necessarily translate to a 1-mm movement
in the X-Y plane.
Because they find translating the box-like device cumbersome, some surgeons use
instead a single-plane x-y manipulator or offset-rotating discs mounted on the arc sup-
porting the electrode system manipulator (Figure 2.9). Such substitution represents
a tertiary method of translocation. It is critical that one appreciate two tendencies:
(1) this horizontal plane, x-y, is unlikely to share an orientation with the X-Y plane
based on the AC-PC line; and (2) any changes in the angles ϕ and φ change the orien-
tation of the x-y horizontal manipulator relative to the X-Y plane. The fact that most
surgeries involve relatively limited movements of manipulators in the x-y plane serves
to reduce disparities. Translocations of the manipulators following the first trajectory
are relative to the first trajectory. The impact of disparities between the x-y plane of
the manipulators and the X-Y plane based on the AC-PC line are thus minimized. The
x-y plane of the tertiary translocations is in the plane orthogonal to the long axis of the
electrode system.
(a) (f) (g)
x
Electrode
system
(b)
(c) (d) (e)

E* D*
FIGURE 2.9 Schematic representations of two methods for translating the volume of tissue
accessible by electrode system (represented by the tall vertical cylinder in A, B, F, and G) in
the horizontal plane. The first system consists of a stage that is movable in the x and y axes.
These axes are orthogonal to each other in the manner of a Cartesian system (A). Another sys-
tem consists of two circular discs or cylinders that can rotate (B). The smaller disc or cylinder
rotates on the larger disc or cylinder, but its center of rotation is offset relative to the center
or rotation of the larger disc or cylinder (C). The smaller disc contains the electrode system
whose long axis is orthogonal to the surface of the disc or cylinder. Two factors in combina-
tion determine the position of the electrode in the horizontal plane based on the degree of
rotation of both discs (D* and E*). Positions available are shown in D and E. The position of
the electrode system in the horizontal plane rests on one arc of a series. The circular burr hole
isolates a series of cylinders of accessible brain (one assumes in this case that there is in use a
secondary translating system in the horizontal plane). Potential cylinders relative to the target
represented by a star are shown for a target-centered system (F) and for a system in which the
center of rotation sits above the head (G).
Another primary translocation system employs a rotational system whose cone

apex—the origin of the spherical coordinate system—is mounted over the skull of the
patient (Figure 2.7). Designed to “project” the base of the cone ventrally into the brain,
this system situates the target at a point on the base of the cone (in this case the base
of the cone is not planar but a portion of a sphere). Changing the two angles at the
manipulator effects translocation horizontally (in the plane of the base of the cone).
Movement on this horizontal plane (imagined for simplicity’s sake as a flat horizontal
plane) is calculated according to the following formula:
Distance moved by electrode system horizontally ∝ √((L*sineϕ)2 + (L*sineφ)2),
where L is the length of the electrode system from the manipulator to the tip of the
system, and ϕ and φ are the angles of the spherical coordinate system. The primary
method of translocation in some systems, this rotational system is often coupled with
MRI or CT scan to determine angles ϕ and φ. Fiducials mounted on the skull prior to
obtaining images allow one to make this determination. Detectable with MRI and CT
scan as well as ultrasound and reflectance, these fiducials constitute the navigation sys-
tem in the operating room. Detectable fiducials also adorn electrode system manipu-
lators (in one version of the frameless system). Wherever they appear these markers
are co-registered. Co-registration allows the operator to locate the target on the imag-
ing and relay the information to a second operation tasked with adjusting ϕ and φ and
directing the electrode systems to the target. Secondary translocations built into the
platform (usually situated in an x-y plane) enable operators to determine additional
trajectories (Figure 2.9). The x-y plane of the secondary translocations lies in the plane
orthogonal to the long axis of the electrode system.
One observes that the depth of the electrode system increases the change in posi-
tion. As the electrode system advances, therefore, any errors introduced at the moment
of setting the angles will increase errors in the horizontal plane. This places operators at
a significant disadvantage. No such magnification of error occurs with rotational sys-
tems whose focus, or cone apex, coincides with the target; operators may alter angles
and reach the target without fail.
To accomplish initial targeting, another system uses a prefabricated system
(Figures 2.4 and 2.10) that combines preoperative imaging with skull-mounted fidu-
cials. Using imaging, the manufacturer constructs a bespoke platform that the sur-
geon mounts to the patient’s skull. Because in this case the primary translocations
are fixed, guiding changes to trajectories are built-in secondary translocations, which
typically occupy an x-y plane orthogonal to the long axis of the electrode system.
Targeting
platform
Y
Y
PC
AC
Y
Fiducial
Mounting points
FIGURE 2.10 Schematic representation of a system that relies solely on a horizontal method of

translocation. Fiducials are applied to the skull, and MRI or CT scans obtained. Image-guided
systems direct fabrication of a platform that when attached to the fiducials previously fixed
to the skull, directs a cylinder of brain access toward the target. Translation in the horizontal
plane effects movement of electrode systems (see Figure 2.9).
ISSUE OF BRAIN SHIFT
A caveat: All foregoing discussion of image-based targeting concerned scans obtained

preoperatively. Issues attending the opening of the skull or dura have yet to be addressed.
The relative orientation of the brain to the external landmarks established in preopera-
tive imaging likely changes once the patient is arranged for surgery in operating room.
Surgeons and intraoperative neurophysicians must therefore expect some consequent
spatial error in targeting.
Brain shift owes to a number of causes, not the least of which is the patient’s physical
arrangement as she undergoes preoperative imaging. Whereas the patient lies supine
during an MRI or CT scan, in the operating room she occupies a semirecumbent
position, her head higher than her hips. Belief has it that the brain floats in cerebro-
spinal fluid and thus has neutral buoyancy. But this is not exactly the case. The brain’s
native weight of approximately 1400 g is not eliminated when it is suspended in cere-
brospinal fluid; approximately 25 g remain to it (Noback, Strominger, Demarest et al.
2005). Though small, this weight is enough to cause the brain to shift one way from one
position—while the body housing it lies supine for imaging studies, for instance—and
a second way from another: say, while the body housing it lies semirecumbent during
surgery. The extent of the shift depends on the relative volume of the cerebrospinal fluid
relative to the brain. In patients with brain atrophy, for example, one might observe a
much greater shift in distance.
Loss of all cerebrospinal fluid would result in a dramatic force increase—from 25 g
to 1400 g, to be exact. Admittedly, a complete loss is an unlikelihood, but merely imag-
ining it serves to emphasize the implications even a slight loss carries. Any air present
increases in volume as it warms to body temperature, and thus also increases downward
force owing to gravity. Of utmost importance, then, is minimizing cerebrospinal fluid
loss and the amount of air entering the subarachnoid space. For example, air can enter
the subarachnoid space via the burr hole and open dura and then float to the frontal lobes,
where it becomes trapped, the patient’s semirecumbent position making this possible.
Breathing, coughing, sneezing and cardiac systole produce changes in intracranial pres-
sure that may evacuate cerebrospinal fluid. Upon relaxation of this pressure, air can enter.
Initially at room temperature (20oC, for example), air that becomes trapped over
the frontal lobe warms to the body’s temperature of 37oC and thus experiences a 17-fold
increase in the product of pressure (P) and volume (V), as determined by the ideal gas
law, PV = nRT (R = universal gas constant; n = moles of gas; T = temperature). Air
enters the subarachnoid space at 1 atmosphere (atm) and remains at this pressure as
it warms, because the burr hole does not form a pressure seal. Tension pneumoceph-
aly thus results from the 17-fold increase in volume, one cc of air expanding to 17 cc
A B
e
FIGURE 2.11 Postoperative MRI scan performed approximately 24 hours following DBS targeting
of the ventral intermediate nucleus of the thalamus. One observes (A) the DBS lead’s extremely
anterior final position (d) vis-à-vis the ventral intermediate nucleus of the thalamus. One
observes also that the lead abuts the posterior limb of the internal capsule. Spread to the internal
capsule likely explains speech impairment with stimulation. In B appears a pneumocephalus
(e). Remnants of the microelectrode trajectories (c) appear in the medial globus pallidus interna.
after it becomes trapped over the frontal lobes. The brain shifts to accommodate the
expanded volume of air, increasing discrepancy between the targets and external land-
marks (Figure 2.11). Rather than using a sealant, such as fibrin glue, some surgeons use
a felt-like pad in order to save expense. In this author’s experience, use of such porous
felt-like pads to not protect against cerebrospinal loss, rather it may increase the loss by
acting as a wick.
Intracranial hematomas can cause brain shift. The exact direction of consequent
brain shift is difficult to determine. Loss of neuronal action potentials and a drop in
background activity in the micro—or semi-microelectrode recordings—the latter of
which occur despite preservation of electrode impedance—may indicate intracranial
hematoma, which one can confirm by inspecting electrodes and guide cannulas for
traces of blood. Though one may be tempted simply to place the DBS lead in the best
trajectory or rely on image-guided navigation, any brain shift caused by hematoma is
likely to result in suboptimal placement of the DBS lead.
ISSUE OF ANESTHESIA AND SEDATION
As will be discussed in subsequent chapters, providing important information for tar-

get localization is the ability to correlate changes in neuronal activities with behaviors
that come in response to a light touch or other peripheral nervous system stimulation.
Important also is the patient’s ability to report sensations associated with electrical
stimulation through the microelectrode, semi-microelectrode, and macrostimula-
tion, because this ability assists target location and helps to assure tolerable attempts
at postoperative therapeutic DBS. If the patient is sedated or anesthetized, gathering

these reports may prove difficult.
By affecting the firing characteristics of the neurons during microelectrode or
semi-microelectrode recordings, many anesthetic agents interfere with neurophysi-
ological mapping. Some surgeons use short-acting agents during frame placement
and creation of the burr hole—two highly uncomfortable portions of the surgical
procedure—because, given the short plasma half-life of the agents, they believe that
the central nervous system effects wear off prior to commencement of intraoperative
neurophysiological monitoring. If, for example, change in the plasma levels requires
approximately five half-lives to reach 99% of the state, an agent whose plasma half-life
is 15 minutes requires 75 minutes to be 99% eliminated from the plasma.
An important distinction obtains, however, between plasma half-life and clearance
from the brain. Lipophilic agents and many agents of other kinds accumulate in the
brain, a third-space that may prolong their clearance. Highly lipophilic propofol, for
example, rapidly redistributes in the brain and body. Though in terms of sedation its
reversibility is reportedly on the order of tens of minutes, propofol possesses a plasma
half-life of 3 to 12 hours. Indeed, the P300 EEG potential in patients sedated with pro-
pofol but otherwise conscious was abnormal (Sneyd, Samra, Davidson et al. 1994). In
some cases, the P300 did not return for 60 minutes. Reports on the effects of propo-
fol on subthalamic neurons contradict several studies demonstrating a reduction in
neuronal activity. As one study has shown, reduction in subthalamic neuronal activity
returned to baseline by 9.3 ± 4.0 minutes (Raz, Eimerl, Zaidel et al. 2010).
Midazolam
Though midazolam possesses a half-life of 1.8 hours (for older patients it is a bit longer),
electrophysiological measures have shown its effects to last hours beyond its discontinu-
ation (Bührer, Maitre, Crevoisier et al. 1990; Hotz, Ritz, Linder et al. 2000). A case from
this author’s experience involves four patients given intravenous midazolam during frame
placement, unbeknownst to the author. Of the four patients, three underwent ventral
intermediate nucleus DBS surgery, and in each of the three the author was unable to detect
neurons by use of microelectrode recordings whose impedances and other parameters
were nominal. The effects of midazolam on two of these three patients were reversed with
flumazenil, which could not be used on the third, his recent history of alprazolam use in
fairly large doses making an abrupt withdrawal response a concern. One must bear in mind
that flumazenil possesses a shorter plasma half-life than midazolam—0.7 to 1.3 hours, to
be exact—and thus may need to be administered in repeated doses.
Morphine
The effects of morphine and other opiates on microelectrode and semi-microelectrode

recordings do not appear to garner much discussion in the literature. Two patients from
this author’s experience received morphine during frame placement for subthalamic
nucleus DBS surgery (the first patient) and ventral intermediate nucleus DBS surgery
(the second). Marked reductions in neuronal activities characterized both cases, and in
both cases administration of naloxone reversed the morphine’s effects. One must bear
in mind that naloxone’s short half-life of one hour may make necessary repeated doses
and that withdrawal symptoms may occur rather abruptly.
A third patient from this author’s experience received both morphine and mid-
azolam. Microelectrode recordings whose nominal electrode parameters were normal
failed to detect any neuronal activity. Small traces of blood on the microelectrode tip
were observed when the microelectrode was withdrawn. The patient’s neurological
examination remained unchanged, but physicians aborted the surgery out of concern
about a possible intracerebral hematoma, though no evidence of such appeared on an
immediate MRI scan. The error leading to the halt of surgery, in other words, arrived
from the effects of morphine and midazolam on neuronal activity.
Dexmedetomidine
Though few studies have examined the specific effects of dexmedetomidine on neuronal
discharge patterns, numerous case reports exist that establish the demonstrability of neu-
ronal responses in the globus pallidus interna and subthalamic nucleus, which makes pos-
sible microelectrode and semi-microelectrode recording for the purpose of physiological
mapping.
POSTIMPLANTATION STUDIES
Studies focusing on use of intraoperative CT or MRI scans and other surgical tech-
niques evidence the fact that DBS continues to evolve. New implanted pulse gen-
erators (IPGs), leads, and extensions are being developed. For all of its evolution,
however, the success of DBS surgery may owe in part to idiosyncrasies of the sur-
geons, physicians, and neurophysiologists involved in it, as well as the institution
in which it is performed. Each institution must therefore implement and maintain a
vigorous quality-assurance program, one that demands cooperation from healthcare
providers involved in each stage of a patient’s care. Standard outcomes measures must
include symptom-based treatment, estimates of functional capacity, quality of life,
and adverse events, while assessments must include patients’ medical regimen—the
combinations of medications and the status of their administration (currently taking
or not taking certain medications)—and DBS therapy status (currently undergoing
it or not undergoing it). This author’s experience has shown him that postoperative
medication reductions following DBS for Parkinson’s disease offers insight into the
effectiveness of DBS as practiced by any specific team.
Determining the actual placement of DBS leads requires postoperative imaging,
whether by standard postoperative MRI or some other means. Because one can merge
it with a postoperative MRI or CT scan, a high-field-strength MRI obtained preopera-
tively offers the best view on DBS lead placement.
Best practice also prescribes that one obtain postoperative anterior-posterior and
lateral X-rays of a patient’s skull, cervical spine, and chest. These X-rays help one to
ascertain that the hardwire occupies its correct position and is properly connected and
they serve as useful material for an easy baseline study for comparison should any emer-
gency arise. Comparison between immediate postoperative X-rays and skull X-rays
obtained during an emergency allows one to check the intracranial tip of the DBS lead
for any evidence of migration or fracture. Figure 2.12 shows a fracture of the DBS lead
(appearing on the left and indicated by the arrow). One notes also that, as shown in
the lateral cervical spine X-ray, the connector between the DBS lead and the exten-
sion resides in the neck and not on the skull. The connector’s migration thereto may
have contributed to the lead’s fracture. Such a development requires that one repeat the
entire DBS surgery in order to rectify it.
FIGURE 2.12 Lateral cervical spine X-ray showing a fracture in the left DBS lead (indicated by
the arrow). One notes that this fracture may owe to the fact that the connector between the
DBS lead and the extension wire is not over the skull but in the neck.
SUMMARY
It behooves the intraoperative neurophysiologist to understand the initial targeting, as

this informers her what to expect during the intraoperative neurophysiological monitor-
ing. Also, determining whether a trajectory represents the optimal trajectory for the DBS
lead is relatively straightforward. More complicated is the decision where to place the sub-
sequent trajectories should they prove necessary. The responsibility for determining sub-
sequent trajectories typically befalls the intraoperative neurophysiologist. Translations to
achieve subsequent trajectories can be complicated, particularly by changing the regional
anatomies around the trajectories, by the choice of translocating mechanisms.
REFERENCES
Bührer M, Maitre P, Crevoisier C, et al.: Electroencephalographic effects of benzodiazepines. II.
Pharmacodynamic modeling of the electroencephalographic effects of midazolam and diazepam.
Clin Pharmacol Ther 48: 555–567, 1990.
Hotz M, Ritz R, Linder L, et al.: Auditory and electroencephalographic effects of midazolam and
alpha-hydroxy-midazolam in healthy subjects. Br J Clin Pharmacol 49: 72–79, 2000.
Noback C, Strominger N, Demarest R, et al.: The Human Nervous System. Totowa, NJ, Humana Press,
2005.
Raz A, Eimerl D, Zaidel A, et al.: Propofol decreases neuronal population spiking activity in the subtha-
lamic nucleus of Parkinsonian patients. Anesth Analg 111: 1285–1289, 2010.
Sneyd JR, Samra SK, Davidson B, et al.: Electrophysiologic effects of propofol sedation. Anesth Analg
79: 1151–1158, 1994.
/ / / 3 / / / BASIC CONCEPTS OF
ELECTRICITY AND
ELECTRONICS
INTRODUCTION
The brain is basically an electronic device, because its constituent neurons electronically
encode, process, and relay information. As such it provides the basis for intraoperative neu-
rophysiological monitoring, the method by which one accesses these electrical activities.
For its part, Deep Brain Stimulation (DBS) affects those electrical activities. Effective use
of intraoperative neurophysiological monitoring and DBS, then, requires an understand-
ing of electricity and electronics. Electricity refers to electrical charges and their properties,
and electronics to the manipulation thereof. Examples of the second include manipulation
of neurons in the brain to affect information and use of artificial systems for recording or
influencing neurons’ electrical charges.
ELECTRICITY
A term that relates to electrical charges, electricity admits of variety, a fact that individuals
interested in intraoperative neurophysiological monitoring must appreciate. One typically
imagines electricity as electrons flowing through copper wire or some other conductor.
These elementary subatomic particles possess a negative electrical charge of –1.6 × 10-19
coulombs (a coulomb is a unit of measure of electrical charge). The electrons occupying
outer shells of conductors of electrical current, such as electrical wires, are “loose” and
mobile. Thus, they may begin to flow. Under the influence of an electrical, magnetic, or
electromagnetic field, electrons may move from one conductor atom to another, travel-
ing from the negative (cathode) end of the electrical field to the positive (anode) end.
Movement of electrons is fundamental to electrodes and the recording and stimulating
devices used in intraoperative neurophysiological monitoring.
48
3. Basic Concepts of Electricity and Electronics / / 49
This chapter discusses how the various fields, electrostatic, magnetic, and electro-
magnetic, are involved in moving electrical charges from the neurons in the brain into
the electrodes for recording neuronal activity and how these fields are used to stimulate
the brain. These same fields in the environment can drive electrical charges, into and
out of the recording systems to produce artifact and noise, which can interfere with
intraoperative neurophysiological monitoring recordings or effect abnormal stimula-
tion of the brain. Indeed, understanding how electrostatic, magnetic, and electromag-
netic fields are generated and how they contribute to artifact and noise greatly aids
elimination of artifact and noise.
Electrons, and thus their use in electrical devices such as amplifiers, can be affected
by fields of three kinds: electrostatic, magnetic, and electromagnetic. The electrostatic
charge is due to the electrical charge itself and, thus, is present even if the electron is at
rest. Consider the experience of vigorously rubbing a balloon on a piece of fabric and
then placing the balloon against the wall. It appears to cling to the wall. In the process
of rubbing the balloon, the balloon picks up electrons from the fabric and becomes neg-
atively charged. Even though the excessive electrons just “sit there” they still exert an
attracting force on the wall. Thus, an electrostatic force can have an effect on an electron
such as one moving through a conductor.
Should the electron move, it will generate a magnetic field. An electron, even at rest
(if this were possible), creates a magnetic field because the charged electron is spinning.
Because the electron is acting like a miniature magnet, it will be affected by a magnetic
field. If the electron moves in an accelerating (or decelerating) manner through space it
creates an electromagnetic field. Similarly, an electromagnetic field can have an effect
on electrons causing them to move. Each kind of field has the ability to affect electrons
flowing in the electrodes and amplifier systems.
In brains flow ions, however, rather than electrons. Atoms characterized by an
unequal number of positive (protons) and negative (electrons) charges, ions achieve
this condition as a result of the deficiency or excess of negatively charged electrons
“orbiting” their nucleus, which number either more or less than their protons, respec-
tively. Atoms possessing fewer electrons than protons acquire a positive charge. Known
as cations, these atoms in their movement generate positive electrical current (current is
the flow of electrical charge over time). Sodium (Na+) and potassium (K+) are two com-
monly encountered cations. Atoms possessing more electrons than protons acquire a
negative charge. Known as anions, these atoms in their movement generate negative
current. Chloride (Cl-) is a commonly encountered anion. Neurons and other brain
cells generate electricity by creating positive and negative electrical currents involving
ions, which requires an electrical force similar to the field generated when one attaches
copper wire to a battery.
Recording neuronal electrical activity with electrodes requires conversion of elec-

trical current, as positive and negative ions in the brain, to the flow of electrons into
and out of the electrode tip. The flow of electrons toward a target or device is a nega-
tive current while flow away is construed as a positive current, though discussions of
the electronics involving electrons do not usually speak of positive currents. Similarly,
stimulation through a metal electrode tip requires conversion from an electrical change,
such as an excess or deficit of electrons at the electrode tip during the negative (cath-
odal phase) and positive (anodal) phase. Generally, there are two mechanisms in the
conversion from changes in ion concentrations in the brain to flow of electrons in the
electrodes: capacitive and faradaic. With stimulation, a capacitive exchange involves
generating an imbalance of electrical charges around the electrode tip creating an elec-
trostatic field (similar to the rubbed balloon described above). This electrostatic field
causes ions in the electrolyte solution (for example, water with ions such as Na+, K+,
and Cl-) to move, with the positive ions moving toward the electrode’s tip and negative
ions away from the tip with negative (cathodal) current. The opposite would occur with
the positive (anodal) current (Figure 3.1).
A buildup or an excess of electrons, or removal of electrons from the tip of the elec-
trode, also can create an electrostatic field that drives a chemical reaction, where cer-
tain atoms will accept an electron and become a negatively charged ion while others
will donate an electron and become a positive ion (Figure 3.2). The process of giving up
an electron is called oxidation, and gaining an electron is reduction. As these occur in
pairs, the general reaction is called Redox reaction.
During recording of electrical charges, electrons are induced to move in the metal
conductor of the electrode tip by a buildup of electrical charges in ions around the
FIGURE 3.1 Schematic representation of how a negatively charged stimulation electrode tip

during negative (cathodal) current results in a flow of ions that then generate action potentials
in nearby axons. In this case, ions already exist but migrate in different directions under the
influence of the electrostatic field due to buildup of negative electrical charges on the tip.
FIGURE 3.2 Schematic representation of how a negatively charged stimulation electrode tip

during negative (cathodal) current results in a flow of ions that then generate action poten-
tials in nearby axons. In this case, the ions do not exist prior to stimulation but are created by
electrostatic fields that are generated with stimulation. The reactions go in one direction to
create the ions with the first phase and then are reversed during the second phase. Incomplete
reversal could result a residual charge that has not been reversed and lead to tissue damage.
electrode tip by neurons in the vicinity. This buildup of either negative or positive ions
in numbers in excess of those when the neuron is not firing generates an electrostatic
field that moves electrons into or out of the metal electrode tip.
Whether capacitance or Redox reaction brings about an exchange of electrons
depends on the nature of the conductor, for example, whether the conductor is made of
tungsten, platinum-iridium, or silver, and the stimulation parameters. Because Redox
reactions can generate metal ions that can be deposited in the brain tissue, silver elec-
trodes are not used in microelectrode recordings (or stimulation), the toxicity of silver
ions presenting unacceptable risk. The nature and method of electrical charge transfer
carry significant implications for the fidelity of recordings of electrical signals. (These
implications receive further discussion below.)
ELECTRONICS
Electronics involves the purposeful and controlled movement of electrical charges con-
sisting of electrons or ions. For example, one may create a source of heat by sending
electrons through a wire in quantities sufficient to warm it due to friction. Increasing
this flow can warm it to glowing, thus creating a source of light as well. Rapid reversal
of electrons flowing through the wire of an electromagnet mechanically connected to
a membrane produces sound. Control of electrical current in the form of ions across
the cell membrane of a neuron generates an electrical pulse that carries or modifies
important information in the brain. The question then is how neurons control the flow
of electrical charges in order to do their work. The work of a neuron relates to controlled
changes in the voltage across the neuron’s membrane that is the charge outside the neu-
ron compared to inside the neuron.
Conceptually, the problem for the neuron is to generate a force that will move
electrical charges, like a battery, and then a means for controlling the flow of electri-
cal charges, like a switch. As will be seen, the force that will move electrical charges
relates to the differences in concentrations of anions and cations on the outside com-
pared with the inside of the neurons. For example, if there is a relative excess of cations,
such as Na+, outside the neuron compared with cations inside the neuron, there will be
a voltage across the neuronal membrane. This is analogous to a battery, which will be
described subsequently.
However, the force that moves ions across the neuronal membrane is not the voltage
difference or polarity across the neuronal membrane but rather the differences in the
relative concentrations of ions outside compared with inside the neuron (the concen-
tration gradient). There is relative greater concentration of Na+ ions outside the neuron
compared with inside the neuron. Thus, there is a force that will attempt to move Na+
ions to the inside of the neuron. The converse is true of other ions such as K+, where
there is a greater concentration inside the neuron compared with outside the neuron.
It is the drive from the differences in concentrations that causes the flow of electrical
charges between the inside and the outside of the neuron and, thus, determines the
voltage across the neuronal membrane and, consequently, the ability to record neuro-
nal discharge activity. However, the electrical charges, in the form of ions, also are sub-
ject to other forces such as electrostatic, magnetic, and electromagnetic forces. Thus,
the voltage across the neuronal membrane can be affected by electrostatic charges, for
example, and by the concentration gradient across the neuronal membrane. As will be
discussed latter, the control of the flow of electrical charges is determined by special
channels or pores in the neuronal membrane; these pores are controlled by the voltage
across the neuronal membrane and act as switches for the flow of electrical current in
the form of ions.
THE ELECTRONICS OF RECORDING AND STIMULATING SYSTEMS
Known as electromotive force, the force in electronics that impels electrical charges is
measured in volts. Voltage difference throughout the conductor determines electromo-
tive force. A battery, for example, generates electromotive force or volts (Figure 3.3).
The force provided by the separation of charges, in the case of a battery, is insuf-
ficient to move the electrical charges through the semipermeable membrane barrier.
+ –
+ +
– +
+ – –
– + –
+ +
+ – –
– +
+ + –
+ –
– –
FIGURE 3.3 Schematic representation of a battery’s construction. The two figures represent
containers holding ions in solution. The image on the left depicts a random distribution of pos-
itive ions (cations) and negative ions (anions). Randomly distributed, these cations and anions
in their movement produce no net electrical field. The image on the right depicts the conse-
quence of placing a barrier in such a way as to allow the solvent to flow across it. Depicted also
is a mechanism that concentrates the positive charges to the left of the barrier and negative
charges to the right. In so doing it produces a difference in the charges on each side of the
barrier. This separation of charges creates an electrical force that drives positive ions to the
right and a force that drives negative ions to the left. The presence of the barrier, however,
prevents ions from moving from one side of the barrier to the other. Source: Reproduced with
permission from Montgomery (2010).
One imagines a container that holds two electrodes, each occupying one half of it.
Either electrode interacts with ions on its side of the barrier (Figure 3.3). The electrode
on the right side of the barrier accepts electrons from negative ions, while the elec-
trode on the left donates electrons to positive ions. In each half, a force moves electrical
charges to the electrode from ions, or to ions from the electrode. Each movement of this
sort is known as a half-cell potential. The force of a half-cell potential depends on the
nature of the ions and the electrode.
The situation in which a conductor connects the two electrodes is illustrative of
a half-cell potential. In such a situation, the difference in half-cell potentials at each
electrode becomes the force that moves electrons through the connector (one assumes
here that a metallic conductor is in use) (Figure 3.4). The flow of the electrical charges,
or current, is measured in amperes. One ampere, or amp, equals 1 coulomb of charge in
motion for 1 second. Intraoperative neurophysiological monitoring currents are typi-
cally measured in milliamps (1 milliamp equals .0001 amp) and microamps (1 micro-
amp equals 1 × 10 –6 amp).
Certain forces resist the flow of electrical charges in conductors. For example,
resistance, a force measured in ohms (Ω), opposes the flow of electrons. The flow of
electrical current (I) through a conductor depends on force (V) and resistance (R).
(Unfortunately, the naming of the direction of current flow is problematic. The origi-
nal, and now conventional, definition is the flow of the positive current that is opposite
Switch Switch
open closed –
– –
+ – + –
–
+ –
+ + –
+ – +
+ – + –
+ – + –
+ – + –
– –
– – +
FIGURE 3.4 Creation of electrical potentials by the exchange of electrical charges among ions
occupying either side of the barrier. Metallic conduction connecting the two electrodes gener-
ates a force or potential capable of moving electrons from one electrode to the other. A switch
allows one to control of the flow of electrons between electrodes. Source: Reproduced with
in direction to the flow of electrons. If one were to imagine an electron moving in one
direction, for example to the right, one can think that this leaves a “hole” in the elec-
trons that results in a positive charge. Thus, as electrons move in one direction, to the
right for example, the positive “hole” moves in the opposite direction, left for example.
This book will discuss current flow in terms of the direction and magnitude of the flow
of electrons.) Ohm’s law, where V = IR, defines this relationship.
The flow of water provides an analogy to the flow of electrical current as it is con-
trolled by the electronics. The situation depicted below involves a reservoir that holds
water and that has a hose connected to it (Figure 3.5). The height of the reservoir
determines the hydrostatic force, the generation of which pushes water through the
hose. The height of the reservoir is analogous to voltage in electrical circuits. The hose
resists the flow of water in the same way that a metallic conductor resists the flow of
electrons. The diameter of the hose determines the resistance: the smaller the diame-
ter, the greater the resistance. Thus, the amount of water expelled by the hose depends
on the height of the water reservoir and the diameter of the hose. One may increase
the flow by either raising the reservoir (or voltage in the case of an electrical circuit) or
increasing the diameter of the hose (reducing the resistance in the electrical circuit).
Conversely, one may lower the flow by reducing the height of the reservoir (voltage)
or reducing the diameter of the hose (increasing the resistance). Should resistance
increase, one may maintain the same rate of flow by increasing the hose’s diameter or
the reservoir’s height.
C
A B
FIGURE 3.5 Illustration of Ohm’s law by way of water-propulsion analogy. A reservoir supplies
hydrostatic force to drive water in it through a hose. The degree to which the flow of water
meets resistance depends on the diameter of the hose: the smaller the diameter, the greater
the resistance. Provided the reservoir’s height remains constant, use of a small-diameter hose
decreases the amount of water exiting, and use of larger-diameter hose increases it (A and
B). Should one return to using a small-diameter hose, she may maintain the amount of water
exiting the hose by increasing the reservoir’s height (B and C). Source: Reproduced with per-
mission from Montgomery (2010).
ALTERNATING OR FLUCTUATING CURRENT
Discussion of resistance to this point has focused on direct current (DC) circuits, that
is, circuits whose current is constant. In the situation of DC circuits, the resistance is a
function of how tightly the atom holds onto its electron, the tighter the hold, the greater
the resistance. However, most of the electrical currents one encounters in intraopera-
tive neurophysiological monitoring—extracellular action potential and local field
potentials, for example—vary. Alternating or fluctuating currents and voltages oppose
the flow of electrical charges in an altogether different way than do direct currents,
and their opposition is complicated by two factors, inductive reactance and capacitive
reactance, in addition to resistance. The value of the opposition to the flow of current in
situations of fluctuating currents is the impedance (Z) and depends on the speed (fre-
quency) of changes in current and is measured in Ohms (Ω). Ohm’s law holds for situ-
ations of fluctuating currents, but the equation uses impedance rather than resistance.
Expressed as an equation, Ohm’s law thus becomes V = IZ.
One component of impedance, inductive reactance gives rise to induction.
Induction is a consequence of the fact that flowing electrical current generates a mag-
netic field and that magnetic fields in turn generate electrical current but in the oppo-
site direction, thus contributing to the flow of current. The movement of electrical
charges in a conductor creates a magnetic field that affects the same conductor, which
in turn generates a flow of electrical charges. As long as the magnetic field does not
fluctuate it will not drive electrical charges to move. For that reason DC circuits do not
create indication. One can imagine a magnetic field as containing lines of magnetic
force (Figure 3.6) that radiate from the north and south poles of the magnet. So long as
N S
FIGURE 3.6 Schematic representation of the iron-filing experiment demonstrating lines of

magnetic force.
the conductor does not move with respect to the lines of magnetic force, no electrical
forces will be generated in the conductor. If the conductor does cross lines of magnetic
force, either by the magnet or conductor moving, electrical current will be generated
in the conductor and oppose the normal or intended current flow. Electrical plants
powered by coal or nuclear energy spin conductors housed around large magnets in
order to generate electricity. The conductor must cross lines of magnetic force. This
can be accomplished by moving the conductor, as is often done with electrical genera-
tors, or by moving the lines of magnetic force.
Even a motionless conductor with alternating or fluctuating electrical current flow-
ing within it will create fluctuating or moving lines of magnetic force (Figure 3.7).
Generated by electrical current moving through and interacting with the conductor
itself, moving lines of magnetic force create a potential for electrons’ movement in
the conductor. Yet the potential generated by moving lines of magnetic force actually
oppose the original current. This is termed inductive reactance. Thus, the net current
flow is going to be less, thereby implying an increased opposition (impedance) to the
intended flow of electrons. Acceleration of the alternating current’s fluctuations (fre-
quency) meets with acceleration of change in the lines of magnetic force. Thus, the
faster the intended electrical current is fluctuating, the faster the lines of magnetic
force will change. These faster fluctuating changes in magnetic force will cause greater
induction of current in the opposite direction, giving rise to greater resistance to the
intended flow of electrical charges. Increase in frequency, hence rate of change, leads to
increased impedance.
Capacitive reactance, which is related to the actions of a capacitor that stores elec-
trical charge, also contributes to impedance. A typical electronic capacitor consists
of two conductors separated by nonconductive material (Figure 3.8). The electrical
– + Voltage or
electromotive
force from neuron
Conducting wire
Negative charge
Expanding then
contracting lines
of magnetic force
Voltage or + –
electromotive
force from induction
FIGURE 3.7 Schematic representation of the relationship between inductive reactance and
impedance. Electrical current flowing through the conductor (top figure) generates a magnetic
field whose force moves in circular fashion along the conductor in the direction indicated by
the thumb of the hand gripping the conductor (the curled fingers represent the direction of the
magnetic force’s circular movement). As depicted in the middle figure, the lines of magnetic
force expand and collapse, inducing an electrical current in the conductor as they do so. The
curled fingers of the left-hand rule represent the direction of lines of magnetic force, which is
the same as the direction indicated by the fingers of left-hand rule appearing in the top figure.
(Unfortunately, the naming of the direction of current flow is problematic. The original, and now
conventional, definition is the flow of the positive current that is opposite in direction to the flow
of electrons. If one were to apply the conventional definition then the right hand and not the left
hand would be used to show the direction of the magnetic field.) The thumb of the hand appear-
ing in the lower figure indicates the direction of a secondary electrical current generated by the
changing lines of magnetic force. This secondary current moves counter to the original electri-
cal current. In so doing, it resists the original flow of current and thus contributes to impedance.
charge moves through the conductor leading to the first plate of the capacitor and
accumulates on that first plate (Figures 3.8A and B). This accumulated charge gener-
ates an electrostatic field that creates an electrical current in the conductor leading
away from the capacitor by “pushing” charge of the same polarity from its place on the
second plate. Electrical charge accumulating on the first plate of the conductor begins
to resist any further flow of electrical current from the source toward the first plate
of the conductor (Figures 3.8C and D). There occurs as a consequence a reduction in
net force exerted against electrical charges moving in the first conductor leading to
the capacitor and in the current leading away from the capacitor. With this reduction
comes, on the other side of the capacitor, an arrest of electrical current, which at this
point is said to be saturated (Figure 3.8E).
Voltage or
– + electromotive
Force from neuron
A Conducting wire
B
+ – Voltage or
electromotive
force from
C
build up of
charges
D
Negative charge
FIGURE 3.8 Schematic representation of the development of capacitance. Voltage applied in a

manner depicted by the hand in (A) causes electrical charges to flow toward the capacitor (B).
Electrical charges begin to accumulate on the first plate of the capacitor (C), producing a voltage
in the opposite direction. This accumulation of charge continues until the voltage on the first plate
of the capacitor comes to equal the voltage of the source (D). At that point the capacitor reaches a
point of saturation, making further flow of electrical charges impossible. When the source voltage
drops to zero, or ground (represented by the gray hand), the electrical charges leave the first plate
(discharging the capacitor), moving in the opposite direction from that in which it moved during
charging (E). It continues to move in that direction until the capacitor is fully discharged (F).
In the case of a fluctuating electrical voltage or current, the moment the voltage
pushing electrons on the first side of the capacitor reaches ground or reverses, electrical
charges accumulated on the first plate flow to the conductor that led originally to the
first plate, producing an electrical current in the opposite direction. This effect mani-
fests on the second plate of the capacitor, sending flow in the opposite direction, that is,
to the other side of the capacitor.
The effects of accumulating charges on the capacitor resisting further flow of elec-
trons to and through the capacitor is called capacitive reactance and occurs anytime
an energy flow meets a change in the medium that affects its flow. The resistance to
flow of an alternating voltage or current source is impedance. As discussed previously,
impedance is related to three factors: (1) the inherent resistivity of the conductor to the
flow of electrical charges, which is analogous to resistance in DC circuits; (2) induc-
tive reactance; and (3) capacitive reactance. These factors will have significant implica-
tions for recording and stimulating neurons, which will be discussed in greater detail
in chapter 5.
In situations in which a fluctuating or alternating current forms a square wave or takes

some other specific shape (Figure 3.9, “source voltage”), the voltage and current accumu-
lating on the second plate represent the signal following its passage through the capacitor.
One observes that the shape of the source is different from that of the signal exiting the
capacitor. An example is when stimulating using a square wave pulse. A voltage increase on
the first plate to the level of voltage at the source reduces current flow through the capacitor
(Figures 3.9B and B′ ). Flow of current ceases once voltage on the capacitor comes to equal
the source voltage (Figures 3.9C and C′ ). When the source voltage drops below the volt-
age on the capacitor, current begins to flow in the opposite direction (Figure 3.9D and D′ )
until such time that the charge on the capacitor comes to equal the source voltage, which
itself has reached zero, or ground (Figure 3.9F and F′ ).
Capacitive reactance depends on frequencies in the source voltage. In the case of
inductive reactance, impedance associated with inductive reactance increases as fre-
quency of the alternating source voltage increases. Capacitive reactance decreases
C’ D’
Voltage across
capacitor
B’ E’
A A’ F’
0
B
A’
C
Current across
B’
capacitor
D C’ E’ F’
0 D’
E
Voltage
Source
0
FIGURE 3.9 Schematic representation of capacitive reactants’ effect on the changes in the
source waveform. The source consists of a square-wave pulse that rises to some voltage
before returning to ground (0 volts). Caused by the source voltage, a second voltage mani-
fests across the capacitor, sending current flowing across the capacitor. Increasing source
voltage is answered by increasing electrical charges (A), which accumulate on the first plate
of the capacitor (B), and which engender voltage whose influence sends current flowing in
the opposite direction. When the accumulated voltage comes to match the source voltage (C),
movement of electrical charges ceases and current across the capacitor moves to 0 (C′). Upon
reduction of the source voltage—the source voltage’s return to ground (0 volts), in this case—
the capacitor begins to discharge (D), producing current in the opposite direction (E and E′)
until such time as the capacitor is fully discharged (F and F′).
as frequency increases. Slower frequency source voltages permit greater saturation

of the capacitor, a process whose completion renders impossible further current flow
(Figure 3.9). This results in signal loss. Rapid frequencies produce no saturation
or do so only for brief periods. Yet these periods, though brief, result in signal loss
(Figure 3.10).
Not the repetition rate but the frequency of the waveform of the signal affects
inductive reactance, capacitive reactance, and impedance. Frequencies within the
waveforms are given by the Fourier transform of the signal (Figure 3.11). A train of
square waves pulsing at 100 pulses per second (pps), for example, does not have a fre-
quency of 100 Hz. Rather, extremely high frequencies characterize the wave’s leading
and trailing edges, as evidenced by its sharp rise and fall. Impedance thus has a greater
effect on a train of square wave pulses than it does on a sinusoidal wave pulse train of
equal frequency. On frequencies of the source voltage, inductive reactance and capaci-
tive reactance have differing effects. Determining that reactance which dominates in
any particular circumstance proves problematic. Use of the coiled-conductor DBS lead
and extension in one commercially available system, for example, results in extremely
different inductive reactance than does the uncoiled conductors in another system.
Capacitive reactance and inductive reactance, along with resistance, combine
to create impedance, which proves a factor regardless of whether the neuron being
Voltage
Low frequency
Current
Signal
Recorded
Signal
Voltage
High frequency
Current
Signal
Recorded
Signal
FIGURE 3.10 Schematic representation of the effects of frequency on capacitive reactance.
One observes that low frequency allows the capacitor to saturate, as evidenced by the loss
of current near the end of the signal. This loss of current results in truncation of the signal, as
evidenced by the shorter duration of the recorded signal. Higher frequencies prevent satura-
tion and thus result in no loss of signal.
Neuronal action potential

Waveform
Component frequency A
Component frequency B
FIGURE 3.11 Schematic representation of the effects of frequency-dependent impedance

on the shape of a waveform. The original waveform has a shape “built” of two sine waves
of different frequencies. The original waveform could be considered as comprising two
waveforms. Impedance due to inductive reactance will have greater effect on the higher
frequency component of the original waveform and, consequently, will alter the shape of the
waveform. For example, the original wave neuronal action potential waveform will appear
the same as component frequency A. In the case of square pulses, as used for stimulation,
the square wave pulse contains very high frequencies and consequently will be greatly
altered by the impedance unless compensatory mechanisms are used, such as constant
current stimulation.
recorded or a stimulator supplies the source voltage. In cases in which a neuron

or ensemble of neurons is the source, high impedance alters and degrades extra-
cellular action potentials or local field potentials. In cases in which stimulation is
the source, high impedance degrades the pulse. The effect of the stimulation pulse
has three characteristic features: (1) it is based on the current delivered during the
pulse; (2) it is measured as the area under the electrical current-versus-time curve
(Figure 3.9); (3) it is the amount of charge delivered. High impedance thus reduces
the effectiveness of the stimulation pulse. A constant-current stimulator adjusts the
source voltage to ensure a constant current being delivered.
SUMMARY
Because the brain is basically an electrical device, its electronics offer a way to control
the electricity of its systems, that is, the ensembles of neurons that form it and that func-
tion variously as resistors, capacitors, and transistors. Gaining this control requires that
one observe some important parameters—the force that moves electrical charges, the
resistance in DC circuits and impedance in AC circuits that oppose the flow of electri-
cal charges, and capacitance that stores and then releases electrical charge, among oth-
ers. Later chapters address the effects of these parameters on the generation of neuronal
activities, recording and stimulation equipment, and the electrical environment.
REFERENCE
Montgomery EB, Jr.: Deep Brain Stimulation Programming: Principles and Practice. New York, Oxford
/ / / 4 / / / ELECTRODE RECORDINGS
Neurophysiology
INTRODUCTION
This chapter reviews the sources of electrical signals that are the basis for intraopera-
tive neurophysiological monitoring. Though many of the basic principles discussed in
chapter 3 apply to neurons, one should not gather the impression that neurons stand
unique among cells of the brain or body for their ability to generate electrical charges.
Neurons do distinguish themselves, however, in their amenability to dynamic manipu-
lation of their electrical charges.
A separation of charges across the cell membrane prompts changes in electrical
voltages around a neuron, which in turn generate the electrical signals recorded dur-
ing intraoperative neurophysiological monitoring. Upon separation and with changes
in the membrane structure, the charges (ions, specifically) begin to flow through the
cell membrane, thus becoming an electrical current. Able to generate current across
their entire membrane, neurons produce voltage whose changes constitute the electri-
cal signal recorded during intraoperative neurophysiological monitoring. Each region
of a neuron, however, possesses unique properties that determine the type and course
of recordable electrical signals.
THE NEURON
A schematic representation of a neuron appears in Figure 4.1. The neuron consists

of a cell body, or soma, which houses the nucleus and other machinery necessary for
the neuron’s health. The projection out from the cell body, the axon, mediates the
information the neuron emits. Sometimes extremely long, an axon may run from
the top of the cerebral cortex to the bottom of the spinal cord, ending at a terminal
63
Dendrite
Synaptic
terminal Cell
body
Axon
Synaptic
terminal
FIGURE 4.1 Schematic representation of a neuron, which consists of a cell body housing a

nucleus and cellular machinery. Depicted are two types of branches: (1) an extremely long
single axon, which conducts information from one neuron to others; and (2) dendrites, which
receive information from other neurons. The axons from other neurons make contact with the
dendrites at the axon’s synaptic terminal.
in contact with the subsequent neurons downstream. Its design is such that it trans-
mits neuronal information with high fidelity down its entire length. Terminations
of the axons onto subsequent neurons are known as synapses because of an inter-
vening gap that uses subsequent chemical transmission to relay information. These
contacts convey information from one neuron to another. (Notable exceptions to
this are gap junctions, whose connections directly convey electrical energy from
one neuron to the next.) In addition to making contact with the cell body, axon ter-
minals typically also connect with other branches known as dendrites, which along
with the cell body receive information from other neurons. (Instances also exist in
which axons make synaptic contacts with other axons to provide presynaptic inhibi-
tion and serve other functions.)
SIGNALS USED IN INTRAOPERATIVE NEUROPHYSIOLOGICAL MONITORING
The signal recorded in intraoperative neurophysiological monitoring arrives from

a change in voltage caused by electrical current. (Chapter 3 contains discussion of
the nature of batteries.) Like a battery, a neuron requires a separation of charges. This
separation of charges owes to a greater concentration of sodium (Na+) ions outside
the neuron and a greater concentration of potassium (K+) ions inside it (Figure 4.2).
The concentration of Na+ outside exceeding the concentration of K+ inside, a greater
4. Electrode Recordings / / 65
+ + Na+ Na+
+ + Na+ +
+ + Na+ Na
+ Na+
Battery –
K+
K+
–
Segment of Segment of
neuronal neuronal
axon axon
Outside Outside
Inside Inside
FIGURE 4.2 Schematic representation of a neuron as battery. Generated by a separation of

electrical charges (see chapter 3), the battery-like voltage in the neuron produces a greater
concentration of sodium (Na+) ions outside the neuron than within it, as well as a greater
concentration of potassium (K+) within it than outside it. The difference in Na+ concentration,
however, exceeds that of the K+ concentration. A relatively greater positive charge thus accu-
mulates outside the neuron, and a relatively greater negative charge accumulates within it.
Source: Reproduced with permission from Montgomery (2010).
net positive charge builds on the neuron’s surface. Mechanisms in the neuronal mem-
brane expels three Na+ for every two K+ it admits establishes and maintains the differ-
ences in ion concentrations.
A conductor permits electrical current to flow between a battery’s cathode (negative
contact) and anode (positive contact). (Unfortunately, conventional notation has current
flowing from the positive [anode] to the negative [cathode]. The direction of electrical
current described here is based on the direction of the flow of electrons in metal con-
ductors and the flow of negative ions in the brain.) A switch interposed in the conductor
(Figures 4.3 and 4.4) allows one to control the flow of electricity for the purpose of doing
useful work. Whereas a battery’s electric current consists of electrons, a neuron’s electric
current consists of ions—atoms whose charge is determined by an excess or deficiency
of their electrons relative to their protons. The electric current of ions flows through the
cell membrane. The cell membrane’s lipid bilayer, however, resists the flow of electrical
charges (Figure 4.4). Special transversal protein channels whose pores or openings allow
ions to flow through the membrane must therefore open in the cell membrane for current
to flow (Figure 4.5). In the resting state, the channels or pores close to prevent current
flow. From the halted flow follows no change in voltage or emission of signal. (One notes
that absence of signal does not entail absence of information; like spaces between letters
and words, absence of signal contributes to the production of information.)
+ +
+
+
–
–
–
–
– –
–
Segment of
neuronal
axon
Outside
Inside
FIGURE 4.3 Schematic representation of the means of controlling, for the purpose of perform-
ing meaningful work, flow of electric current from the negative to the positive pole of a battery
by interposition of a switch. In order for a neuron to perform its work of generating, its mem-
brane must similarly feature some sort of switch to control its flow of ions. Source: Reproduced
with permission from Montgomery (2010).
Switch Switch
open closed –
no flow – flow – –
– – –
Na+ Na+
Na+ Na+ Na+
Na+ Na+
Na+ Na+
Na+
+ – + –
FIGURE 4.4 Schematic representation of a neuron’s “switch,” which, appearing in the form

of a valve penetrating the membrane, controls electric current. The figure on the left depicts
a closed valve. The figure on the right depicts a valve held open by some obstruction. In a
manner analogous to closing an electrical switch, the neuron generating an electrical signal
permits ions (sodium [Na+] in the present case) to flow through it. Source: Reproduced with
Ion
Ion
Outside
Lipid bilayer Inside

Receptor or
channel
FIGURE 4.5 Schematic representation of a neuronal “switch” and “battery.” The lipid bilayer
resists the inward or outward flow of ions. Spanning the lipid bilayer are protein structures
whose channels open or close to permit outward or inward flow of ions.
The constituent phenomena of neuronal current manipulation occur in the

following sequence: (1) opening and closing of protein channels, (2) changes in
voltage, and (3) emission of an electrical signal containing information. Essential
for creating, processing, and conveying information, the precise control of chan-
nels can be achieved in a number of ways. One way involves release of a chemical
neurotransmitter by an action potential from the axon’s synaptic terminal. This
neurotransmitter diffuses across the synapse and binds with a receptor on the post-
synaptic membrane’s protein channel (Figure 4.6) to cause a change in the latter’s
structure. With this change, the channel’s pore opens to permit the flow of electric
current. Channels that open in this way are known as ligand-gated ion channels (the
neurotransmitter is the ligand). A second way involves a neurotransmitter that binds
to another receptor. The establishment of the bond triggers a cascade of protein
enzymatic steps that culminates in change to a protein channel. Thus altered, the
channels pore opens, allowing ions to flow (Figure 4.7). G protein-coupled channels
exemplify this second way.
GRADED VERSUS ACTION POTENTIALS
There generally are two types of changes in the postsynaptic neuronal membrane
potentials. One is localized to the vicinity of the synapse that has just released its
Modify to show synapse
Neuro- Ion
transmitter Ion
Outside
Receptor
binding
Inside
Lipid bilayer
Receptor
or channel
FIGURE 4.6 Schematic representation of ligand-gated ion channels. In an instance of ligand

gating, a specific neurotransmitter released from another neuron interacts with a binding
receptor on the protein channel. Each neurotransmitter interacts with a single binding recep-
tor (or a few receptors) in a manner analogous to the interaction between a key and lock (the
neurotransmitter is the key, and the receptor the lock). Two different types of neurotransmit-
ters and binding receptors are illustrated, each associated with a different effect of ion flow.
Neuro-
transmitter Ion
B
Receptor
binding
Lipid bilayer
Receptor or
channel Protein
FIGURE 4.7 Schematic representation of a G protein-coupled channel. In an instance of G pro-
tein coupling the neurotransmitter binds to a separate protein (or a different part of the chan-
nel protein). Instead of directly causing the ion channel to open, the receptor protein initiates
a cascade of enzymatic reactions to produce a chemical inside the neuron that causes the
channel’s opening.
neurotransmitter. There is a brief flow of current that allows a change in the ion con-
centrations such that the outside of the neuron becomes less positive, resulting in
a decrease in the voltage or electrical potential across the neuronal membrane (the
process of depolarization), or more positive, resulting in an increase in the voltage or
electrical potential (the process of hyperpolarization). Depolarization tends to make
the neuron more excitable, that is, it increases the likelihood the neuron will gener-
ate a signal that will be transmitted to the next neuron (the signal is referred to as
an action potential). Hyperpolarization decreases the excitability of the neuron and,
hence, decreases the likelihood of generating a signal. The change in the membrane
electrical potential is greatest just under the synapse. As the name graded potential
implies, the change in membrane potential spreads out over the membrane, decreas-
ing as it spreads (Figure 4.8).
By contrast, action potentials do not diminish in amplitude over distances and time.
While graded potentials typically are generated in the cell body (soma) and dendrites,
action potentials are generated in the axons. Thus, action potentials traveling long dis-
tances in the axon must be able to maintain the fidelity of the information being trans-
mitted. The fidelity is maintained by an active regenerative process along the axons.
This regenerative process involves channels in the neuronal membrane, but these chan-
nels open or close when the electrical potentials in the neuronal membranes reach
Voltage
A B C D
FIGURE 4.8 Schematic representation of a graded potential caused by passive diffusion of

ions to contiguous regions of the neuronal membrane. Segment A at time 1 (Figure A), for
example, indicates an initial influx of anions into the neuron. In this example, the anions enter
through a ligand-gated ionic channel, as might occur in the postsynaptic membrane follow-
ing neurotransmitter release from the presynaptic terminal. The ions that enter through the
postsynaptic membrane passively diffuse through adjacent regions (B, C, and D), causing a
graded depolarization of decreasing amplitude as the distance from the postsynaptic mem-
brane increases.
certain threshold values. These channels are called voltage sensitive or voltage-gated
channels. The regenerative process involves active positive and negative feedback to
these voltage sensitive channels.
Graded potentials typically occur in the neuronal membranes of most dendrites
and cell bodies because they lack voltage-gated channels and therefore also lack regen-
erative capability. Voltage change initiated in one part of the cell membrane does not
initiate—whether by opening up voltage-gated channels or by some other means—
active voltage change in the rest of the cell membrane, the dendrites, or the cell body.
Rather, the flow of ions causing the voltage change dissipates either inside or outside
the neuron by passive diffusion (Figure 4.8). It is similar to the spreading wave on the
surface of a pool of water after a stone has been dropped into the water. Similarly, the
magnitude of the changes in the electrical membrane potential decreases over time.
Graded postsynaptic potentials provide computational capacity. The information
contained in one postsynaptic potential may interact with information contained in
other postsynaptic potentials. Consider the analogy of simultaneously dropping two
stones into the pool of water fairly close together. Where the waves from each stone
overlap, the amplitude of the wave will increase. A graded potential of 0.4 combining
with another graded potential of 0.3, however, produces a potential of 0.7; and a graded
potential of 0.3 combining with another graded potential of -0.5, produces a potential
of -0.2. This can be understood as a convergence of ion concentration changes as the
ions diffuse outward from the postsynaptic cell membranes in a process known as spa-
tial summation (Figure 4.9).
Postsynaptic potentials may interact over time, as well. Those potentials under
synaptic contact, for example, diminish over time. Should a subsequent postsynap-
tic potential initiate on the heels of lingering remnants of a prior potential, however,
the two potentials may sum. Also, should another nearby postsynaptic potential ini-
tiate before the first postsynaptic potential subsides, the two potentials may interact.
Interaction in either instance is known as temporal summation. Consider the analogy
of dropping two stones into a pool of water, one after the other but in the same location.
A wave from the second stone can “catch up” with a wave from the first stone, and con-
sequently the combined amplitude will be greater.
The water analogy for both spatial and temporal summation illustrates the case if the
two synaptic events are of the same polarity, that is, both are depolarizing or both are
hyperpolarizing. In these cases, the amplitudes will be greater where the waves overlap.
However, if the polarities are different, one depolarizing and the other hyperpolarizing,
the effect will be to cancel each other. This provides even greater computational power.
Voltage
Voltage
A1 B1 C1 D1 E1 F1 G1
Voltage
Voltage
A2 B2 C2 D2 E2 F2 G2
Voltage
Voltage
A3 B3 C3 D3 E3 F3 G3
FIGURE 4.9 Schematic representation of spatial summation. One imagines a situation in which
there occur two simultaneous postsynaptic potential changes at time 1 at segment A (Figure
A1) and segment G (Figure G1). Increased anions associated with the opening of channels at
A1 begin to diffuse toward segments B2, C3, and D3. Increased anions entering the neuron at
G1 begin to diffuse toward F2, E3, and D3. At D3 ions converge, causing ion concentrations to
increase at D3 and culminating in a spatial summation.
The case is different, as will be seen, for action potentials. For example, one action
potential with a value of 1 (a value that indicates presence), when combined with
another action potential of a value of 1, produces a value 1. The neuronal membranes
serving action potentials exist in one of two states: (1) rest potential or (2) action poten-
tial. These states are analogous to digital computer binary code, which consists of ones
and zeroes. In the case of action potentials, the ground, or zero, voltage corresponds to
the zero of binary code, and positive voltage corresponds to the one.
Spatial and temporal summation process the various information streams that flow
from numerous other neurons and that enter a neuron through inputs on its dendrites
and cell bodies. The resultant new information, which is analog in nature, is relayed to
the next set of neurons through the axon, whose output is digital in the sense that its sig-
nal consists of either the presence or absence of an action potential. Converting analog
inputs to digital outputs are thresholds in the various voltage-gated channels located in
the axon and at the junction of the axon and the cell body. (The junction is also known
as an axon hillock or action potential initiating segment.)
Action potentials maintain their voltage and, hence, information as the action
potential travels along the axons to the next neuron. A regenerative process is neces-
sary to maintain the action potential amplitude, and this process involves positive
and negative feedback that are offset in time. In a positive feedback loop, once the
process is started the results of that process further increases the process. In the case
of action potentials, once a change in initiated in the membrane potential, typically
a depolarization, voltage sensitive channels begin to open up, allowing a flow of ions
that causes further changes in the neuronal membrane channels, for example, further
depolarization. These subsequent channels cause more channels to open, causing fur-
ther depolarization and even more channels to open. This is much in the manner of a
chain reaction. Negative feedback occurs when the result of the initial process shuts
the process down or when the results of the initial process produce a competing result.
This will become clearer when the flow of ions that generate the action potential are
subsequently described.
Fundamental to the generation of action potentials are certain protein channels,
known as voltage-gated channels, that are sensitive to the voltage across the cell mem-
brane and may remain closed until the membrane voltage reaches a certain threshold
value, at which point they open to allow electrical current to flow. Further increase
in membrane voltage opens additional channels, which triggers a membrane poten-
tial’s spread to adjacent membranes in a positive feedback. An ongoing process, this
neuronal action potential regenerates as change in membrane voltage in one area initi-
ates changes in membrane voltage in other areas (Figure 4.10). Because it possesses
voltage-gated channels, the axon cell membrane undergoes regeneration of this sort,
which enables it to conduct the action potential down its length without signal loss.
Understanding this conversion process requires that one appreciate certain details
of the regenerative process characteristic of an axon’s action potential. Typically, the
graded potentials resulting from synaptic inputs sum to cross a threshold sufficient to
open voltage sensitive ionic conductance channels with the subsequent generation of
an action potential. The action potential generation typically involves ions Na+ and K+,
though calcium (Ca++) along with chloride (Cl-) or other negatively charged compan-
ion ions may participate as well. As described above, more Na+ concentrates outside
the neuron than K+ concentrates within it. The neuron thus becomes some 70 milli-
volts (mv) more negatively charged internally relative to its charge externally. A resting
membrane voltage of -70 mv inside prompts the majority of voltage-gated Na+ and K+
channels to close (Figures 4.11).
Na+ or other positive ions tend to enter the neuron as a result of the influence of the cell
membrane’s concentration gradient; this decreases the negative charge of the membrane
Voltage
Voltage A1 B1 C1 D1
A2 B2 C2 D2
Voltage
A3 B3 C3 D3
Voltage
A4 B4 C4 D4
FIGURE 4.10 Schematic representation of the regenerative process for creating an action poten-
tial and conducting it along the neuronal membrane. A, B, C, and D depict four contiguous sec-
tions of a neuronal membrane for times 1 through 4. Figure A1 shows a local region in which the
anionic channel opens to allow Na+ or other anions to enter the neuron. This influx of anions pro-
duces the change in membrane voltage shown in the graph. The increase in ions subsequently
diffuses to segment B1 and begins to depolarize that region. When the B segment is depolarized
above threshold, the voltage-gated anion channel opens, allowing additional anions to enter the
neuron occupying this region (shown in B2). The anions then diffuse toward segment C2. Once
sufficient numbers of anions have diffused to segment C at time 3 (C3) to reach threshold, the
voltage-gated anion channel opens to allow more anions to enter the neuron at C. These ions dif-
fuse to segment D at D3. When sufficient depolarization occurs at D4, the anion channels in seg-
ment D open. This process continues along the entire length of the axon or other neural element.
at rest and drives it toward the positive membrane potential in a process known as depo-
larization. The depolarization insufficient for crossing the threshold, the voltage-gated
channels at this point undergo no change. Continued polarization, however, causes the
Na+
Channels
Na+ closed, Na+

activated K+ Na+ Na+
Na+ open,
Activated
Intracellular voltage
B
Na+ open, K+ K+
0
Inactivated
A
C
K+ closed, D
activated
–70
K+ open,
activated
Time
FIGURE 4.11 Schematic representation of action potential generation. At the point at which
the neuron achieves its resting membrane potential, or voltage, the ion conductance channels
through which ions enter or exit the neuron stand closed. Sodium ions (Na+) therefore cannot
enter the neuron, and potassium ions (K+) cannot exit it. A subsequent increase in the resting
voltage or potential (A) reaches the threshold at which the Na+ conductance channels open
to allow Na+ to enter the neuron. The influx of Na+ further reduces membrane voltage; this
is known as depolarization. With depolarization additional Na+ conductance channels open
in a chain reaction that generates the action potential. Further depolarization arrives at the
threshold at which K+ conductance channels open to allow K+ ions to exit the neuron (B). This
efflux of K+ ions drives the membrane voltage toward negative. Polarization also serves to
deactivate Na+ conductance channels (C), which prevents any further influx of Na+. From con-
tinued efflux of K+ follows achievement of negative voltage greater than that of the membrane
at rest. The neuron at that point can be said to have achieved hyperpolarization. At this point,
the Na+ channels are closed and become reactivated (D). Source: Reproduced with permission
from Montgomery (2010).
voltage-gated Na+ channels to open to an influx of Na+ ions (Figure 4.11), thereby causing
further movement of the membrane potential such that the inside of the neuron becomes
positive for a period.
As the neuronal membrane grows progressively more positive, the voltage-gated Na+
channels do not simply close; they deactivate. At the same time, voltage-gated K+ chan-
nels open to allow K+ ions under the influence of the concentration gradient to exit the
neuron. With this efflux, the neuron’s internal voltage grows more negative (Figures 4.11)
until it reaches a point at which the potential across the membrane reverses. This reversal
causes the neuron’s internal voltage to grow even more negative than its voltage during
rest. This entire process is known as hyperpolarization.
During a span known as the refractory period, the neuron’s depolarization deactivates
Na channels, thus preventing any subsequent polarization and, by extension, generation of
+
an action potential. This is negative feedback. Hyperpolarization subsequently reactivates

the Na+ channels, thus restoring the ability to generate an action potential. Restoration
of the normal resting membrane potential occurs as a consequence of the action of a
pump-like mechanism that expels three Na+ ions for every two entering K+ ions.
Injury current owing to physical trauma to the neuron during microelectrode
and semi-microelectrode recordings presents a special case. A tear in the cell mem-
brane caused by an electrode admits a massive influx of Na+ ions that depolarize the
neuron in a continuous manner. The resulting depolarization exceeds the threshold
for the generation of the action potential. Expiration of the refractory period sees the
neuron membrane remaining depolarized above threshold. But the neuron nonethe-
less discharges another action potential, and it continues to discharge them until such
time as the tear in the cell membrane seals or the neuron lyses. With the latter even-
tuality, water enters the neuron with the Na+ ions, causing it to burst. Though both
microelectrode or semi-microelectrode recordings and lysing neurons produce regular
patterns of action potentials at unusually high frequencies, bursting neurons’ pattern
is extremely brief, lasting a few tenths of a second. The pattern characteristic of neuron
lysis consists of regular high-frequency action potentials whose decreasing amplitude
owes to the diminishment of extracellular Na+ concentration. When heard, its sound is
that of brief buzz. To avoid confusion with other bursting patterns, one must keep the
microelectrode or semi-microelectrode still until she observes that neuronal activities
have remained stable for 20 to 30 seconds.
RECORDING NEURONAL ACTION POTENTIALS
Recording an action potential requires that one measure the difference in the voltage or
potential across two points. The difference between a battery’s cathode (negative pole)
and anode (positive pole), for example, determines its voltage. Microelectrode recording
systems accordingly feature two contacts: (1) an active contact; (2) a reference or indiffer-
ent contact. Two contacts consisting of the same material and occupying the same electri-
cal field produce no recordable difference in voltage (Figure 4.12A), because each contact
needs to encounter a different electrical field potential or voltage in order to register a sig-
nal. In what are known as bipolar recordings, the two contacts are situated fairly close to
the neural element. Bipolar recordings convey some advantage (B). (Chapter 5 contains
detailed discussion of this advantage.) In what are known as monopolar recordings, an
extremely large contact may sit some distance away from the neural element that is the
source of the signal, and an extremely fine contact may sit close to the source.
The extremely small voltages generated by axons defy recording with anything but
high amplification and fine electrode tips. Use of such electrodes, however, results in
A Voltage
B
Electrode
Neural Electrical
element Ground
field
High
Medium voltage
Low
FIGURE 4.12 Schematic representation of different electrode recording circumstances.

Situations A and B depict two electrodes in the vicinity of the neural element being recorded.
This arrangement is known as bipolar recording. C depicts a single electrode sitting close to
the neural element and an extremely large electrode, presumed to be at ground voltage, sitting
some distance from it. This arrangement is known as monopolar recording. The two electrodes
in situation A encounter the same voltage in the electric field and therefore record no signal.
Membrane
voltage
Electrical field
Ground
Axon hillock
Axon
Cell body
Dendrite
FIGURE 4.13 Schematic representation of an antidromic action potential. An action potential

initiates when depolarization of the neuronal membrane potential at the axon hillock reaches
threshold. A regenerative process conducts the action potential orthodromically along the
axon. However, the action potential backfires into the cell body and dendrites, the force of
which conducts antidromically thither in the manner of a graded potential to produce the type
of signal recorded by microelectrodes and semi-microelectrodes. The dendritic tree’s larger
volume serves to amplify the resulting electric field to a range that permits typical techniques
to record it.
high impedance. This, coupled with the high gain required, presents typical microelec-
trode recordings with significant difficulties, which are discussed in Chapter 5. One
must then decide which neural elements to record. The usual elements are the den-
drites and cell body, as action potentials in the axons are of too low voltage and too
brief to allow recordings under the typical circumstances (Figure 4.13). The direction
traveled by action potential generated at the axon hillock—namely, down the axon to
the synaptic terminals—is known as orthodromic conduction. At the same time, the
action potential backfires into the cell body and dendrites. Its movement in this reverse
direction is known as antidromic conduction. Two factors serve greatly to amplify the
antidromically conducted action potentials and thus make possible recording by use
of typical electrodes and amplifiers: (1) the volume occupied by the dendritic tree and
cell body; (2) relative synchronization of changes in voltage owing to the antidromic
invasion of the dendrites and cell body by the action potential.
RECORDING LOCAL FIELD POTENTIALS
A similar situation occurs with synchronized postsynaptic graded potentials generated

by synaptic inputs onto dendrites and cell bodies. In this case, extracellular recording
methods cannot record individual postsynaptic potentials. (Intracellular electrodes do
have the capability of recording them.) However, when postsynaptic potentials are syn-
chronized over hundreds or thousands of neurons, their summed voltages or potentials
one may record with larger electrodes (their largeness allows them to record hundreds
or thousands of neurons) having lower impedances and gains (Figure 4.14). These
summed voltages are known as local field potentials.
Inferring the underlying neuronal physiology from local field potentials proves
difficult. One may only claim with any accuracy neuronal activity of some type does
indeed occur. Scientists, however, do not limit themselves to this single claim. They
make others. Some scientists state, for instance, that the 20-Hz oscillator found in the
local field potentials during DBS intraoperative neurophysiology reflects neuronal
activity at 20 Hz. What may be technical artifact these scientists interpret as fact. In so
doing, they confuse ontology with epistemology and thus commit a categorical error,
that is, an error in which a term that takes on a particular meaning in one context is
presumed to retain this meaning when deposited in other contexts.
A local field potential indicates the presence of synaptic inputs to a set of neurons.
One must bear in mind, however, that inputs do not necessarily betoken outputs, the
finished products of a processing of inputs. One must also bear in mind that a local
field potential consists of a sum of synaptic activities. As such, it indicates an average
Elec Ground
trod
e
Postsynaptic electrical field
Dendrite
Membrane potential
Cell body
Action potential in synaptic termianl Axon
FIGURE 4.14 Schematic representation of mechanisms underlying a local field potential.

Occurring in many neurons, synchronous postsynaptic potentials sum to create a much larger
electrical field one may record with larger electrodes.
and therefore may not reflect actual present phenomena. (This second caveat receives
detailed discussion in chapters 9–11.) The example of a bimodal distribution of 10
adults and 10 infants presents a helpful example: The average height of the group of 20
individuals may be five feet, but no one member of that group stands exactly that tall.
Extracting useful information from raw local field potentials in subcortical struc-
tures proves difficult. To date, specific waveforms in the raw local field potentials lack
the same diagnostic value possessed by specific waveforms in the raw electroencepha-
logram (EEG), which essentially records local field potentials in the cortex. A spike has
specific meaning, as do a 3-s spike and slow wave complexes. The origin of the signal
in the local field potential can convey important information. Typically, this is accom-
plished by identifying phase reversals of a signal between two recording electrodes;
comparison of signals recorded at multiple contacts and from simultaneous field poten-
tials localize the signal source (Figure 4.15). Should one identify a target local field
potential, she may localize it by use of phase reversal across an electrode array.
Local field potentials

A
B
Electrical field
Electrical potential
generated by neural
element E
Neural element
F
FIGURE 4.15 Schematic representation of the determination of a local field potential signal gen-
erator’s location by use of an array of electrodes. Shown is a linear array of electrodes labeled
A B, C, D, E, and F, whose source or signal generator appears between electrodes C and D. As
a result of differential amplification (a topic receiving detailed treatment in chapter 5), the
waveform recorded in C appears as the mirror image or reverse of the waveform recorded in
D, which suggests that the source of the waveform must lie between electrodes C and D. The
progressive reduction in the amplitude of the waveforms from C to A and from D to F also sup-
ports the conclusion that the waveform source is localized between C and D.
Other important information contained within the local field potentials relate
to the frequencies contained within the local field potential. The shape of the local
field potential waveform can be considered as some mixture of sine waves of differ-
ent frequencies. Thus, a Fourier transform converts the local field potential into a
series of sine waves of different amplitudes and frequencies (Figures 4.16 and 4.17).
The amount of any single frequency (called power) can be displayed as a spectro-
gram (Figure 4.18). Local field potentials can be characterized by their associated
spectrograms.
PRACTICAL PRINCIPLES OF MICROELECTRODE RECORDINGS
An ability to detect neurons during microelectrode and semi-microelectrode record-

ings amounts to an ability to sample neuronal activities in a statistical sense. Key is
whether the sample (the neurons being recorded) reflect the population (the struc-
ture to be identified). For this reason, sampling is fraught with important issues. One
such issue involves the size of the neuron generating the recordable action potential.
The size of the signal generated is a function of the size of the neuron’s dendritic tree.
A*
B y 90°
0°
y D
Time (in seconds or degrees)
C
Amplitude
A**
Time (in seconds or

y = sine (theta) E
1 cycle, 1 period
degrees)
y = cosine (theta)
or y = sine (theta + 90°)
FIGURE 4.16 Schematic representation of periodic functions such as sine waves. Consider a

race car on a circular track (A). When viewed from above, the car is seen as driving in a circle.
When viewed from the side or level of the race track, the car appears to be moving back and
forth. Depending on the perspective from the side view, the car may be seen as starting at the
middle of the race track and the back and forth (B) appearing movement would trace a sine
wave over time (C). If viewed from a different perspective (A*), the car would be perceived
as starting at one end and the movement back and forth (D) over time would trace a cosine
wave (E). The diameter of the race track determines the amplitude of the car movement back
and forth and would be the amplitude of the sine and cosine waves. The time it takes the car
to make one complete circle of the race track, when viewed from above, and the time it takes
for the race car to move from and then back to the starting point (and going in the same direc-
tion) would correspond to the period of the oscillation (circular movement when viewed from
above) or the length of a single sine or cosine wave. The number of circles or complete back
and forth movements per second would be the frequency and is equal to 1 divided by the
period. Source: Reproduced with permission from Montgomery (2010).
Most microelectrodes and semi-microelectrodes recordings are biased toward larger

neurons, and even at that only a small fraction of neurons will be identified in the
recordings. Consequently, caution needs to be exercised when making inferences
from the neurons recorded at the microelectrode tip and the neurons characteristic of
the structure being recorded.
A second issue, which is related to the first, involves the distance one must move
the microelectrode or semi-microelectrode before she can assure herself that she is
in fact recording a different neuron. Should she observe persisting patterns of neu-
ronal activities describing the same waveform, odds are that she is still recording the
same neuron, notwithstanding the fact that she has advanced her microelectrode or
semi-microelectrode. This author makes a practice of advancing the electrode at least
250 μm before he begins to record new neuronal activities.
C
A
B •
•
FIGURE 4.17 Building on the situation described in Figure 4.16 of the characteristics of a race
car moving on a circular track and producing sine waves when viewed from the side, now
imagine that the race car is on a small circular track (A) which itself is circling on a larger
track (B) with that track circling on an even larger track (C). The time it takes for the race car to
traverse a single cycle on the small track will be short and when viewed from the side or level
of the track will appear to have a high frequency. The race car also will be circling on track B
at a slower frequency and circling on track C at an even lower frequency. When viewed from
the side, the back and forth movement of the race car will be seen as complex as shown in
D. However, the complex movement shown in D can be considered as a sum of the back and
forth movements on the individual tracks (see Figure 4.18). Source: Reproduced with permission
from Montgomery (2010).
Original waveform
Spectrogram
Power
Frequency
FIGURE 4.18 A schematic example of a spectrogram. The complex waveform shown is the
same as that of Figure 4.17, which itself is a composite of three sine waves. Thus, the com-
plex waveform can be considered as the sum of the component sine waves. Breaking down
a complex waveform into its components is called a Fourier transform, whereas construct-
ing a complex waveform from simple sine waves is called an inverse Fourier transform. The
spectrogram shows the breakdown of the complex waveform into its three components. The
amount of each component is represented as the power at that frequency.
Other issues involve the density, or number of neurons per unit volume, of brain
tissue within the range of the microelectrode tip and a neuron’s discharge frequency.
A slowly discharging neuron may require several seconds of recording in order to
detect an extracellular action potential. In cases of slowly discharging neurons, sev-
eral seconds may elapse before one detects a recordable extracellular action potential.
Recording “on the fly”—that is, recording as one advances the electrode—reduces the
time per electrode trajectory, provided the electronic descends slowly enough to stand
a reasonable chance of recording low-frequency neuronal extracellular action poten-
tials. Though preferable to a method that requires one to advance and halt an electrode,
“on the fly” recording sometimes engenders its own difficulty in the form of mechanical
or electrical interference.
A microelectrode or semi-microelectrode sometimes exerts a shearing force on
adjacent brain tissue. Specifically, it drags tissue along with it as it advances. Two
forces are at work: that of the stretching brain and that of the friction generated by the
advancing electrode. The first eventually comes to exceed the second, at which point
the tissue returns to its earlier place with a jerk. The danger of shearing force runs
particularly high with simultaneously using several electrodes situated relatively small
distances between them. Though generation of shearing force does not prevent the
microelectrode or semi-microelectrode tip from recording activity as it approaches a
neuron, the subsequent retraction of shearing force may cause the recording to be lost.
When such a loss occurs, withdrawing the electrode a small distance may allow one to
recover the recordings.
Postoperative MRI scans analyzed by this author have shown that use of multielec-
trode arrays injure the brain to a notable extent. Yet, whether multielectrode arrays
may be said to increase risk of adverse outcomes remains unclear. Because movement
of the microelectrode or semi-microelectrode through the brain may cause physiolog-
ical as well as anatomical changes, recording neuronal activities as electrodes are with-
drawn may prove less successful than recording them as the electrodes are advanced.
PRACTICAL PRINCIPLES OF LOCAL FIELD POTENTIAL RECORDINGS
To the best of this author’s knowledge, methods for recording local field potentials
for target localization have neither been perfected nor standardized. The discussion
to follow therefore relates to the general principles of local field potential recordings.
An important unresolved matter involves two considerations: (1) the spatial resolu-
tion of information content amenable to sampling by local field potential recordings;
(2) the size of the electrodes used to localize field potential recordings. Summing
postsynaptic potentials over a quantity of neurons sufficient for recording a signal

requires larger electrodes. Information loss occurs, however, in such instances as
when the requisite surface area of the local field potential electrode surpasses in size
the spatial distribution of one unit of information. For example, consider the homun-
cular representations as units of information. One unit of information may relate to
the arm representation, while the leg representation may relate to a different unit of
information. If the local field potential recording volume spans both homuncular rep-
resentations, or both units of information, the actual result will be some confusing
combination of the two (Figure 4.19). It remains unclear whether the types of elec-
trodes used in local field recordings, such as the contacts on a DBS lead, possess spatial
resolution sufficient for identifying specific regions in the sensorimotor homunculus
of the target, or even for differentiating sensorimotor from nonmotor regions of the
nucleus that contains the target.
Again, it remains unclear whether there exists in raw local field potentials any sig-
nal of sufficient specificity and sensitivity for identifying a target. Investigators have
considered frequency content and other secondary measures. In the case of DBS for
Parkinson’s disease, for instance, they have particularly scrutinized the power (amount)
Electrode A
Signal from
electrode A
Electrode C
Electrode B
Signal from
electrode B
Signal from
electrode C
FIGURE 4.19 Schematic representation of the effects of spatial resolution of local field poten-
tial recordings. Consider the motor homunculus. Electrodes A and B are small relative to the
side of the homunculus such that electrode A can record from the upper extremity region
while electrode B can record from the lower extremity region. Electrode C is large relative
to the homunculus and will be pick up local field potentials generated in both the upper and
lower extremities and the signal recorded will be some “average” of the local field potentials
from the upper and lower extremities. In this hypothetical case, the waveforms associated
with the upper extremity are exactly same shape as that for the lower extremity but opposite
in polarity. This means that when the two are “averaged” together to form the signal for the
large electrode C, the result will be no signal. At the least, this hypothetical case raises con-
cerns and cautions relative to the use of local field potentials for target localization.
Electrode A
Averaged signals
From electrode A
Singals from
electrode A
Averaged signals Electrode B

From electrode B
Signal from
electrode B
FIGURE 4.20 Schematic representation of the hypothetical use of evoked potentials to localize
the homuncular representation. In this case the patient is asked to repeatedly move the arm
and just the arm. The local field potentials from both electrodes A and B are recorded during
each trial. As can be seen, the signals from each trial may be difficult to interpret. However,
when averaged over multiple trials, a clear signal emerges from electrode A that is not seen
from electrode B. The signal in electrode A is correlated with movement of the arm and con-
sequently, it may be inferred reasonably that electrode A, as opposed to electrode B, is in the
arm homuncular representation.
of components of the raw signal occupying the 20-Hz frequency band. Methods involv-
ing secondary measures require that one apply Fourier transformations—operations
predicated on the idea that at specific frequencies complex signals are decomposable
into a series of sine or cosine waves—to determine the power of any single frequency in
a raw local field potential. Some investigators have found that the 20-Hz frequency in
the Fourier transform increases in brains of patients with Parkinson’s disease.
Averaged local field potentials evoked by movement may help to localize the sen-
sorimotor region. Figure 4.20 shows a hypothetical example of how movement-related
evoked potentials may be used to target specific homuncular representations.
Motor-evoked potentials that achieve sufficient resolution, specificity, and sensitivity
prove useful to the task of identifying DBS targets.
SUMMARY
Used to relieve a wide range of neurological and psychiatric disorders, DBS exploits
the fact that the brain is basically an electrical device. One may also make use of the
electrical properties of neurons (the brain’s fundamental constituent) in order to locate
the appropriate target for the DBS lead. Making use of electrical signals requires that
one understand the way in which these signals are generated and the methods by which
they can be recorded.
The neuron is essentially a battery and a circuit for generating signals, albeit com-
plex. The circuitry contains different mechanisms that modulate the signals generated,
which are in turn mediated primarily by protein channels traversing the neuronal
membrane’s lipid bilayer. The electric current controlled by these channels consists of
ions and affects the voltage signal one records during intraoperative neurophysiologi-
cal monitoring. The types of channels belonging to a neuron’s various parts possess
unique electronic capabilities. Each capability one may assess by use of microelectrode
and semi-microelectrode recording of extracellular action potentials in the axons, local
field potential recordings of summed postsynaptic potentials in the dendrites and cell
bodies, or some other technique of electrical recording.
REFERENCE
/ / / 5 / / / MICROELECTRODE AND
SEMI-MICROELECTRODE
RECORDINGS
Electronics
INTRODUCTION
Chapter 3 covered electricity’s fundamental characteristics, and chapter 4 covered its

applicability to intraoperative neurophysiological monitoring of neuronal activities.
The present chapter covers electricity’s manipulation for the purpose of measuring
and analyzing neuronal activity. Quite complicated is the business of converting the
extremely low voltage signals generated, in the case of extracellular action potentials,
by neurons and, in the case of local field potential recordings, by combined postsyn-
aptic graded potentials. The many available “turnkey” systems greatly simplify the
routine measurements of extracellular action potentials and local field potentials, but
monitoring frequently requires troubleshooting. As helpful and responsive as device
manufacturers strive to be, the individual performing the monitoring ultimately finds
herself bearing the responsibility for rectifying any malfunctions. Indeed, she finds
herself obliged to understand the basic concepts underlying the electronic devices. In
doing so, she not only becomes an effective troubleshooter; she also ensures successful
monitoring. Any individual unable or unwilling meet this responsibility is not recom-
mend to performed monitoring of patients.
As discussed in chapter 1, at issue is the maximal spatial volume that provides the
fundamental unit of electrophysiological information necessary for the optimal local-
ization of the target. As suggested in chapters 1 and 4, for relatively large structures
such as the globus pallidus interna and the ventral intermediate nucleus of the thala-
mus, the size of the sensorimotor region is large with respect to the effective volume of
86
5. Microelectrode and Semi-Microelectrode Recordings / / 87
stimulation available on current DBS electrodes. Further, the homuncular representa-

tion is laid out in this larger volume. Consequently, one cannot just place the DBS elec-
trodes in the sensorimotor region in the globus pallidus interna or ventral intermediate
thalamus and expect that the appropriate homuncular representation will be within the
volume of tissue activation during DBS.
Consider the hypothetical situation shown in Figure 5.1, in which local field poten-
tials are used for target localization. In this case, the patient has predominantly upper
extremity problems and so this region of the homunculus is the appropriate target. As
can be seen, if the electrode is very large relative to the arm region of the homunculus,
it may not be able to differentiate the local field potentials unique to the arm representa-
tion and, consequently, may not precisely define the target. It cannot be assumed that
semi-microelectrode recordings offer the necessary spatial resolution. Microelectrode
recordings clearly are capable of the needed spatial resolution.
The issue of spatial resolution also relates to the resolution of the manipulation of
the electrodes. Consider the situation with the subthalamic nucleus. The subthalamic
nucleus presents a small DBS target, measuring a mere 5.9 mm in the anterior–poste-
rior direction, 3.7 mm in the mediolateral dimension, and 5 mm in the dorsoventral
Electrode A
Signal from
electrode A
Electrode C
Electrode B
Signal from
electrode B
Signal from
electrode C
FIGURE 5.1 Schematic representation of the effects of spatial resolution of local field potential
recordings. Consider the motor homunculus. Electrodes A and B are small relative to the side
of the homunculus, such that electrode A can record from the upper extremity region while
electrode B can record from the lower extremity region. As electrode C is large relative to
the homunculus, it will pick up local field potentials generated in both the upper and lower
extremities and the signal recorded will be some “average” of the local field potentials from
the upper and lower extremities. In this hypothetical case, the waveforms associated with
the upper extremity are exactly same shape as those for the lower extremity but opposite
in polarity. This means that when the two are “averaged” together to form the signal for the
large electrode C, the result will be no signal. At the least, this hypothetical case raises con-
cerns and cautions relative to the use of local field potentials for target localization.
direction (Richter et al. 2004). The target becomes yet smaller when one recalls that the
actual target is the sensorimotor region, a subset of the volume in question. Assuming
that the smallest move from one trajectory to a new trajectory is 2 mm, if one is in
the middle of the sensorimotor region, then any movement will be at least 2 mm and
risks placing the DBS electrodes outside of the sensorimotor region of the subthalamic
nucleus. For that reason and the fact that volume of tissue activation relative is rela-
tively large compared with the volume of the sensorimotor region, there is little reason
to attempt to specify the exact homuncular representation in the subthalamic nucleus.
Cases where the fundamental volumes of information may amount to approxi-
mately 250 μm, the level of the neuron’s dendritic tree, require one to use a microelec-
trode. A typical microelectrode tip has a cone shape, its base measuring some 20 μm in
diameter, or, roughly, the span of three red blood cells placed side by side. The micro-
electrode’s extremely small tip generates high impedance that may significantly dis-
tort the extracellular action potential waveform. On the order of 50 μV to 100 μV, the
voltage charge generated by extracellular action potentials and recorded by microelec-
trodes may require amplification of 10,000 times or so. This high gain amplifies not
only the extracellular action potential but also any electronic “noise” and artifact in the
microelectrode recording systems and the environment.
High impedance and gain cause electrodes to pick up a great deal of electronic “noise”
and artifact. Filtering microelectrode recordings thus becomes necessary in order to dis-
tinguish extracellular action potentials’ actual waveform. Filtering may distort the wave-
form of the extracellular action potentials, however, and may also distort electronic “noise”
and artifact, which one may mistake for actual neuronal extracellular action potentials.
Intraoperative neurophysiologists must understand and control all of these factors.
AMPLIFIERS
The operational amplifier (op am) serves here as a paradigmatic example of an amplifier.
A typical operational amplifier features two inputs: (1) a noninverting input, designated
positive (V+) (this input is discussed below), and (2) an inverting input, designated nega-
tive (V–). The output is designated Vout. Supply voltage powers the operation amplifier.
This supply voltage has positive (VS+) and negative (VS-) sources (Figure 5.2).
Differential in this configuration, the amplifier emits a signal (Vout) whose value
is the difference between the inputs (V+ and V-) multiplied by the gain (G), accord-
ing to Vout = G (V+ - V-). If V- does not equal zero, then V+ does not equal Vout. In other
words the emitted signal Vout differs from the signal being recorded. If V+ equals V-, then
the output Vout equals zero (Figure 5.3). Certain unique advantages and disadvantages
Vs+
V+
Vout
V–
Vs–
FIGURE 5.2 Schematic representation of an operational amplifier. The amplifier’s two inputs

(V+ and V-) are situated in comparison. The output (Vout) is the difference between the two
inputs. Supply voltages typically draw power from a positive (VS+) and negative (VS-) source.
characterize outputting the difference between the two inputs in the differential ampli-
fier. These advantages and disadvantages are discussed below.
A series of feedback circuits allow one to control amplifier gain by changing resis-
tance. The emitted signal out (Vout) may not exceed the supply voltages (VS+ and VS-);
indeed, it often falls short of them. Increasing gain beyond the supply voltages results in
saturation and signal loss (Figure 5.3).
A
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
Vin V+
2 Vout 0V
1.8 V–
1.6
1.4
1.2
1
0.8
0.6 Resistorg Resistorf
0.4
0.2
0
Signal ground
B
2
1.8 Vs+ 2
1.6 1.8
1.4 1.6
1.2 1.4
1 Vin V+ 1.2
0.8 1
0.6 Vout 0.8
0.4 Vs– V– 0.6
0.2 0.4
0 0.2
0
Resistorf
Resistorg
Signal ground
FIGURE 5.3 Schematic of an operational amplifier whose negative feedback passes through

a resistor (Resistor f). The gain or amplification is determined by the ratio of Resistor f and
Resistorg. Adjusting the value of Resistor f also adjusts the amplification or gain. Configured as
differential, the amplifier in A generates output whose value is the difference between the two
inputs. Because the two inputs are exactly the same, the output equals zero. The amplifier in
B demonstrates the effect of “clipped” output that results when an input’s signal exceeds the
positive (VS+) and negative (VS-) supply voltages.
IMPEDANCE AND IMPEDANCE MATCHING
During intraoperative neurophysiological monitoring, a signal must pass through ampli-

fiers, filters, and other devices before arriving to the neurophysiologist (Figure 5.4). These
devices themselves consist of operational amplifiers, capacitors, resistors, and other com-
ponents. The connections or interfacings among all of these components have effects on
the signal beyond what would be desired. These effects, one of which is impedance, can
alter the signal (see Chapter 3). Impedance results from the resistance of the conductors
(related to how tightly the outer electrons are held by the atom) and reactance. Electricity
involves two kinds of reactance: (1) inductive and (2) capacitive (see chapter 3).
A common phenomenon of energy that passes through different media, reactance
is best illustrated by another common phenomenon: reflection. Someone holding a
flashlight may illuminate an object situated on the other side of a single-paned window
by directing the flashlight’s beam at that object. Some light, however, fails to reach the
object; it is instead redirected. That is, an amount of electromagnetic energy is lost in
the attempt to illuminate the object through the glass. This happens as a consequence
of different impedance. Air impedes light far less than glass. At the point of the two
media’s interface there results reactance in the form of reflection.
C D
E
A F G
A to D Computer
converter system
FIGURE 5.4
Schematic representation of devices used in microelectrode and
semi-microelectrode recordings. The neuron is represented in A. The microelectrode or
semi-microelectrode is represented in B. Represented in C is a unity-gain amplifier known as a
cathode follower or high impedance probe, which is discussed in greater detail below. Rather
than amplifying the signal, the purpose is to match microelectrode or semi-microelectrode
impedance at the input to the input impedance of the next stage of amplification, which is
represented by amplifier D. Filters represented by E receive the output of amplifier D. The
analog-to-digital (A to D) converter (F) in turn receives the analog signal from filters E and
digitizes it. The digital signal passes next to the computer system, which analyzes and dis-
plays it.
A double-paned window only increases the amount of reflection or impedance.

A flashlight beam shone through it experiences a doubling of the reactance experi-
enced by the beam shone through the single-paned window. That is, the beam passes
through air, glass, air again, glass again, and air yet again on its way to the object; and at
each site of interface of the two media the beam loses energy. The amount of light lost
with each reflection being small, not much effect is observed. The recording systems
require electrical current flowing through multiple points of contact that may result
in electrical energy passing from on medium with a certain impedance into another
with a different impedance (impedance mismatch). This passing between mediums
with different impedances produces pronounced reactances that not only change the
recordings of the desired signal but also create and magnify artifact and noise.
Reactance occurs when an energy signal enters a medium of different impedance
compared to the medium just left but also will experience reactance when exiting the
medium into one that has different impedance. That is why one speaks of an input and
output impedance of an electrical difference and why both the input and output imped-
ances should be matched to the preceding and subsequent devices.
Minimizing impedance mismatch helps to preserve the fidelity of signal recording and
reduces artifact and noise. The risks and effects of mismatched impedances are particularly
significant if one component in the recording system is at high impedance. In the case of
microelectrodes, the small electrical contact surface area usually means high impedance.
Further, the impedances typical of the subsequent components, such as amplifiers and fil-
ters, are usually low. Figure 5.5 illustrates the cumulative effect of impedance mismatches,
which one may describe as the difference between the actual voltages. This difference con-
sists of electrical energy generated in the neurons (Vactual in Figure 5.5), and the voltage
registered, upon passing through the entire amplifier system (Vmeasured 2, Figure 5.5), by a
microelectrode (Vmeasured 1, Figure 5.5) or another measuring device.
The error or loss of signal that occurs between the neuron and the microelectrode
tip or at some other electrical interface is approximately related to the ratio of the
impedances (Z 0 / Z1 in Figure 5.5). Equal impedances (Z 0 = Z1 in Figure 5.5) lead to
minimal signal error. Extremely high output impedance in neurons owes to the fact
that the lipid bilayer of the dendrites and cell body is highly resistive to the flow of
electrical charges that occurs when the action potential backfires, or back-propagates
(see c hapter 4). Microelectrode tip impedance (Z1 in Figure 5.5) must therefore be
extremely high—a fortuitous requirement given the fact that the small electrode tips
necessary to isolate the extracellular action potentials in a single neuron also generate
high input impedance.
Vactual
Z2
Vmeasured 1
Z0 Z1
Z3
Error in V = Vactual–Vmeasured 1
~ (Z0/Z1)*Vmeasured 1
Vmeasured 2
Error in V = Vmeasured 1–Vmeasured 2
~ (Z1/(Z2+Z3))*Vmeasured 2
FIGURE 5.5 Schematic representation of a series of electrical devices used to record neuro-

nal activity. First, current flows through the neuronal membrane and encounters impedance
(Z0). Electric current generated by the neuron produces a second current that flows through
the conductor to the electrode tip via capacitive coupling (see Chapter 3). The current again
encounters impedance at the electrode tip (Z1) and as it flows from the electrode (Z 2). The sig-
nal encounters yet more impedance as it exits the electrode and enters the amplifier (Z 3). The
signal recorded and displayed may therefore differ from the signal generated by the neuron.
The difference owes to error in V, which is related to all impedances in the path.
Similar concerns apply to the electrical interface between the microelectrode and
the rest of the system’s amplifier devices. For example, mismatch between the input
impedance of the microelectrode (Z1 in Figure 5.5) and the microelectrode output
impedance (Z2 in Figure 5.5), may cause signal loss. Also, because impedance depends
on the frequencies contained in the signal (see chapter 3), the signal becomes distorted
as its component frequencies are altered. Consequently, one would like the output
impedance of the microelectrode to be extremely high as well.
Next, one must consider the output impedance of the microelectrode relative to the
input impedance of the rest of the amplifier system. An amplifier’s typical input imped-
ance may be on the order of 20,000 ohms, or 20 kiloohms (20 kΩ), which does not match
the microelectrode’s impedance of 600 kΩ (0.6 mΩ). These circumstances require that
one use a special preamplifier known as a cathode follower or high-impedance probe,
a unity-gain amplifier that does not boost the signal but rather matches the microelec-
trode’s output impedance with its high input impedance and the input impedance of
the rest of the amplifiers with its low output impedance.
A sudden change in the microelectrode input impedance, such as that which occurs
when insulation breaks off and leaves the electrode with a larger surface area, might
cause a significant reduction in the impedance and a mismatch that alters the signal.
Increased impedance may also result from fracturing of the microelectrode tip. From
this results mismatch and consequent alteration of the recorded signal. One must
therefore check microelectrode impedance if there should be any sudden change in the
recording quality.
Recording systems for local field potentials typically involve much lower impedances,
because local field potentials have a much larger surface area compared with the micro-
electrode. In such instances, cathode follower preamplifiers are generally not necessary.
The amplifier gains required of microelectrode and semi-microelectrode record-
ings higher than those required of local field potential recordings, extraneous signals
may be detected. Known as “noise” or artifact, these extraneous signals, which tend to
interfere with the recording of the intended signal, may arise from the target volume of
brain tissue, the amplifier system, or the environment. Of these three potential sources,
the third, in most cases, proves to be the more difficult one.
Necessary for establishing the homuncular representation of the recording site,
identifying the extracellular action potentials of individual neurons may prove difficult
in the case of microelectrode recording. Depending on electrode tip size and imped-
ance, semi-microelectrode and local field potential recordings do not encounter a simi-
lar problem, but they sacrifice resolution, particularly in terms of differentiating the
homuncular representation. Using microelectrodes with finer tips can reduce “noise”
emitted by adjacent neurons. Using too fine a tip, however, makes it difficult to isolate a
specific neuron and hold it for recording and analysis.
“Noise” internal to the amplifier systems comes in the following forms: (1) thermal
noise arising from the spontaneous and random movement of electrical charges (ions or
electrons) whose movement produces currents that are registered by microelectrodes
and (2) “shot” noise owing to the fact the electrical charge is quantized in units of a sin-
gle electron, which causes microelectrode recordings to display fast (high-frequency)
transients (transitions from a state of X electrons to X+1 electrons back to X electrons).
Of unknown etiology, “flicker” noise is signal frequency dependent: The amplitude
of noise stands in inverse proportion to frequency (~ 1/f). “Flicker” noise is thus great-
est at low frequencies, particularly at frequencies related to so-called DC recordings
(discussed below). Crosstalk between adjacent lines carrying different signals in close
proximity may affect each other by electrostatic (capacitive coupling) and magnetic
noise (induction). This holds true especially in cases of extreme difference among
impedances and voltages between the lines. Extreme differences exist, for example,
between analog and digital conductor (electric communication) lines. Digital lines
typically have higher voltage than analog lines, digital line voltage ranges between 0
and 5 volts. Rapid transition between 0 and 5 volts in digital lines generate extremely
high-frequency electromagnetic fields that may produce inductive and capacitive cur-
rents (see chapter 3). One is thus advised to keep these lines as separate as possible and
shielded, if appropriate (shielding is discussed in chapter 6).
Analog multiplexers can be a source of noise. Many systems generate multiple ana-
log input signals that must be converted to a single A/D (analog to digital) unit. In
this case, the A/D converting unit “samples” each analog channel at a discrete time.
A relatively large number of sampled analog signals may experience considerable
change in the analog signal value between sequential samples that appears as a sudden
or high-frequency change in the recording. A “sample and hold” circuit employed by
some systems preserves the analog sample value at the commencement of a series of
A/D conversions until the series for that sampling is complete.
A/D conversion is also prone to “digitization” noise, which arises from binary digi-
tization of a continuous analog signal (the analog input). For example, a 0.3-v analog
signal translates to a binary signal of 10011. The last bit of signal, however, will “toggle”
between 0 and 1, that is, it will oscillate between 10011 and 10010. The oscillation reg-
isters as high-frequency noise in the digital line and represents the least significant bit
of information in the signal, particularly in those instances where the analog and digital
communication lines lie in close physical proximity.
One can take certain steps to reduce the amount of “noise” and artifact affecting
the signal of interest (see Chapter 6). It is highly unlikely, however, that any amount of
prevention will completely eliminate noise or artifact. One must therefore have at the
ready filtering and other ways to deal with them.
FILTERING
Filtering is based on the fact that any complex signal, such as the analog signals
recorded and analyzed in intraoperative neurophysiological monitoring, can be
decomposed into a series of component sine waves of different frequencies, ampli-
tudes (power), and phases by use of a Fourier transform (see chapter 4, Figures 4.16
and 4.17). Many sources produce noise of various frequencies. These frequencies
may be higher or lower than those of the signal of interest, that being extracellular
action potentials in the use of microelectrodes and semi-microelectrodes. Movement
artifact and 60-Hz power-line noise in microelectrode or semi-microelectrode
recordings, for example, typically have lower frequencies than those of the signal of
interest. One may therefore filter them quite effectively at those frequencies. Radio
frequency and digital-line noise frequency are typically much higher than the com-
ponents of the signal of interest.
Filters are designed to reduce the amplitude or power in a specific frequency or

range of frequencies. Consider a complex signal that is sent to a filter. The complex sig-
nal can be decomposed into its component sine waves as shown in Figure 5.6 A–C. The
performance of the filter is shown in the graph as a sigmoid (backward “S”-like figure)
that indicates the degree to which a signal at any given frequency is reduced. Signals
of a frequency to the left of the downslope of the curve are not reduced, while those to
the right of the downslope are significantly reduced. The output signal from the filter
similarly can be decomposed into its component sine waves, but, as can be seen in com-
parison with the input signal, the frequencies represented by C* and D* are reduced
in amplitude compared with the input signal C and D. As can be seen, the amplitudes
of the low frequencies are unaffected, and consequently this type of filter is called a
low-pass filter. The frequency at which the amplitudes are significantly reduced is called
the cutoff frequency.
A B C D 2.5
2 2 2 2
1.5 1.5 1.5 1.5
1 1 1 1
0.5 0.5 0.5 0.5
0 0 0 0
Low pass
1
Output gain
Frequency
2 2
1.5 1.5
1 1 2
0.5 1.5
0.5 1 2.5
0.5
2
0
1.5
0
1
0.5
0 0
A* B* C* D*
FIGURE 5.6 Schematic representation of the effects of a filter on signals entering the filter.

Consider a complex signal that is decomposed into its component sine waves, for example
by a Fourier transform (see chapter 4), in A, B, C, and D. The filter reduces the amplitudes of
some of the components but not others as seen by comparing A, B, C, and D to A*, B*, C*,
and D*, respectively. The ratio of the amplitudes between A, B, C, and D to A*, B*, C*, and D*,
respectively is represented by the graph appearing as a backward “S.” As can be seen, fre-
quencies to the left of the downslope are unchanged, whereas those to the right are reduced.
The frequency associated with the downslope is the cutoff frequency. As the low frequency
components are unaffected, this type of filter is called a low-pass filter.
The cutoff frequency can be changed with a corresponding change in the output
to change which frequencies are reduced and by how much (Figure 5.7). This can be
very useful. For example, consider a signal of interest whose fastest component, in
a Fourier transformation, is somewhat less than the frequencies contained in noise
or artifact. One could adjust the cutoff for a low-pass filter to be just above the maxi-
mum frequency of the signal of interest but less than the frequency of the noise. For
example, consider where the maximum frequency in the signal of interest is 20 Hz
but the recordings are contaminated with AC power-line noise at 60 Hz. One could
have the cutoff tuned to 25 Hz thereby reducing the 60-Hz noise without greatly
affecting the 20-Hz signal of interest. However, if the signal of interest also is at 60
Hz, then this filter would reduce the signal of interest at the same time as it reduced
the noise.
A high-pass filter (Figure 5.8) is the converse of the low-pass filter in that frequen-
cies above the cutoff are preserved while those below the cutoff are reduced. The effects
A B C D 2.5
2 2 2 2
1.5 1.5 1.5 1.5
1 1 1 1
0.5 0.5 0.5 0.5
0 0 0 0
Low pass
1
Output gain
Frequency
2
1.5
1 2
1.5
0.5 1
0.5 2
1.5
0
1
0 0.5 2.5
2
1.5
1
0 0.5
0
A* B* C* D*
FIGURE 5.7 Schematic representation of the effects of a filter on signals entering the filter but
with the cutoff moved to the left as compared with Figure 5.6. As can be seen, frequency A is
not affected but frequency B* is now reduced compared with the situation described in Figure
5.6, a Fourier transform (see chapter 4), in A, B, C and D. The amplitudes of frequencies C*,
and D* remain reduced.
A B C D 2.5
2 2 2 2
1.5 1.5 1.5 1.5
1 1 1 1
0.5 0.5 0.5 0.5
0 0 0 0
High pass
Output gain
Frequency 2.5
2.5 2
2 1.5
1.5 1
2
1.5 1
1 0.5 0.5
0.5 0 0
0
2
1.5
1
0.5
0
A* B* C* D*
FIGURE 5.8 Similar to Figure 5.7, the schematic representation is of a high-pass filter. A com-
plex signal (not shown) is decomposed by a Fourier transform into its component sine waves,
A, B, C, and D. The output of the filter is a complex signal that now has the component sine
waves shown by A*, B*, C* and D*, respectively. As can be seen, the amplitudes of sine waves
A and B have been reduced to A* and B* respectively, while the amplitudes of sine waves C
and D remain unchanged.
of low- and high-pass filters on a complex signal are shown in Figure 5.9. It is important
to appreciate that filters also may have an effect on the signal of interest, such as the
neuronal action potential, particularly distorting its shape (Figure 5.10).
A notch filter combines a low- and high-pass filters so that a small range of fre-
quencies are reduced while frequencies above and below that range are unchanged
(Figure 5.11). Notch filters often are used specifically to reduce the amplitudes of the
AC power-line interference, often called the 60-Hz notch filter.
An important concept, harmonics denotes frequencies that are whole-number mul-
tiples or fractions of the primary frequency of interest. A 20-Hz frequency, for example,
has an initial supraharmonic (harmonic frequency above the frequency of interest)
of 2*20 (40) Hz, a second supraharmonic at 3*20 (60) Hz, and so on. The 20-Hz fre-
quency of interest also has a subharmonic at 20/2 (10) Hz. The amplitudes of these har-
monic frequencies are typically less than the primary frequency of interest, but filters
may affect the harmonic frequencies of the primary frequency of interest as well.
Construction of signal of interest Effects of filters on noise and

and noise Signal of interest
A 1.82 D 2
1.8
1.6 2 H1.6
1.4
1.2
1.4
1.2 1.5 1
1 0.8
0.8 1 0.6
0.6 0.4
0.4 0.5 0.2
0.2 0
0 0
B 1.82 E G 1.82 2
1.6
1.4 2 1.6 I 1.8
1.6
1.4 1.4
1.2 1.2 1.2
1 1.5 1 1
0.8 1 0.8 0.8
0.6 0.6 0.6
0.4 0.5 0.4 0.4
0.2 0.2 0.2
0 0 0 0
2.5
C F 2 J 2
2 1.8
1.5 1.6
1.5 1.4
1 1.2
1 1
0.5 0.8
0.5 0.6
0 0.4
0 0.2
0
FIGURE 5.9 Construction of Signal of Interest and Noise. Schematic representation of a signal

of interest (A) that is affected by noise (B). The Fourier transform shows that the recorded
signal (B) is made up of sine waves C through F. The sine waves D and E are thought to be
components of the signal of interest while those of C and F are noise. Subtracting components
C and F from the recorded signal (B) produces the signal of interest (A).Effects of Filters on
Noise and Signal of Interest.Schematic representation of filters imposed on the signal (com-
posed of the signal of interest and noise) B and G. Figure H shows the effects of a high-pass
filter whose cutoff is set slightly above the low-frequency artifact (C in Figure 5.9). Figure
I shows the effects of a low-pass filter whose cutoff is set slightly below the high-frequency
noise (F in Figure 5.9). The bandpass filter consists of low- and high-pass filters imposed on
the signal of interest (J).
Neuronal action potential

waveform
Component frequency A
Component frequency B
FIGURE 5.10 Schematic representation of the potential effects of filters on the neuronal action
potential, particularly distorting its shape. Most neuronal action potentials are not a simple
sine wave and consequently contain multiple sine wave components. In this case the neuronal
action potential is constructed of two sine waves, as an example. If the cutoff of a low pass fil-
ter is just greater than component frequency A, the effects of a low-pass filter will change the
shape of the neuronal action potential to that of component A. A high-pass filter will change
the shape of the neuronal action potential to that of component A, both different from the
original shape of the neuronal action potential.
A B C D 2.5
2 2 2 2
1.5 1.5 1.5 1.5
1 1 1 1
0.5 0.5 0.5 0.5
0 0 0 0
Low pass High pass

1
Output gain
Frequency
2.5
2 2
2
1.5 1.5
1.5
1 1
1
0.5 0.5 2
1.5 0.5
0 0 1
0.5
0
0
A* B* C* D*
FIGURE 5.11 Schematic representation of notch filter that is composed of a low- and high-pass
filter that results in a reduction of frequencies within a specific range such that frequencies
above and below the range are unchanged. In this example, a complex signal is sent into the
filter (unseen) and is decomposed into its component frequencies, A, B, C, and D. As can be
seen, only frequency C is reduced (C*), while the other frequencies above and below C are
unchanged.
AC- AND DC-COUPLED AMPLIFIERS
AC-coupled (as distinguished from DC-coupled) amplifiers are commonly used for
intraoperative microelectrode and semi-microelectrode recordings. The “AC” and “DC”
designations refer not to alternating and direct current, but to the frequency compo-
nents possible in the output, which begin at zero frequency, that is, a flat, unchanging
voltage of any value.
Such frequency responses in DC-coupled amplifiers may aid recording of local field
potentials and extremely low frequencies. These amplifiers, however, tend to “drift,”
that is, they tend to move slowly to different levels. Drift may require frequent adjust-
ment. This is true particularly when the DC-coupled amplifier has only just been
switched on, its warming up affecting the electrical resistance in components.
The AC-coupled amplifier output typically filters low frequencies or DC levels
and therefore better resists drift and other sources of noise or artifact. Because low
frequencies are less of a concern in microelectrode and semi-electrode recordings,

AC-coupled amplifiers confer an advantage.
COMMON-MODE REJECTION
Differential amplifiers enjoy an important advantage with respect to noise and artifact.
Essential to this advantage is the concept of common-mode rejection. The difference
between the two input signals constitutes the differential amplifier’s output signal. Any
signal component occupying both inputs cancels (Figure 5.12).
The property of common-mode rejection can be exploited to remove artifact,
provided the latter is common to both inputs (Figure 5.13). In the case of the two
inputs, optimal common-mode rejection confronts a problem: Two electrodes must
register the exact same noise. If one electrode registers the noise at twice the ampli-
tude or power registered by the other electrode, common-mode rejection reduces
the noise but does not eliminate it.
For the two electrodes to register the same noise or artifact, impedance or some
other characteristic of the two electrodes must be identical. As shown in Figure 5.5, the
Inactive or reference
electrode 2
Active electrode
A B C D
Inactive or reference electrode 1
FIGURE 5.12 Schematic representation of the employing common mode rejection to remove

noise. In the figure, the circle shows the electrical field generated by the signal of interest,
such as an extracellular neuronal action potential. The background represents the noise and
as can be seen the noise changes from the left to the right (A–D). The active electrode in zone
A records from both the background noise A and the signal of interest. Inactive or reference
electrode 1 records the background noise in A only. However, the inactive or reference elec-
trode 2 records a different noise (C). Thus recording between the active and the inactive or
reference electrode 1will reject the noise in common resulting in a “cleaner” signal. However,
recording between the active and the inactive or reference electrode 2 will have residual noise,
as the two electrodes are recording different noise. When the active electrode is paired with
the inactive or reference electrode 2, the output will contain the signal of interest but also resid-
ual noise, which will be the difference between the noise of zone A and the noise of zone C.
Straight connectors Twisted pair connectors

Active electrode Active electrode
Differential amplifier
Differential amplifier
Output Output
Reference or indifferent Reference or indifferent

electrode electrode
FIGURE 5.13 Schematic representation of the effects of common-mode rejection. The dif-

ferential amplifiers have two inputs. The output of the differential amplifier is the difference
between the two inputs. In the figure to the left appear two signals. The first signal contains
background noise as well as an extracellular action potential. The second signal contains
only the noise, albeit at a reduced amplitude. The output of the differential amplifier con-
tains the signal, extracellular action potential, and noise. The figure to the right depicts both
inputs. Each input contains equal amplitude noise, whereas only one contains the extracel-
lular action potential. The noise thus “subtracts out,” leaving only the extracellular action
potential.
measured voltage, which is a function of the actual voltage, is modified by the ratio of
output and input impedances encountered at the electrode. If noise has the same volt-
age but the two electrodes register different impedances, each electrode will have a dif-
ferent measure of the noise or artifact. This difference will be relayed to the differential
amplifier. As a consequence, noise or artifact will remain.
Similarly important is the distance of each of the two electrodes that run inde-
pendently from the differential amplifiers’ inputs to the source of noise. Just as long
electrical wire offers greater resistance than does short wire, a greater amount brain tis-
sue occupying the distance between the electrodes offers greater resistance than does
a smaller amount, and from this results a different signal. Optimum common-mode
rejection requires a large distance between the sources of artifact or noise and a short
distance between the two recording electrodes, which must be as electrically identi-
cal as possible. Such optimum conditions, however, often prove impossible to create.
Adequate distance can be established between implanted electrodes and any source of
noise or artifact, for instance, but noise originating in the implanted patient can present
a problem. Examples of artifact or noise arising from the patient include electromyo-
graphic activity, particularly from the muscles in the head, and microelectrode chafing.
(Other sources are discussed in chapter 6.)
MONOPOLAR RECORDINGS
Recordings in a bipolar configuration make use of common-mode rejection, because both

electrodes are electrically close to each other. Also possible are recordings in a monopolar
configuration. In this configuration, the electrode left uninvolved in recording the sig-
nal of interest—which is known as the reference or indifferent electrode, and which runs
to the inverting input—lies, in an electrical sense, infinitely far away. Even monopolar
recordings, however, utilize differential amplifiers. This utilization rests on the assump-
tion that the indifferent or reference electrode possesses a zero, or ground, voltage.
The patient, rather than a six-foot copper alloy rod or some other external structure,
typically grounds the indifferent electrode. Doing so allows one to avoid leakage or
the formation of a ground loop current, the latter of which may electrocute the patient.
Situating the ground in the patient reduces this risk.
Important for both recording and stimulation, the task of electrically isolating the
patient from external power lines supplying electricity is aided by the fact that batteries
power the devices physically attached to the patient, and thus the amount of electrical
energy that would propagate to the patient is limited. One may also use isolation trans-
formers (Figure 5.14), which permits no physical contact between the external power
lines and the components that touch the patient.
Transformer
Voltmeter A
Voltmeter B
Conductor
Primary winding
Secondary winding
Magnetic flux
FIGURE 5.14 Schematic representation of a transformer one may use to isolate the power
supply from the power to devices connected to the patient. The transformer consists of two
physically separated coils of wire. These coils encircle a cord that transmits magnetic energy.
Connected to the AC power line, the first coil (primary winding) represents a potentially infi-
nite source of electrical current. Electrical current flowing through the primary winding causes
a magnetic flux and induces electrical current and voltage in the secondary coil, which may be
located in one of devices attached to the patient. Physically isolated, these circuits protect the
patient from higher voltages and currents in the AC power line.
The reference or indifferent contact used in monopolar recordings must have a zero
voltage. One therefore should not place the contact over muscle or other tissue capable of
generating electrical signals because doing so may introduce noise or artifact arising from
electrical potentials. One must instead place the indifferent electrode over a bony promi-
nence in such a way as to prevent intermittent disruption of the electrical connection.
AUDIO MONITORING
Most amplifier systems include an audio monitor that transmits processed microelec-
trode and semi-microelectrode recordings to a loudspeaker. One may identify extracel-
lular action potentials, noise, and artifact by their characteristic sounds. Three kinds
of neuronal activity patterns—spontaneous, stimulation induced, and movement
induced—characterize specific anatomical regions that aid identification of the opti-
mal DBS target. These patterns are more readily detected by the ear than by the eye.
Effective audio monitoring requires that one use an effective audio monitor that fea-
tures low-pass, high-pass, and notch filters. Often there is very high frequency but low
amplitude background noise that produce a hissing sound that can make discrimina-
tion of the signals of interest difficult. A type of filter, squelch control reduces extremely
high-frequency, low-amplitude background hiss, provided the signal-to-noise ratio is
favorable. The speakers’ dynamics should occupy a range of approximately 20 Hz to 24
kHz. The system should also feature an audio jack for headphones one would need for
contending with any ambient noise.
OTHER USEFUL FEATURE
One may implement a variation of squelch controls prior to transmitting the signal to
the audio monitors. A level or window discriminator isolates the waveforms associated
with the extracellular action potentials of interest (Figure 5.15). For example, one may
set an amplitude or voltage level slightly above a level deemed to be noise but below
the maximum amplitude of the extracellular action potentials of interest. Each time the
recordings exceed the threshold the audio monitor receives a signal to produce a clicking
sound. The pattern of clicks thus represents the pattern of extracellular action potentials.
Disadvantages attend use of the output of the spike discriminator for the signal to
the audio monitor. Left undetected might be extracellular action potentials of smaller
amplitude, which likely represent neurons in the background and some distance from
the electrode tip. These neurons may also alter their activity in response to movement
or stimulation. This alteration of activity aids identification of the recording site.
B
C
D
FIGURE 5.15 Schematic representation of a spike discriminator used to “clean” the signal to the
audio monitor. A raw microelectrode recording (D) sent directly to the audio monitor contains
noise whose amplitude nearly equals the amplitude of the spikes. This noise makes it difficult to
discern the extracellular action potentials. A threshold discriminator (C), however, detects those
waveforms that exceed the threshold and that produce a pulse (B). The train of pulses transmits
to the audio monitor (A). One hears as a result only a “click” for each waveform. On this click she
must rely as a way of distinguishing the signal from the raw microelectrode recording.
Level-discriminator output sent to a device allows one to determine the instantaneous

frequency (the reciprocal of the interspike interval) or the average frequency, which may
be of aid in identifying the recording location. Reported frequencies may not actually
reflect the discharge frequency of a particular neuron, because multiple neurons may be
recorded simultaneously. But the average frequency does offer some rough indication.
D B
FIGURE 5.16 Screenshot of microelectrode recording system made during DBS surgery.

A shows the microelectrode recording (the image’s incomplete availability owes to its having
been “painted” as the screenshot was taken). B shows the continuous display of neuronal
activity extracted from the raw microelectrode recordings. This continuous display enables
direct visualization of the firing patterns. Direct visualization in turn enables one to assess
regularity of neuronal bursts and other phenomena. C shows the instantaneous discharge
frequencies of the neurons in the microelectrode recordings. D shows the input from the
accelerometer attached to the user’s hand in order to capture the time of onset of the behav-
ioral event. E shows the peri-event histogram of neuronal activity relative to the behavioral
event. The histogram clearly demonstrates a relationship between the behavior and a change
in neuronal activity. Demonstrably similar changes with each trial, as evidenced by the raster,
validates the histogram (E).
The physiological characterization of a neuron’s activities depends on demonstrated

presence or absence of any correlation between changes in the neuronal activities and
a behavioral event. The behavioral event may be peripheral nerve receptor activation,
such as occurs in passively rotating a joint or stroking the skin, or it may be an active
joint rotation initiated by the patient. One estimates this event by listening to the neu-
ronal discharge activities occurring during it. Yet this method proves problematic.
Some centers have developed systems for constructing a peri-event raster and histo-
gram of neuronal activities centered on a specific behavioral event over repeated trials
of the behavioral event (Figure 5.16).
SUMMARY
Devices used in intraoperative neurophysiological monitoring share a common design.

An electrode that interfaces with the neural tissue connects to an amplifier. A unity-gain
preamplifier aids one in matching the impedances among the microelectrode or
semi-microelectrode and the subsequent amplifier systems. Though common to the
recording environment, noise and artifact may be minimized through a number of elec-
tronic “tricks,” such as filters and common-mode rejection in differential amplifiers.
These devices and “tricks” require that one attend closely to impedance and other basic
principles of electronics.
REFERENCE
Richter EO, Hoque T, Halliday W, et al.: Determining the position and size of the subthalamic nucleus
based on magnetic resonance imaging results in patients with advanced Parkinson disease. Journal
of Neurosurgery 100(3): 541–546, 2004.
/ / / 6 / / / NOISE AND ARTIFACT
INTRODUCTION
An unavoidable part of all intraoperative neurophysiological monitoring, noise and arti-

fact can be minimized to a point at which they no longer interfere with procedures, pro-
vided one identifies their source and type. Noise and artifact may equally be mistaken
for neuronal activities. Artifact and noise may be prevented or, at the least, mitigated.
This chapter discusses noise and artifact, identifies their respective sources, and presents
methods for prevention and mitigation. For the most part, noise and artifact relate to
electrostatic, magnetic, and electromagnetic interference. (The types of noise discussed
in chapter 5—thermal noise, shot noise, flicker noise, multiplexor noise, and analog-to-
digital [A/D] conversion noise—receive no further mention in the present chapter.)
Fundamentally, noise and artifact are due to unwanted movement of electrical
charges in the recording systems. Thus, noise and artifact are due to the forces that move
electrical charges, particularly electrons in the recording systems. Unfortunately, these
forces also are what move electrical charges to generate the signals of interest, such as
the neuronal extracellular action potentials and local field potentials. As discussed in
chapter 3, because electrons carry a negative electrical charge, they are made to move by
repulsion from a negative charge or by attraction to a positive charge, and these charges
cause electrostatic fields. Because electrons have negative electrical charge and they
are spinning on their axis, their spin creates a magnetic charge (a dipole of north and
south magnetic poles) and consequently, changing magnetic fields can move electrons.
An electromagnetic field, or radiation, contains both electrostatic and electromagnetic
components and hence, can effect electron movements in two ways.
Understanding the sources of electrostatic, magnetic, and electromagnetic forces,
other than those generating the signal of interest, greatly facilitates their identification
and mitigation, preferably by elimination. As will be demonstrated, certain circumstances
are more likely to produce the different forces, and, further, the exact circumstances will
10 6
6. Noise and Artifact / / 107
produce artifact and noise whose characteristics match, in some way, the characteristics of
the electrostatic, magnetic, and electromagnetic forces that produce the artifact and noise.
For heuristic purposes, noise will be defined as interference that is continuous
and often periodic, which means that noise tends to vary in a characteristic pattern or
rhythm. Artifact, on the contrary, will be defined as brief and arrhythmic, although the
artifact can reoccur in a periodic or rhythmic fashion. What is meant by arrhythmic is
that within the artifact there is no recurring stereotypic pattern. There will be exceptions,
but those exceptions are relatively few and by being few are fairly easily remembered.
NOISE
Perhaps the most typical and most ubiquitous example of noise is the interference
caused by AC power lines. The electrical current carried by the AC power line typically
is a sine wave where the current increases, decreases, reverses and increases, decreases,
reverses again and increases in a continuing pattern (Figure 6.1). Current is the flow of
electrons per second and is analogous to the velocity of an object, for example, a car.
When velocities increase or decrease, they are said to accelerate and decelerate, respec-
tively. As discussed in chapter 5, when an electrical charge accelerates (or decelerates)
it produces an electromagnetic field or radiation that will go out into the environment.
If unchecked, the electromagnetic field radiated into the environment will invade
objects in the environment. If the electrons orbiting the atoms are tightly bound to the
atoms that make up the object, the electrons will not move. However, if the orbiting
electrons are loosely held by the atoms in the object, such as in a metallic conductor,
A 1 2 3 4 5 6 7 8 9 B AC power
line
Conductor Conductor
in recording in recording
system system
+
Current
+
Current
– 0
– 0
FIGURE 6.1 Schematic representation of electromagnetic interference caused by AC power

lines. A shows a cycle of current flow in a power line. As the current increases, there is an
acceleration of electrons (A 1–3) then a deceleration (A 3–5) and then an acceleration in the
opposite direction (A 5–7) and then a deceleration (A 7–9). During the accelerations and decel-
erations electromagnetic radiation is given off, and when the electromagnetic field invades a
nearby conductor (B), a current in the opposite direction is created. If the conduction is part of
the electrophysiological recording systems, a noise will be seen and heard.
those electrons will move and produce an electrical current in the conductor. This is
because the electrostatic component of the electromagnetic radiation will cause the
electrons in the conductor to move. In addition, the magnetic component of the electro-
magnetic radiation also will cause electrons in the conductor to move. The movement
of electrons induced in the conductors of the recording systems will generate noise.
The AC power lines running to nearly every electrical device in the operating room
represent the most common source of noise. The electrons induced to move in the
recording systems will move in phase with the movement of the electrons in the AC
power lines. Thus, the frequencies of the noise will mirror the frequency of the AC power
lines, which are at 60 Hz in North America and 50 Hz in Europe. The resulting 60-Hz
noise enters the audio monitor, where it manifests as a hum. So common is power-line
noise that most amplifier systems and audio monitors feature a 60-Hz notch filter (see
chapter 5). Harmonics of the 60-Hz AC power line may also contribute to any noise.
Depending on their propensity for radiating a 60-Hz electromagnetic field, certain
devices are likelier than others to produce noise. A wire coil of a kind used in electro-
magnets and transformers (Figure 6.2), for example, produces a strong electromagnetic
field with AC inputs (or magnetic field with DC inputs) and thus tends to produce noise
as well. Used in power supplies, transformers change the 120 volts of the typical AC
power line to another needed voltage (or between other voltages)—5 volts DC for most
computers, for instance, or + and – 12 volts DC for many operational amplifiers. The
Transformer
Voltmeter A
Voltmeter B
Conductor
Primary winding
Secondary winding
Magnetic flux
FIGURE 6.2 Schematic representation of a power supply transformer. The transformer consists

of a core that conducts magnetic fields. The primary winding moves electrical current to the
transformer and generates the magnetic flux. The magnetic flux moves through the secondary
winding to induce a current and voltage in the circuit connected to the secondary winding. The
ratio of windings determines voltage in the secondary circuit (voltmeter B). A fewer number of
turns on the secondary winding than on the primary winding indicates lower voltage in the sec-
ondary circuit (voltmeter B) compared with the primary circuit (voltmeter A). The magnetic field
generated by the transformer may spread far beyond its source. A magnetic field that reaches a
nearby conductor induces an electric current in it, and from this encounter results noise.
cathode ray tubes (CRTs) in televisions and some computer monitors, on the contrary,
require voltages much higher than 120 volts.
The transformer relies on induction to change the voltage. As shown in Figure 6.2,
a typical transformer consists of two coils placed in proximity: a main coil, or primary
winding, through which runs the AC power-line current to generate a fluctuating
60-Hz magnetic field; and a subordinate coil, or secondary winding that the magnetic
field “invades,” the effect of which is 60-Hz induction of current.
The numbers of turns of wire on the primary and secondary coils (or windings) are
different. A greater number of turns of the primary winding vis-à-vis the secondary
winding results in lower voltage exiting the transformer. A smaller number of turns of
the primary winding vis-à-vis the secondary winding results in greater voltage exiting
the transformer. Transformers whose exiting voltage is lower than the voltage entering
it are known as step-down transformers. Transformers whose exiting voltage is higher
are known as step-up transformers.
Because they require voltage higher than that which courses through the AC line,
fluorescent lights use step-up transformers and are therefore a major source of 60-Hz
noise. Cathode ray tubes, which have the same requirement, also produce 60-Hz noise.
Liquid crystal displays (LCDs), however, do not need step-up voltage transformers and
therefore produce less noise than do CRTs.
In order to minimize effects on monitoring, one must keep power supplies and
other devices as far away from the recording systems as possible. Failing this, one may
shield power supplies and other devices in order to reduce the number of magnetic
fields emitted into the local environment.
When situated in proximity to electrical current, electric motors and other devices
containing wire coils produce significant electromagnetic fields. Microelectrode and
semi-microelectrode drivers, warm-air circulating blankets, intravenous fluid pumps—
these common fixtures of neurophysiological operating rooms feature motors that may
cause artifact by propagating electrical current to the patient.
Not at all exhaustive, the above list of artifact sources offers a sense of the ubiquity
of noise. One must therefore remain mindful of the many devices and situations she
may need to consider when troubleshooting noise. Indeed, even light, radio transmis-
sions, and pager and mobile phone signals have an electrical and a magnetic compo-
nent. Yet, whether they produce significant noise depends on their frequency. Digital
signal lines, for example, operate at frequencies well above 60 Hz. Because imped-
ance and capacitance act as filters (see c hapter 5), however, the electromagnetic fields
of digital signal lines may reach a frequency low enough to interfere with electrical
recordings.
Conductor A
–
Electrostatic field
Conductor B
+
FIGURE 6.3 Schematic representation of capacitive coupling between two conductors.

Conductor A carries a current productive of a negative voltage, which in turn produces an
electrostatic field that “pushes” or repels electrons in nearby conductor B. These electrons
thus driven away, conductor B acquires a voltage and current that may produce noise.
Capacitive coupling may also produce noise (Figure 6.3). Increasing and decreasing
numbers of electrons in a segment of conductor, as well as other kinds of fluctuating
electrostatic charges, can induce the same fluctuation in another separate conductor,
even in the absence of contact or the presence of insulation between them (see portion
of Chapter 3 under the heading “Capacitors”). Voltage carried by a conductor may thus
propagate to a nearby conductor through capacitance.
Capacitive coupling noise plagues particularly devices that generate electrostatic
charges. For example, a CRT, which houses a highly negatively charged and heated
cathode that emits electrons, generates large electrostatic charges. A positively charged
anode attracts the emitted electrodes, which accelerate as they move toward it. They
strike the phosphorescent front of the CRT. This impact generates the light necessary
for the image to appear. Any source of static electricity can produce interference, but
typically artifact rather than noise is produced.
As discussed previously, moving lines of magnetic force can cause electrons to flow in
a nearby conductor. If that nearby conductor is part of an electrophysiological recording
system, interference may be produced. However, a conductor moving relative to a static
magnetic field still can induce electrical current in the conductor. A typical example is
when conductors, such as the cables that connect the electrodes to the amplifiers, sway
in the magnetic fields generated in the environment. Typically, the interference is con-
tinuous and at the same frequency of the swaying conductors, typically low frequency.
Differentiated from artifact based on its greater persistence, noise arrives for the most
part from electrical sources. Mechanical sources, however, can generate artifact suf-
ficiently continuous to produce artifact indistinguishable from noise. Such noise can be
generated by rhythmically changing mechanical effects—mechanical vibrations transmit-
ted to the connectors, particularly those connecting the preamplifier (cathode follower or
high-impedance [Z]‌probe) and the microelectrode or semi-microelectrode, for example.

These vibrations produce a rhythmic change in the physical connection, which in turn
brings about in the electrical connection a change of the sort produced when one causes a
bulb to flicker by wiggling a loose wire to its socket.
Such disruptive vibrations may emanate from loudspeakers, large electric motors,
patient’s speech, the act of treading the floor near the patient and recording equipment,
patient’s respiration, or force arriving from other parts of the building that propagate to
the patient.
Any rhythmic change in the spatial relationship of a microelectrode to its target
neurons can produce noise. Cardioballistic effects—pulse waves generated by the
heart—move the brain relative to the microelectrode or semi-microelectrode. As that
brain moves with each cardiac contraction, the electrode tip moves closer to the target
neurons one moment, further from them the next. As a consequence of this change in
distance there occurs a change in the resistance of the brain tissue interposed between
the electrode tip and the neuron and a concomitant change in the voltage recorded.
The recorded voltage ebbs with every increase of the electrode’s distance from its target
neurons and recrudesces with every decrease, the result of which is a rhythmic voltage
signal. One may time this signal to the patient’s pulse.
Ground loop currents represent yet another source of noise. The mistaken belief that an
electrical ground generates no voltage falls before the fact that small electrical potential dif-
ferences may obtain between different grounding sources. Two or more connected devices
possessing different ground potentials may produce noise as a result of current flow from
one ground through the ground with a lower ground potential into an electrical device
causing artifact (Figure 6.4). The same holds true for a patient connected to a ground.
ARTIFACT
Defined for heuristic purposes as transient, artifact is primarily mechanical in nature and
involves unstable electrical connections. A phenomenon characterized by intermittent
bursts of sound, “electrode popping” occurs as a result of poor mechanical connection,
most often that between the microelectrode or semi-microelectrode and the preamplifier.
A second phenomenon, microdrive chatter, occurs when a microelectrode or
semi-microelectrode slides unevenly and intermittently within its guide cannula. From
this results an irregular mechanical vibration that persists for the duration of the micro-
electrode or semi-microelectrode’s movement.
Yet another source derives from the use of mechanical gears whose teeth intermesh
in order to transfer rotations that drive the electrode. Wear or damage causes the gears
A Connection between devices

with ground loop noise
Device A Device B Device C Device D
B Connection between devices

without ground loop noise
Device A Device B Device C Device D
FIGURE 6.4 Schematic representations of two methods of grounding connected electrical

devices. A depicts a “daisy chain,” a type of serial ground connection. In such an arrangement
ground-wire resistance increases at each point of connection. Thus, whereas device D meets
single resistance, device A meets the sum of all resistance met by devices A, B, and C. The vari-
ous degrees of resistance produce the voltage observed at each device’s ground connection,
giving rise to a ground loop current. Connecting each device directly to the ground (B), how-
ever, results in little difference of resistance and thus gives rise to no ground current capable of
producing noise. The signal sent between devices will not experience any ground loop noise.
to increase in vibration and thus create artifact. Though artifact ceases the moment
the microdrive stops, for the time that it persists it may slow the microelectrode or
semi-microelectrode’s descent and thereby prolong recordings.
Although not specifically artifact, neuronal injury discharges are brief and unex-
pected repetitive neuronal signal emissions (Figure 6.5). The microelectrode or
semi-microelectrode causes a tear in the neuronal membrane that permits sodium
(Na+) to enter the neuron. From this influx follow a massive depolarization and con-
comitant repetitive action potentials, as water enters with Na+ and causes the neuron
to burst. The repetitive extracellular action potentials end with the neuron’s destruc-
tion. The concern here is that one might take injury currents for artifact.
PREVENTING NOISE AND ARTIFACT
Noise reduction and artifact elimination depend on observing the following sound
electrical practices:
A 1 B
0.6 s
C 1 D
2 3
FIGURE 6.5 Examples of a rapid transient neuronal activity whose decreasing amplitude likely
owes to an influx of Na+ ions and subsequent depolarization as a consequence of a tear in the
neuronal membrane. Example A depicts a decreasing discharge (Arrow 2). The initial spike
(Arrow 1) likely indicates the moment at which the tear occurs. From this follows a rapid flux in
ionic currents and repetitive action potentials as a consequence of subsequent neuron depo-
larization. Example B depicts a progressively increasing repetitive discharge likely initiated
by a small tear in the neuronal membrane. The influx of Na+ ions introduces water to the neu-
ron, causing it to swell. As it swells, the tear grows larger, permitting more Na+ and water to
enter. The increased influx causes further depolarization and an increased rate of discharge.
Discharge ceases upon membrane lysis. C depicts another example. The downward arrows
(1) indicate mechanical artifact as the microelectrode advances and subsequent chaffing as
it settles in position. Arrow 2 indicates microdrive artifact. Mechanical irritation engenders
a sequence of extracellular action potentials that dissipate over time. One risks mistaking
mechanical artifact associated with Arrow 1 for an extracellular action potential. In order to
provide supportive evidence of artifact, the expanded view includes more oscillations than
one would observe in an extracellular action potential. Example D depicts characteristic
high-frequency repetitive discharge of decreasing amplitudes indicative of an injury current.
1. Ensuring that all devices work properly and that the electrode’s impedance and
other electrical properties are properly matched.
2. Making use of the shortest possible cables and connectors to reduce the amount
of noise current caused by induction and capacitance.
3. Making use of twisted-pair cables and connectors to differential amplifiers in order
to ensure that both inputs to the amplifiers detect the same artifact (Figure 6.6).
(Amplifiers with adequate common-mode rejection cancel noise and artifact.)
4. Ensuring firm, steady electrical connections by checking connections for corro-
sion, which is indicative of poor or intermittent electrical contact. (Gold-plated
connectors are less likely to corrode. One must use the tips of the alligator clip,
for example, to connect inputs to the preamplifier to the microelectrode or
semi-microelectrode, or to connect the ground inputs ([Figure 6.7].)
Straight connectors Twisted pair connectors

Active electrode Active electrode
Differential amplifier Differential amplifier
Output Output
Reference or indifferent
electrode Reference or indifferent
electrode
FIGURE 6.6 Schematic representation of the respective effects of straight and twisted pair
connectors from the electrode to the inputs of the differential amplifier. As discussed in
Chapter 5, the differential amplifier output consists of the difference between the two inputs.
If the two inputs share the same noise, as shown for the twisted pair connection, the noise
cancels, leaving the signal. Use of twisted pair connectors ensures that both inputs share
the same noise. In the straight connection, however, one input may encounter a different,
lower-amplitude version of the noise. The subtraction of the two inputs consequently permits
some noise in the outputs to remain.
5. Securing wire connections between the preamplifier and electrodes in such a

way as to prevent the wire’s swaying and thereby producing noise or artifact.
6. Ensuring relative fixity of the electrode apparatus and the head in order to mini-
mize any movement productive of noise or artifact.
7. Making use of bipolar microelectrode or semi-microelectrode recordings in
order to take advantage of common-mode rejection.
8. Ensuring that electrode and preamplifier electrical cables running to the main
amplifier system avoid coming into close proximity with such sources of noise as
the following common operating room equipment:
a. Power supplies
b. AC power lines
Improper Proper
FIGURE 6.7 Schematic representations of the improper and proper method for attaching an alli-
gator clip to a conductor. The distal portion of the clip must make contact with the conductor.
c. CRTs. (Their use is best avoided. If they must be present, then they should be
switched off and unplugged. One must remember, however, that power bricks
for laptop computers and other devices continue to have electric current run-
ning through them when switched off.)
9. Avoiding use of electric motors. (If their presence is unavoidable, they should
be switched off, unplugged, and placed at a distance from the patient and
recording equipment.)
10. Avoiding use of fluorescent lights. (Incandescent lights should be used instead. If
fluorescent lights must be used, they must be shielded. One should bear in mind
that incandescent bulbs are being phased out and replaced by compact fluores-
cent light (CFL) bulbs. In such situations where incandescent bulbs are unavail-
able, LED bulbs should be preferred over CFL bulbs.)
11. Refraining from “daisy chaining” power cables belonging to electrical devices
connected to the patient or to equipment used for electrophysiological record-
ings. (In such an arrangement there may result a ground loop current, because
none of the ground wires share the same ground ([Figure 6.4A].)
12. Minimizing as much as possible any electrical connections between the patient
and any electrical devices. (Electrocautery used by some surgeons requires a
grounding pad that is attached to the patient and to an electrical cable whose
other end connects to the device. One must disconnect the grounding pad when
not in use, preferably at the connection that lies closest to the patient. If left con-
nected, the cable becomes an antenna capable of picking up noise.
13. Avoiding use of operating rooms near sources of large electrostatic or electromag-
netic noise. (No large electric motors driving compressors, sterilization equip-
ment, elevators, X-ray, CT, or MRI should be in the vicinity of the operating room.
If their presence cannot be avoided, then they must be shielded. Failing this, the
operating room itself must be shielded—an expensive remedy, admittedly, and
perhaps not worth the trouble in light of the fact that noise and artifact are also
generated within the operating room.)
SHIELDING
One shields by surrounding electrodes, cables, connectors, and other devices with
metal foil, metal screen, or some other conductive barrier. Doing so reduces the effects
of electrical noise and artifact. The shield takes advantage of induction to reduce noise
generated by an electromagnetic field. The electromagnetic field induces currents in
Conductor emitting EMI
Shielding
Conductor
FIGURE 6.8 Schematic representation of the effects of a shield around a conductor in the pres-
ence of electromagnetic interference (EMI). The source of the EMI radiates electromagnetic
radiation (field) that ordinarily would induce currents in the nearby conductor, likely result-
ing in noise. In the case of a shielded conductor, the electromagnetic field generated by the
source induces electrical current in the shield that counters the effects of the electromagnetic
field, thereby reducing the noise.
the shield that counteract the electromagnetic field (Figure 6.8). Shielding also may
block an electrostatic field by absorbing the electrostatic field and shunting the capaci-
tively coupled charges to ground (Figure 6.9).
One may shield individual cables. Many electrical cables, for example, consist of
wires wrapped in metal foil, a design that for the most part protects them from electro-
static and electromagnetic noise. Yet the metal foil itself becomes a capacitor capable
of generating capacitive-coupling noise. Grounding the foil shield bleeds off any elec-
trostatic charge accumulated on the foil surface. The foil shield is typically grounded
at one end in order to avoid ground loop currents.
A Faraday cage (also known as a Faraday room), whose ceiling, walls, and floors
consist of conductive material, proves effective in terms of reducing electrostatic and
Conductor emitting
electrostatic field
Conductor
Shielding
Ground
FIGURE 6.9 Schematic representation of the effects of a shield around a conductor in the pres-
ence of electromagnetic interference (EMI). The source of the EMI radiates electromagnetic
radiation (field) that ordinarily would induce currents in the nearby conductor, likely resulting
in noise. In the case of a shielded conductor, the electrical charges in the shield as a conse-
quence of capacitive coupling are “drained off” by the ground.
electromagnetic noise and artifact; but it requires cautious use, because it may increase
capacitive-coupled noise and artifact. Some construct a double-sided Faraday room,
whose inner shielded surface is electrically isolated from its outer shielded surface.
Doors compromise a room’s shielding and must therefore bear double-layered shield-
ing. Specifically, the door’s inner surface must be electrically continuous with the inner
shielded surface of the room, and the outer shield surface of the door must be electri-
cally continuous with the outer surface of the room.
Similar but of opposite polarity to the electrostatic field, a second field arises on
the shield’s surface. This occurs in a manner similar to the capacitor described in
chapter 3. Like a capacitor, the shield’s surface becomes saturated, thus negating the
effects of the electrostatic noise.
GROUNDING
Grounds used in electrical recordings are of three types: (1) signal ground, (2) chassis

ground, and (3) earth ground (Figure 6.10). Each serves a different purpose and is sus-
ceptible to different noise and artifact problems.
Input +
Input –
Signal
ground
Conductive
guide
cannula
Chassis
Indifferent ground
electrode
Active electrode
Micro-or semi- Earth

microelectrode ground
FIGURE 6.10 Schematic representation of different grounds. The microelectrode and the guide
cannula are connected to the differential amplifier. The ground on the guide cannula is not in
electrical contact with the chassis ground. The chassis ground is connected to the earth ground.
The signal ground relates to amplifiers and other devices’ voltage inputs. Voltages
are relative in the sense that they reflect the differences in electromotive potential, that
is, the energy to move electrical charges. In the present instance, this movement is that
of electrons from cathode to anode. One imagines a situation in which two sets of volt-
ages or signals enter as inputs to a differential amplifier. The voltages of one set are +1
volt and +4 volts relative to earth ground. The output from the differential amplifier is
+3 volts. In the other set the voltages are -1 volt and +2 volts relative to earth ground.
The output of the differential amplifier is likewise +3 volts. One may resolve the ambi-
guity between +1 and +4 or –1 and +3 volts by referencing the active or noninverting
inputs and the indifferent, reference, or inverting inputs to a zero-voltage measure. In
this case, the signal ground acts as this reference.
Bioelectrical signals of interest generated in the patient ultimately must be trans-
mitted to the recording or stimulating device both of which typically are powered by
a connection to the AC power line. If there is a direct electrical connection from the
patient to the recording or stimulating systems to the patient, there is the risk of a short
circuit, which means that the full and potentially unlimited electrical current of the AC
power line could be injected into the patient.
To eliminate this risk, one may isolate devices directly attached to the patients from
the AC power line or other potentially infinite electrical current. Because most devices
feature a battery or another independent power source, the maximal current is limited.
(Safety mechanisms also serve to limit maximal current.)
A chassis ground may prevent dangerous currents and voltages from reaching the
patient. Metallic or conductive enclosures around electrical devices may conduct elec-
tricity. For example, a short circuit in a conductor may propagate an electric charge or
current to the patient via the conductor’s contact with a metallic enclosure unless the
patient is isolated from the device power supplies and chassis ground and if the imped-
ance from the chassis to earth ground is very low. The chassis ground functions to shunt
the electrical charge or voltage to the ground and away from the patient, provided it is
highly conductive in terms of its material and size (diameter).
A single constant zero-voltage ground on hand is of great value, because, as men-
tioned earlier, all voltages are relative. But this ground should be isolated from the
chassis ground as described above. In cases in which the earth serves as this reference,
the ground established is known as an earth ground. An earth ground requires that
one connect the grounding point to a low-resistance, high-current conductor that car-
ries capacity to the earth. A buried six-foot copper alloy rod provides the best means of
establishing an earth ground. No such ground available or practical, a pipe belonging
to a building’s infrastructure suited the purpose in the past, but many pipes found in
buildings are now made of plastic. One may no longer assume therefore that a section
of metal pipe at the point of attachment is in sufficient electrical continuity with metal
pipe underground.
REDUCING NOISE AND ARTIFACT
This section deals with noise and artifact that prevails against the aforemen-
tioned measures. The first issue is to recognize whether the microelectrode or
semi-microelectrode’s signal is noise, artifact, or neuronal activity. Most often the
pattern of the noise and artifact permits unambiguous identification. In such circum-
stances in which the pattern does not permit identification, another method becomes
necessary.
1. One must first determine whether the signal suspected of being noise persists
as the electrode is moved. Noise or artifact frequently persists and neuronal
activities increase or decrease with the electrode’s changing proximity to the
neurons.
2. Any apparent noise that persists after one has grounded the inputs is artifact, its
site of generation lying somewhere between the grounding site and the audiovi-
sual display.
3. Once one has determined the signal to be noise or artifact, she must then deter-
mine whether it is continuous, nearly continuous, or intermittent.
4. Noise, by convention in this text, is continuous or nearly so, and one must deter-
mine its frequency. A humming sound suggests a 60-Hz noise (see Item 4.1).
A hissing sound suggests high-frequency, irregular noise (see Item 4.2). One
must ascertain next whether extracellular action potentials wax and wane in
time with the patient’s pulse (see Item 4.3).
4.1. The humming suggestive a frequency of 60 Hz requires that one take the
following actions:
4.1.1. Ensuring that all the electrical devices and electrode apparatus are
in proper working order and are properly connected.
4.1.2. Ensuring that the electrode impedance occupies the proper range,
replacing the electrode if necessary.
4.1.3. Ensuring that the preamplifier is properly grounded to the metallic
guide cannula or another electrode apparatus.
4.1.4. Ensuring that the amplifier systems’ 60-Hz notch filter and audio
monitors are engaged.
4.1.5. Ensuring that no AC power lines and cables of other devices lie near
any cables attached to electrode recording equipment.
4.1.6. Ensuring that the electrical operating room table is switched off and
unplugged.
4.1.7. Switching off any fluorescent lights, including fluorescent light sub-
stitutes for incandescent light bulbs.
4.1.8. Disconnecting any grounding pads connected to the patient, pref-
erably at a location closest to the patient.
4.1.9. Switching off overhead operating lights.
4.1.10. Replacing any faulty microelectrode or semi-microelectrode.
4.1.11. Switching off and unplugging any unneeded electrical devices.
4.1.12. Systematically toggling operating room circuit breakers, ensur-
ing beforehand that the power to the electrode recording systems
remain on. (One may import, if necessary, a power source external
to the operating room.)
4.1.13. Adjusting, when the situation warrants it, audio monitor output in
order to render audible only the neuronal action potentials. (Some
amplifier systems are capable of detecting any extracellular action
potentials that cross a certain threshold. In such cases, the extracel-
lular action potential becomes barely audible, noise making distin-
guishing it difficult ([Figure 6.11].).
4.1.14. Increasing the high-pass filter cutoff frequency to a prime-number
frequency greater than 60 Hz in order to avoid harmonics of the
60-Hz noise (79 Hz, for example).
B
C
D
FIGURE 6.11 Schematic representation of a spike discriminator used to “clean” the signal to the
audio monitor. A raw microelectrode recording (D) sent directly to the audio monitor contains
noise whose amplitude nearly equals the amplitude of the spikes. This noise makes it difficult to
discern the extracellular action potentials. A threshold discriminator (C), however, detects those
waveforms that exceed the threshold and that produce a pulse (B). The train of pulses transmits
to the audio monitor (A). One hears as a result only a “click” for each waveform. On this click she
must rely as a way of distinguishing the signal from the raw microelectrode recording.
4.1.15. Inspecting any electrical devices in the vicinity of the operating

room for any potential source of noise and attempting to switch off
and unplug those sources. 4.1.16. Aborting the surgery or
implanting the DBS lead by use of macrostimulation to guide loca-
tion should reduction or noise and identification of extracellular
action potentials prove impossible.
4.2. Hissing suggestive of high frequency electrostatic or electromag-

netic noise requires that one take the following actions:
4.2.1. Implementing above items 4.1.1 through 4.1.14 in order to determine
whether the noise is simply noise or actually signal. (Intentional
or unintentional filtering can cause the 60-Hz signal to sound like
high-frequency noise.)
4.2.2. Determining whether high-frequency transmission lines, digital
communication channels, Ethernet, USB, monitor, or printer cables
have produced high-frequency noise either through inductive or
capacitive coupling.
4.2.2.1. Ensuring that the cables described above do not lie near
electrodes, cables, and recording devices.
4.2.2.2. Switching off these devices when possible.
4.2.2.3. Reducing low-pass-filter frequency to a point low enough to
eliminate noise while leaving unaffected the ability to rec-
ognize the extracellular action potentials.
4.2.2.4. Aborting the surgery or implanting the DBS lead by use of
macrostimulation to guide location should reduction or noise and
identification of extracellular action potentials prove impossible.
4.3. The waxing and waning of the extracellular action potential
in time with the patient’s pulse, which are suggestive of cardioballis-
tic effects related to movement of the brain relative to the electrode
with each heartbeat, require that one to take the following actions:
4.3.1. Ensuring the seal of the burr hole through which the electrode enters
the brain in order to prevent a continuous pressure gradient between
the intracranial contents and the environment.
4.3.2. Advancing the electrode slowly over a small distance in order to

place its tip closer to the neurons and thereby minimize the car-
dioballistic effect.
5. Bearing in mind that patient speech or another specific mechanical event in the
operating room may produce artifacts.
5.1. Attempting to identify any operating-room activity temporally linked to
the artifact and to reproduce the artifact by reproducing the activity.
5.1.1. Requesting that the patient refrain from speaking. (The vibrations
created by patient’s speech propagate to electrical connections—the
preamplifier connector to the electrode, for example—through the
frame to the electrode.)
5.1.2. Checking for noise originating from the motor driving the electrode.
(As the electrode advances, artifact occurs. Worn mechanism gears
also generate vibrations. These problems may require servicing by
the manufacturer. In some cases one can continue neurophysiologi-
cal monitoring by advancing the electrode in short increments and
pausing to record between them. One may also manually advance
the drive should the drive motor cause significant problems.)
5.1.3. Securing, by use of sterile tape or some other implement, any swaying
wires to some structure, such as the frame affixed to the patient’s head.
5.1.4. Requesting that the patient refrain from moving. (Bodily motion
may generate noise.)
5.1.4.1. Compensating for movements of patients who cannot
remain still with a number of measures. (Patients with
tremor owing to Parkinson’s disease or Essential tremor, for
example, may minimize their tremor by holding their limbs
in a certain position. Patients with Parkinson’s disease can
also reduce their tremor by moving the affected limb in a
slow, continuous manner during the recording session.)
5.1.4.2. Increasing the cutoff high-pass-filter frequency in order to
reduce artifact. (Tremor and similar movements are typi-
cally low frequency.)
5.1.5. Restricting all physical movement in the operating room to an abso-
lute minimum.
5.1.6. Listening for noise associated with sound originating in the audio
monitor and taking the following actions should any such noise
be detected. (The sound creates vibrations in the connectors, par-

ticularly connectors between electrodes and preamplifier. The noise
from the audio monitor increases when it reaches the speaker, caus-
ing a positive feedback problem.)
5.1.6.1. Inspecting the connectors for corrosion.

5.1.6.2. Disconnecting and reconnecting the connectors.
5.1.6.3. Directing the speakers away from the electrodes.
5.1.6.4. Using headphones in order to bypass the loudspeakers.
5.1.7. Determining whether the noise appears to be correlated with any
observable physical activity and taking the following actions if it does
indeed appear to be so. Mechanical effects caused by physical activ-
ity and productive of artifact may be present but unobserved. Such
effects, which are often extremely small, may occur in the amplifier
system or other unobservable points of connection.
5.1.7.1. Removing the preamplifier or main amplifier systems in
such instances where they prove the source of artifact, as
suggested by the fact that their inputs are grounded.
5.1.7.1.1. Aborting implantation of DBS lead or surgery in
favor of macrostimulation in such instances where
persistent artifact interferes with monitoring.
5.1.7.2. Grounding preamplifiers to determine whether the mechani-
cal effect originates in the electrode apparatus or its con-
nection to the preamplifiers. (Artifact that disappears upon
grounding preamplifiers suggests a poor electrical connection
resulting from intermittent breaks in continuity is to blame.)
5.1.7.2.1. Inspecting the connectors for corrosion.
5.1.7.2.2. Disconnecting and reconnecting the connectors.
5.1.7.2.3. Replacing the electrodes.
5.1.7.2.4. Aborting implantation of DBS lead or surgery

in favor of macrostimulation in such instances
where persistent noise prohibits identification of
extracellular action potentials.
SUMMARY
Artifact and particularly noise can be ubiquitous and compromise electrophysiologi-

cal recordings. Often, reducing or eliminating noise and artifact can be difficult, but
knowledge of the underlying mechanisms greatly facilitates troubleshooting. Noise and
artifact can be appreciated as the unintended movement of electrons within the record-
ing systems. Forces that move electrons include electrostatic, magnetic, and electro-
magnetic fields, and noise and artifact are the consequence of unintended actions of
these fields. Understanding the nature of these fields leads to an understanding what
kind of artifact and noise would be generated. Knowing what type of field is playing a
role and how that particular field would be generated allows one to identify the poten-
tial sources of the artifact and noise and thus take steps to prevent or minimize artifact
and noise.
/ / / 7 / / / MICROELECTRODE
RECORDINGS
Neuronal Characteristics and
Behavioral Correlations
INTRODUCTION
This chapter presents general and practical approaches to microelectrode and

semi-microelectrode recordings before proceeding to a discussion of the latters’ role
in specific targets. The material in this chapter applies to specific DBS targets in clini-
cal use, as well as those that might be in use in the future. Of a general nature, the
approaches described should not be understood as rigidly applying to any microelec-
trode and semi-microelectrode recordings made in an individual patient by any partic-
ular neurophysiologist. Sole responsibility for proper procedure falls to the physicians
and healthcare professionals caring for the patient; they alone must exercise judgment
in treating their patients.
GUIDING PRINCIPLES
Proper microelectrode and semi-microelectrode recordings consist of two method-

ological elements: (1) correct identification of the anatomy of the microelectrode or
semi-microelectrode trajectory in order to determine optimal DBS lead placement or,
failing this, the direction and distance for a second placement; and (2) optimal effi-
ciency in microelectrode and semi-microelectrode recordings for the sake of minimiz-
ing operating-room time and reducing risk to the patient. The author has distilled the
approaches discussed below from experience. On such occasions as it is practical to do
so, he explains his approach.
125
Critical to interpretation of the anatomical localization of the recording site are

seven specific parameters for neuronal activities: (1) neuronal discharge frequency,
(2) duration of neuronal activity, (3) neuronal discharge patterns’ characteristics and
degrees of regularity, (4) neuronal density (the number of neurons occupying a record-
ing site), (5) neuronal recording site density within a trajectory (the number of sites
of detectable neuronal activity along a trajectory), (6) behavioral correlations, and
(7) unique characteristics of specific nuclei.
NEURONAL DISCHARGE FREQUENCY
A neurophysiologist expedites the microelectrode recordings by making qualita-

tive assessments of the parameters described above, which she bases on viewing the
microelectrode recordings on the monitor or listening to the audio monitor. She
judges the frequency of neuronal activity by estimating the number of extracellular
action potentials that occur during 500 ms of recording. Fewer than five extracel-
lular action potentials occurring over such duration describes a frequency of 10 Hz
or lower (Figure 7.1). (Frequency description does not involve the estimated fre-
quency of an individual recorded neuron; such estimation presents considerable dif-
ficulty for reasons discussed below.) Observation of five or more extracellular action
potentials occurring during a span of 500 ms and experiencing distinct, sustained
pauses between them describes a moderate frequency (Figure 7.2). A nearly continu-
ous train of extracellular action potentials experiencing few pauses describes a high
f requency (Figure 7.3).
0.05 s
FIGURE 7.1 Representative example of a low-frequency, low-density microelectrode record-

ing site. During the 500-ms span there occur only two extracellular action potentials of equal
amplitude.
7. Microelectrode Recordings / / 127
0.5 s
FIGURE 7.2 Representative example of a moderate-frequency, moderate-density microelec-

trode recording site. The example depicts a discontinuous activity of at least three sets of
extracellular action potentials of different amplitudes, which are indicated by the downward
arrows with different fill patterns.
NEURONAL DENSITY WITHIN A SINGLE MICROELECTRODE

RECORDING SITE
Neuronal density within a single microelectrode recording site is based on the number
of neurons occupying it, which is estimated by the variation in the amplitude of the
extracellular action potentials. The presence of only one or two different amplitudes
indicates low density (Figure 7.1). The presence of two to five different amplitudes indi-
cates moderate density (Figure 7.2). The presence of five or more different amplitudes
indicates high density (Figure 7.3).
In order to ascertain sustained activity, one must observe microelectrode recordings
for a reasonable period of time, typically at least 30 seconds. Figure 7.4 shows represen-
tative examples of transient neuronal activities. Transient neuronal activity, which is a
primary feature of the neurons at the ventral tier of the thalamic nucleus, anterior thal-
amus, and other recording sites during intraoperative neurophysiological monitoring
FIGURE 7.3 Representative examples of a high-frequency, high-density microelectrode

recording site. Shown are at least six sets of nearly continuous extracellular action potentials
of different amplitudes.
FIGURE 7.4 Representative example of transient neuronal activities. Neuronal activity begins

abruptly and fades.
for thalamic and subthalamic nucleus DBS or the putamen in globus pallidus interna
DBS, is the onset of neuronal extracellular action potentials that fire repetitively then
reduce or stop. Often this transient activity is due to “irritation” of the neuron, perhaps
by disturbing the neuronal membrane. The normal transient activities must be distin-
guished from other situations where there is a recording and then subsequent failure of
neuronal recordings. For example, as the microelectrode moves through the brain, the
resulting friction pulls the brain along with it until the latter achieves critical tension
and recoils. If this critical tension is achieved at the point at which the neurophysiolo-
gist discovers neuronal activity, the activity may be lost upon recoil, leaving one with a
false impression of transient neuronal activity. Lasting mere tens of milliseconds, this
recoil is itself transient. Primary transient neuronal activity occurs over a longer time.
A source of transient neuronal activity, an “injury discharge” may owe to a slight
tear in the neuronal membrane or some other mechanical cause. A tear allows sodium
(Na+) ions to flow into the neuronal membrane and precipitate depolarization and
action potential generation. The frequency of the action potentials diminishes as the
tear seals. There may occur a characteristically brief burst of rapid neuronal activities
which causes high-frequency discharges that progressively decrease in amplitude and
strike the ear as a diminishing buzz (Figure 7.5).
NEURONAL REGULARITY
Neuronal regularity avails itself to visual and aural assessment. Indeed, the ear enjoys
greater power of discrimination than does the eye. A wide variety of regularities—the
monotonous regularity of neuronal activities within the substantia nigra pars reticu-
lata (Figure 7.6), for example, or the high-frequency regularity of an injured neuron—
produce recurrent patterns detectable by ear.
A 1 B
0.6 s
C 1
D
1
2 3
FIGURE 7.5 Examples of a rapid transient neuronal activity whose decreasing amplitude likely
owes to an influx of Na+ ions and subsequent depolarization as a consequence of a tear in the
neuronal membrane. Example A depicts a decreasing discharge (Arrow 2). The initial spike
(Arrow 1) likely indicates the moment at which the tear occurs. From this follows a rapid flux in
ionic currents and repetitive action potentials as a consequence of subsequent neuron depo-
larization. Example B depicts a progressively increasing repetitive discharge likely initiated
by a small tear in the neuronal membrane. The influx of Na+ ions introduces water to the neu-
ron, causing it to swell. As it swells, the tear grows larger, permitting more Na+ and water to
enter. The increased influx causes further depolarization and an increased rate of discharge.
Discharge ceases upon membrane lysis. C depicts another example. The downward arrows
(1) indicate mechanical artifact as the microelectrode advances and subsequent chaffing as
it settles in position. Arrow 2 indicates microdrive artifact. Mechanical irritation engenders
a sequence of extracellular action potentials that dissipate over time. One risks mistaking
mechanical artifact associated with Arrow 1 for an extracellular action potential. In order to
provide supportive evidence of artifact, the expanded view includes more oscillations than
one would observe in an extracellular action potential. Example D depicts characteristic
high-frequency repetitive discharge of decreasing amplitudes indicative of an injury current.
NEURONAL RECORDING SITE DENSITY WITHIN A TRAJECTORY
The neuronal recording site density—the number of sites of discernible neuronal activ-
ity along a trajectory—is an important parameter. One may gain an appreciation of
the density of given sites well before she arrives at the end of the trajectory. Rather, she
may determine the relative distance between one site exhibiting discernible neuronal
activities by applying the same reasoning behind the principle that the period, p, of a
Regular
Irregular
FIGURE 7.6 Respective representative examples of regular and irregular extracellular action

potentials from microelectrode recordings.
sine wave (time between homologous points on the wave train) or some other signal is
the inverse of the frequency, f, or p = 1/f.
Determination of the neuronal recording site density depends critically on the sam-
pling rate, that is, how often or after what distance a microelectrode recording is made
as the electrode passes through the trajectory. In cases in which the microelectrode is
advanced in discrete increments (increments short enough to permit recording), one
must sample frequently, lest she miss a site of discernible neurons and thus gain a false
impression of trajectory density of recording sites. Too short an increment fails to cre-
ate enough distance between the microelectrode and the previously recorded site. One
continues in this case to record many of the same neurons, and the resulting estimate
will therefore be greater than the actual neuronal density. Regarding the dendritic tree,
which is the source of typically recorded extracellular action potentials, a diameter of
approximately 250–300 μm requires an increment of approximately 250–300 μm in
length. Again, one must permit sufficient time to elapse at each recording site in order
to avoid overlooking low-frequency neuronal activity. Specifically, one must pause at
least 15 to 30 seconds at each site. Even when recording “on the fly,” as the microelec-
trode proceeds along the trajectory, one must remain mindful of the rate of descent,
that is, the distance covered between pauses.
The sensorimotor region of the subthalamic nucleus and other high-density struc-
tures may exhibit nearly continuous neuronal activity. In such cases, one must advance
an electrode to the point at which she detects a change in extracellular action poten-
tial, either in frequency or density, or advance at any incremental distance of at least
250 μm. Otherwise, she risks making two recordings of a single set of neurons.
Certain situations exist in which a slow descent is not needed. For example, an ini-
tial trajectory encounters the subthalamic nucleus and at that depth offers adequate
understanding of the regional anatomy. In such an instance, one may advance subse-
quent electrode trajectories expeditiously to a location situated immediately above
the previously identified subthalamic nucleus. Upon arriving at this location, one may
recommence recording at a slow, deliberate pace. (In-depth discussion of this issue
appears in those chapters covering specific DBS targets.)
AUTOMATED AND SOPHISTICATED ELECTROPHYSIOLOGICAL ANALYSES
It is possible to quantify precisely the aforementioned various neuronal parameters.

For example, there exist software algorithms for discriminating extracellular action
potentials attributable to individual neurons within a recording site. These allow one
to identify the precise frequency of the neuron and the number of neurons at a record-
ing site. Autocorrelograms allow one to determine the exact degree of regularity and
neuronal densities at a trajectory.
One must not fail to appreciate the complexity of developing any automated algo-
rithm to identify the specific anatomical structure. Variance—variability of neuronal
types within a single anatomical structure and between individual patients—is made
more complex by bias in favor of larger neurons and other sampling issues associated
with microelectrode recordings. One must treat such detection and categorization
algorithms as she would any diagnostic test beset by issues of positive and negative
predictive value. That is, she must decide whether they render the intraoperative neuro-
physiological monitoring more or less efficient.
NEURONAL-BEHAVIORAL CORRELATIONS
Extremely important for identifying anatomical structures and specific regions within
a single anatomical structure, neuronal activities correlated with behavior are best
detected by listening for changes in neuronal activities during performance of a behav-
ior. For example, one may listen to neuronal activity as the physiologist passively flexes
and extends the patient’s elbow for purpose of identifying sounds suggestive of change
in activity linked in time to the movement (Figure 7.7).
One may perform any number of behavioral tasks. Most DBS indications currently
address movement disorders. Active and passive joint rotations and most other behav-
ioral events, fortunately, are motor in nature. Brisk joint rotation through the full range
of motion maximizes the ability to detect a correlation. The physiologist must repeat
13 s
20 s
FIGURE 7.7 Representative examples from two microelectrode recording sites that evidence
neuronal-behavioral correlations. The downward arrows indicate phasic increase in neuronal
activity with each joint rotation, for example of the wrist.
testing until she can confidently determine behavioral correlation. One often hears
multiple neurons in the audio monitor—a few extremely large neurons associated with
distinct sounding extracellular action potentials and a few neurons lying a greater dis-
tance from the recording site. The latter produces a hiss. Either may vary in terms of
intensity of respective sound and response to behavioral activation.
At many recording sites one encounters periodic or rhythmic neuronal activities
of varying intensity. Some variation in intensity may just coincide with flexing and
extending the elbow and other behavioral tasks. This could result in a false infer-
ence that the neuron is sensorimotor driven. The likelihood of such a false inference
is made greater if the joint movement is done in a repetitive and regular manner.
Applying the behavioral task or event in an asynchronous manner therefore remains
important.
One may use joint rotations of the jaw and facial movements to identify head
homuncular representations. To this end, the examiner instructs patients to open and
close their mouths and protrude and retract the tongue. The examiner should avoid
giving verbal cues, as speaking makes aural detection of associated changes in neuronal
activity difficult. The examiner instead opens her fist in the view of the patient—a cue
to the patient to open her mouth—and closes it: a cue to the patient to close her mouth
(Figure 7.8). The examiner similarly extends and retracts her index finger as cues for
the patient to protrude or retract her tongue. These cues the examiner must give in an
asynchronous manner.
A B
C D
FIGURE 7.8 Hand gestures used to signal the patient to withdraw the tongue (A), protrude the
tongue (B), close the mouth (C) and open the mouth (D).
Another method of task-coaching involves the examiner’s modeling movements for

behavioral tasks and asking the patient to imitate them. For the patient’s lower extremi-
ties, the examiner’s upper extremity can mimic the joint rotations required of the lower
extremities. The physiologist must take care that joint rotations (active and passive)
do not produce movement artifact. This procedure is to determine whether a neuron
changes its discharge pattern with active compared to passive movements, which is
important when distinguishing recordings in the ventral oral posterior thalamus from
the ventral intermediate thalamus (see chapter 11).
In sensory testing, a patient must distinguish forms of sensory stimulation. With tha-
lamic DBS, joint receptor activations and muscle spindle activations have their specific
anatomical implications. The examiner must perform sensory stimulation with care. The
acts of grasping the hand and rotating at the wrist, for example, activate all three types
of sensations and may thus produce ambiguity and confusion. Repetitive and regular
stimulation leads to habituation, which obscures the effects of neuronal activities.
The examiner may test light-touch receptivity by briskly brushing her fingers or a
wisp of cotton over different parts of the patient’s body. The examiner must take care to
avoid depressing the skin, lest she activate deeper sensory receptors whose associated
neuronal activity changes indicate a different region of the thalamus, for example. Joint
rotations activate joint capsule receptors and muscle spindles, the latter by lengthening
or shortening the muscle around the joint. Determining causal relation to changes in
neuronal activity requires the examiner to begin with joint rotation. Should joint rota-
tion produce a correlated change in the neuronal activities, the examiner then palpates
the muscles whose tendons cross the joint. Identical or similar neuronal responses
to palpation suggest that the neuron’s change in discharge patterns are attributable
to muscle-spindle activations. The absence of identical or similar neuronal responses
indicates neuronal activity changes related to activation of the joint capsule receptors,
provided that light touch did not elicit any change in neuronal activity. During jaw rota-
tions the examiner may palpate a master muscle to distinguish muscle spindle from
joint receptor activations.
Important to globus pallidus interna DBS, identifying the optic tract is accom-
plished by microelectrode recording and stimulation, the latter of which produces
phosphenes. In cases where sufficiently fine microelectrode tips are used, neuronal
activity in the axons of the optic nerve produces a hiss. Considerations attending use
of brisk and large joint rotations in sensorimotor testing similarly apply to use of visual
stimulation of the optic nerve’s neuronal activities. In the latter instance, one may use a
photic stimulator used in EEG laboratories or a similar device.
Behavioral evoked averaged local field potentials may eventually be used to establish
neuronal-behavioral correlations. Other types of stimuli—cognitive tasks or the odd-
ball paradigm, for example—may one day see use in DBS for neuropsychiatric disorders.
UNIQUE CHARACTERISTICS WITHIN SPECIFIC NUCLEI
A number of specific features of neuronal activities carry specific implications for

localization. The globus pallidus externa, for example, often contains so-called
high-frequency-pause neurons, that is, neurons whose highly irregular firing is punctu-
ated by pauses (Figure 7.9). The structure in which these low-frequency bursting neu-
rons appear with high-frequency-pause neurons is the globus pallidus externa.
Border cells surrounding the globus pallidus interna are characterized by their mod-
erate frequency and their unusually lengthy extracellular action potentials (Figure 7.10).
The latter emit a characteristic lower pitch.
ACTION TO TAKE IN THE EVENT OF FAILURE TO ENCOUNTER NEURONAL

ACTIVITY WITHIN A FEW MILLIMETERS OF A TRAJECTORY’S
COMMENCEMENT
Failure to record neurons within the first 4 to 6 mm of a microelectrode trajectory is

not unusual and may owe to microtrauma associated with the placement of the guide
A 12 s
B 2s
C 2s
FIGURE 7.9 Representative examples of high-frequency-pause neurons indicative of the glo-

bus pallidus interna in Parkinson’s disease. (One may not observe high-frequency-pause neu-
rons in the globus pallidus interna of patients with dystonia.) A and B appear at the same
neuronal recording sites but different type scales. C is another example from a different
recording site.
Border cell
10 s
B Non-border cells
3.5 s
FIGURE 7.10 Representative examples of two microelectrode recordings sites (A and B) con-
taining border cells. Downward arrows indicate the border cells in A. Nearly all the extracel-
lular action potentials in B occur in border cells. C shows an expanded-view comparison of a
border cell’s extracellular action potential to a nonborder cell’s. Much broader in the border
cell, the extracellular action potential achieves a characteristically low-pitch sound.
cannula. One must suspect microelectrode malfunction should she proceed further
without encountering neuronal activity. She should confirm that the impedance falls
within expected operational range. Should no artifact follow her gently tapping the
microelectrode’s connecting wires, the problem owes either to grounded inputs or
amplifier failure. Injury currents indicate the presence of neurons and suggest that the
microelectrode or semi-microelectrode is properly functioning.
Of concern is the question of knowing how far to advance an electrode one sus-
pects of malfunctioning before replacing it with a properly functioning one. Should
one traverse the entire microelectrode or semi-microelectrode trajectory, subsequent
microelectrode recordings made along it may not register neuronal activity, because
there may have resulted microtrauma associated with the initial, unproductive elec-
trode penetration. A possibly faulty microelectrode or semi-microelectrode that tra-
verses the entire trajectory, moreover, fails to produce information as to the direction
one ought to move the microelectrode assembly. Microstimulation recommends itself
as a means of possibly overcoming this difficulty.
This author recommends that one replace the electrode should she encounter no
neuronal activity within the initial 10 mm of a typical trajectory beginning 25 mm
above the target. Replacing the electrode allows her to assess the microelectrode or
semi-microelectrode above the critical region of the target without injuring it. A trajec-
tory designed to traverse regions known to contain no neurons stands as an exception. By
use of a shallower approach in the sagittal and coronal planes, some surgeons direct a tra-
jectory through the posterior limb of the internal capsule into the subthalamic nucleus.
(Other issues and concerns attend this approach, and they are addressed in chapter 9.)
Any neuronal activity encountered above a “quiet region,” which is indicative of a
functioning microelectrode or semi-microelectrode, suggests also that an intracerebral
hematoma has been traversed. In the event of such an encounter, one must remove the
electrode from the guide cannula and inspect it for traces of blood. The blood may be
“wiped clean” from the electrode, thus it is important to extend the electrode tip (for
electrodes with retractable tips) to look for blood. The top of the guide cannula one must
inspect for any evidence of extravasating blood suggestive of intracerebral hematoma.
This she does while the cannula remains in the brain. In such instances, some surgeons
leave the cannula in place to allow blood to evacuate rather than enlarge the hematoma.
They then copiously irrigate the guide cannula in order to reduce the likelihood that
blood will clot in the cannula and thus block blood exiting the brain via the cannula.
Any silent region encountered along an initial trajectory and subsequently to an
encounter with neuronal activity owes to one of three possibilities: (1) electrode fail-
ure (one may ascertain this by checking the electrode impedance); (2) traversal of an
anatomical area naturally devoid of neurons, depending on the DBS target (any such
area may be the ventricle or posterior limb of the internal capsule); or (3) the electrode’s
traversal of an intracerebral hematoma. In such instances as one encounters an intra-
cerebral hematoma, it is customary that she continue recording as she withdraws the
electrode. Should she encounter neuronal activity a second time, she may rule out elec-
trode failure as the cause of the region of silence. Should one fail to encounter neuronal
activity as she withdraws the electrode, however, she should not take this to mean that
the electrode has malfunctioned, because the failed recording with withdrawal may
owe to microtrauma.
One must bear in mind when deciding whether to replace a microelectrode or
semi-microelectrode that failure to register neuronal activity may not owe to a faulty
electrode. Immediately replacing the microelectrode or semi-microelectrode prevents
one from identifying the reason for the silence and also causes any information prior to
the silence to be lost. Should she use the same electrode in a subsequent recording during
which she encounters neuronal activity, she may rule out electrode failure as the reason for
the silence encountered in the previous trajectory. Failure to demonstrate neuronal activity
during the subsequent penetration, however, indicates that the issue remains unresolved.
MICROSTIMULATION AND ELECTRODE TYPE
The two most frequently used types of microelectrode or semi-microelectrode con-

sist either of tungsten or platinum-iridium alloy. The latter better withstands micro-
stimulation than does the former. Though a platinum-iridium microelectrode and
semi-microelectrode enable one to microstimulate anywhere within the trajectory, she is
advised to check the electrode impedance after each series of microstimulation. Tungsten
microelectrodes and semi-microelectrodes possess tips whose impedance changes with
microstimulation. One is therefore advised to refrain from microstimulation until she has
completed the electrode recording in the trajectory and begun to withdraw the electrode.
There exist two approaches to the selection of stimulation parameters for micro-
stimulation. One approach mimics the frequency, pulse width, and other stimulation
parameters typically used in DBS. This mimicry is intended to determine whether stim-
ulation with DBS-like parameters at a specific site is predictive of the subsequent clinical
response to DBS. Experience has shown this author that such mimicry possesses rela-
tively poor predictive value; it cannot present the full range of symptomatic benefit one
expects from subsequent stimulation through the DBS system, because the effective
radius of microstimulation, through either the microelectrode or semi-microelectrode,
is too small to activate sufficient neuronal elements.
Microelectrode
20 µm tip exposure
0.6 mΩ impedance
90 µA
250 µm
10 µA
FIGURE 7.11 Schematic representation of showing how a threshold to phosephenes helps

estimate distance. Experience has demonstrated that when a threshold to phosephenes was
90 μamps, the optic tract would be encountered when the microelectrode was advanced
another 250 μm as evidenced by the threshold decreasing to the minimum current. In
this case, 10 μamps represents the lowest possible stimulation current available with the
device used.
Microstimulation does allow one to identify the anatomy in the vicinity of the
electrode tip, provided she uses those stimulation parameters that are most effective
at driving a behavioral response. Stimulation frequencies of 300 pps allow one to make
most advantageous use of temporal summation (see chapter 8) in order to ensure that
electrical signals transmitted through the network find expression in the behavior. By
using a wide-pulse width of 100 μs one activates as many neuronal elements as possible,
according to their chronaxie.
Experience has shown this author that microstimulation threshold on the order of
90 μamps indicates that the electrode tip has come within 250 μm of the target. It has
also shown him that, in a case of a tonic contraction threshold of 90 μamps, advancing
the microelectrode approximately 250 μm brings it into contact with the posterior limb
of the internal capsule. Stimulation of the optic tract to produce phosephenes has led to
similar findings (Figure 7.11).
MACROSTIMULATION THROUGH THE INDIFFERENT OR REFERENCE

CONTACT OF THE BIPOLAR MICROELECTRODE OR SEMI-MICROELECTRODE
Because the indifferent contact on a bipolar microelectrode or semi-microelectrode

is often large, it enables a greater volume of tissue activation. Yet, whatever its size, it
is not nearly as large as a contact in a DBS lead. Problematic, therefore, becomes any
Reference electrode
X µA
Active electrode
250 µm
minimum µA
FIGURE 7.12 Schematic representation of showing how a threshold to phosephenes helps esti-

mate distance. The threshold to phosephenes is recorded as the electrode (in this case the
indifferent or reference) descends toward the optic tract. When the threshold reaches a mini-
mum, the electrode can be assumed to reach the optic tract. The thresholds above the optic
track can be used to calibrate the distance-threshold ratio.
generalization from macrostimulation through the indifferent electrode to an expected

clinical stimulation through the DBS lead. This author recommends that one use such
macrostimulation to identify the regional anatomy. Unfortunately, inferences from
thresholds to behavioral effects to distances in the regional anatomy require a form of
“calibration” that only experience with microelectrodes makes possible (Figure 7.12).
USE OF HORIZONTAL ARRAYS
Current technology allows for multiple simultaneous microelectrode or

semi-microelectrode recordings in the horizontal plane—the two-dimensional plane
orthogonal to the long axis of the trajectory—using a horizontal array. Believing that it
saves time, some surgeons favor a five-electrode cruciform array with a channel occu-
pying the center. Such an array does not save time, however, any economy of effort
negated by the fact that the multiple electrodes in simultaneous use divide the physi-
ologist’s attention.
Use of the arrays occasions other concerns. In some commercially available sys-
tems, for example, the center-to-center distance between channels is 2 mm, the array
itself covering a rectangular area of 2.8 mm by 2.8 mm. The optimal target in some
instances may fall outside this area. Because of the reticence to reposition an array,
many surgeons select one of the five trajectories, these trajectories’ respective degrees
of optimality notwithstanding. There also attends the use of these arrays the risk of
tissue damage: The close proximity of the electrodes to each other greatly increases
the shearing force of which the array is capable. This “bed-of-nails” effect causing tis-
sue shearing one observes in postoperative MRI scans.
In cases involving a single electrode and particular DBS conditions, a neurophysiol-
ogist may find the target in one or two penetrations. One study of targeting the subtha-
lamic nucleus for DBS determined that an average 1.4 penetrations of a single electrode
were required to identify the physiologically optimally defined target (Montgomery
2012). A surgeon who routinely uses a five-electrode array thus makes on average 3.6
potentially unnecessary brain penetrations.
The future may see use of an array of electrodes arranged in a linear fashion, that
is, arranged along the long axis of a single electrode. Whatever its virtues, this innova-
tion, should it ever be realized, would not eliminate the problem of dividing the opera-
tor’s attention. Adding side-looking electrodes may not prove effective. Microtrauma
induced by the leading segments of electrode array may reduce the ability of electrodes
in the subsequent segments to detect neuronal activities.
THE ISSUE OF CONFIDENCE
In an ideal situation, a patient experiences relief from a single DBS-lead implantation

surgery. Standing in the way of this ideal’s achievement is a problem of time: Weeks
may pass before the surgeon knows if any benefits will become of her efforts.
Uncertainty about the optimal placement of the DBS leads also places in a difficult
position the physician or healthcare professional responsible for subsequent program-
ming in the clinic. A clinician may attempt literally thousands of possible combina-
tions of electrode configurations—sets of active cathodes (negative contacts) and
anodes (positive contacts)—and stimulation parameters: voltage, current, frequency,
and pulse width. A postoperative programmer confident about the optimal placement
of the DBS lead is generally inclined to believe that some electrode configuration and
stimulation parameter do indeed confer a therapeutic benefit, and thus she tends to
make a greater effort to discover them. Doubts about the placement of the DBS lead,
however, produce doubts concerning how worthwhile extended efforts of postopera-
tive programming will prove.
Various interests compete in the operating room. The surgeon finds herself under
great pressure to complete the surgery as quickly as possible, primarily in order to reduce
stress and risk to the patient. Because prolonged procedures increase operating-room
costs that may exceed fixed reimbursements, there also lurks the temptation to cut cor-
ners during the procedure. Some surgeons perform simultaneously multiple surgeries.
The claims on her attention are therefore many and urgent. One must recognize and
acknowledge these concerns, which are not theoretical but practical, and which pro-
ceed from the various interests competing in the operating room. Properly to manage
these concerns is to serve the patient’s interest.
DOCUMENTATION
Over 16 years of involvement in intraoperative neurophysiological monitoring for DBS

surgery and many more years conducting microelectrode neurophysiological research
have done nothing to diminish this author’s surprise at the new insights he continues
to gain from his experience in the operating room. Indeed, as technology advances,
intraoperative neurophysiological monitoring evolves.
Learning from experience requires data. This author therefore recommends that
one obtain detailed information from every recording site and every trajectory. To pur-
sue this purpose efficiently, this author uses forms of the kind reproduced in Appendices
D–F. Some intraoperative neurophysiologists take minimal notes—more summaries
than descriptions—perhaps because more extensive documentation they consider
onerous, unnecessary, arbitrary, and immaterial to the outcome. The value of clear and
detailed documentation recommends itself when considered from the broader perspec-
tive of learning with a view to improving future procedures.
SUMMARY
This chapter reviews the general issues and concerns surrounding microelectrode and
semi-microelectrode recordings in order to prepare the ground for discussions of intra-
operative neurophysiological monitoring—microelectrode and semi-microelectrode
recordings specific to select DBS targets, specifically. Of particular emphasis were
these issues and concerns’ underlying rationale. To a certain degree, intraoperative
neurophysiological monitoring is an art, but it is one rooted in reason and fact.
REFERENCE
Montgomery EB, Jr.: Microelectrode targeting of the subthalamic nucleus for Deep Brain Stimulation
surgery. Movement Disorders 27(11): 1387–1391, 2012.
/ / / 8 / / / MICROSTIMULATION AND
MACROSTIMULATION
INTRODUCTION
Macrostimulation through the Deep Brain Stimulation (DBS) lead is done in virtu-
ally every implantation surgery. Others surgeries, which involve microelectrode or
semi-microelectrode recordings, often employ microstimulation through the electrode
tip. In those surgeries that use bipolar microelectrodes, semi-macrostimulation occurs
through the indifferent or reference contact (Figure 8.1).
There are at least two distinct purposes to test stimulation during intraoperative
neurophysiological monitoring. One is an attempt to replicate the physiological effects
of postoperative therapeutic stimulation so that one can prognosticate regarding clin-
ical benefit. The other is to help identify the regional anatomy around the electrode
trajectories to determine whether the trajectory is optimal for implantation of the per-
manent DBS lead and, if not optimal, then in which direction to search for a more opti-
mal trajectory. Both purposes require a basic understanding of electrical stimulation of
the brain in order to interpret the results of this procedure, troubleshoot any technical
problems that might arise, and avoid injury to the patient’s brain. Many of the prin-
ciples of biophysics, electricity, and electronics applicable to recording neuronal activi-
ties, which were reviewed in chapter 3, also apply to electrical stimulation of the brain.
Though the therapeutic mechanisms remain unknown, it is becoming increasingly
clear that excitation is the basic neuronal response to DBS. Different components of
the neuron—axon terminals, axons, cell bodies (soma), and dendrites—respond to
stimulation with a threshold to excitation peculiar to them (Figure 8.2). The response
of a given network of neurons depends on the component involved. Electrical stimula-
tion of the synaptic terminals, for example, may cause a release of neurotransmitters
locally. The majority of neurotransmitters are inhibitory, causing neuronal membrane
142
8. Microstimulation and Macrostimulation / / 143
Bipolar microelectrode
Microelectrode (244 mm)
Reference electrode
connection
Reference electrode
Active electrode
Active electrode
connection
FIGURE 8.1 Schematic representation of a bipolar microelectrode used for recording and stim-
ulation. Microstimulation occurs when passing electrical current through the microelectrode
tip. A form of macrostimulation uses the reference or indifferent electrode.
A
G H
F
E
K
D
I
C B
J
Synapse
Direct connection
Indirect connection
FIGURE 8.2 Schematic representation of potential mechanisms underlying the effects of elec-

trical stimulation, delivered through electrode (B), of a neuron recorded with electrode A. The
electrical stimulation generates an axon potential in the axon located near the stimulating
electrode. The potential’s effect may propagate to the subsequent neuron. This action poten-
tial progresses in an orthodromic manner (C), its subsequent effects related to the time of
neurotransmitter-receptor actions of the synaptic terminal. This effect on the neuron C propa-
gates to neuron J and may ultimately affect neuron E, which is recorded, through a direct or
indirect route (I). The action potential also ascends in an antidromic manner (D) to activate the
neuron near the recording electrode (E) to cause a recordable extracellular action potential.
The action potential moving antidromically (D) can invade an axon collateral (F) that proceeds
in an orthodromic fashion to activate neuron K. This sequence may directly (G) or indirectly
(H) affect neuron E.
hyperpolarization and initially a lower probability of neuronal action potential dis-

charge. For this reason one observes an initial reduction in local neuronal activity.
Many neurons in the basal ganglia–thalamic-cortical system exhibit postinhibitory
rebound excitation and respond to the putative inhibitory neurotransmitter with
rebound excitation in such a way that delayed excitation comes to be perceived as initial
inhibition. Excitation of synaptic terminals may coincide with antidromic activation,
which ascends the axon in a direction that is the reverse of its usual, to cause depolar-
ization of axon collaterals and also extend into the cell body and dendrites. The direct
effects of the stimulation pulse percolate throughout the network of oscillators to cause
repetitive activations over time. Astroglia, arterioles, and other brain cells may respond
to electrical stimulation. The mechanisms responsible for DBS’s therapeutic effect
remain unknown.
BIOPHYSICS OF ELECTRICAL ACTIVATION
Generation of an action potential by a neuronal component requires the neuronal mem-

brane’s depolarization, which causes sodium (Na+) conductance channels to open. This
opening leads to further depolarization (see c hapter 4). In the case of action potentials
generated in response to synaptic input, the neurotransmitter released from the pre-
synaptic terminal adheres to the postsynaptic receptor. Several mechanisms are set in
motion, the results of which are the opening of Na+ conductance channels and con-
sequent depolarization of the neuron in the case of excitatory neurotransmitters. Any
depolarization, for example by summing postsynaptic graded potentials, exceeding a
threshold at the axon hillock (also known as the action potential initiating segment)
causes a further increase in Na+ conductance. From this follows a series of actions that
culminates in an action potential. During stimulation, electrical charges in the form of
ions, from the electrical fields generated by the stimulating electrode tip, depolarize the
neuronal membrane’s previous charge to a point where an action potential is generated
(Figures 8.3 and 8.4).
Unlike pharmacological treatments that mimic or block neurotransmitter function
to either generate postsynaptic graded potentials or prevent them, the depolarization
associated with electrical stimulation typically generates an action potential. The result
is not at all analogous to the effects that might be produced by the typical release of
neurotransmitters, which is why DBS is not analogous to pharmacological therapies.
For example, pharmacological therapies typically do not generate antidromic action
potentials, have effects that propagate readily throughout the neural networks, or
operate on the time scales at which DBS operates.
A1 A2 B1 B2 B3
FIGURE 8.3 Electrical stimulation of the brain requires conversion of electrons that accumu-
late on the tip of the electrode into ions that carry electrical charges necessary to excite neu-
ronal elements, such as the axon. One method involves ions that are already present in the
brain interstitial fluid (A1). Under the electrostatic field, the positive ions migrate toward the
electrode tip while the negative ions are repelled. These negative ions accumulate locally
and reduce the neuronal membrane potential of neuronal elements in the vicinity of the elec-
trode (A2). Another mechanism involves the formation of ions through oxidative and reduc-
tive (Redox) reactions. These ions, once formed (B2), migrate as described above to cause
depolarization of the neuronal elements (B3).
Na+ Na+ Na+ – Na+

– –
Na+ Na+ – +
Na+ Na+ Na+ Na
– Na+
K+ K+ Na+
–
K+ K+
+ – + –
Segment of Segment of
neuronal neuronal
axon axon
Outside Outside
Inside Inside
FIGURE 8.4 Schematic representation of electrical stimulation used to produce an action

potential. The neuronal element is initially polarized, its negative voltage relatively greater
inside the neuronal element than outside it. Applied to the stimulating electrode, the nega-
tive voltage produces Redox reactions that generate negatively charged ions. Electrostatic
charges accumulating on the electrode-brain interface cause these ions resulting from the
Redox reactions and other ions already present to migrate to the neuronal element. This
migration in turn reduces the relative positive charge on the neuronal element’s external sur-
face. From this reduction follows the neuronal element’s depolarization. Sufficient depolariza-
tion opens the Na+ conductance channels, through which flows current to produce the action
potential (chapter 4). Source: Reproduced with permission from Montgomery (2010).
Electrode conductors are typically metal. Some metals undergo Redox reactions,
which deposit metal ions in the brain. Because ions are toxic, physicians avoid using
stainless steel stimulating electrodes on humans. (For the same reason, however, sci-
entists use stainless steel electrodes on laboratory animals, in whose death the exper-
iment ends, because deposits of iron ions aid scientists in localizing the stimulation
location in histological specimens). Stimulation electrodes typically used in intraop-
erative neurophysiological monitoring include platinum-iridium alloys and tungsten.
Platinum-iridium electrodes have a lesser tendency to degrade with repeated elec-
trical stimulation, as the Redox potentials do not produce platinum or iridium ions.
Platinum-iridium electrodes have thus become the standard for procedures that require
one to switch between recordings and stimulations. Because tungsten tends to erode
with stimulation, electrodes made of this metal may change impedances and there-
fore do not recommend themselves for procedures that require one to switch between
recordings and stimulations.
BIOSAFETY
Redox reactions occurring during stimulation may give rise to toxic species. Electrical
stimulation, for example, may convert water (H 20) to hydroxl ions HO-. By revers-
ing the electrical current one may also reverse most Redox reactions. Deep Brain
Stimulation pulses consist of a negative current and a subsequent positive current (by
positive is meant that the electrode accepts electrons). The negative current thus engen-
ders Redox reactions in one direction—H 2O to HO- and H+ for example—while the
positive current transforms HO- (plus H+) to H 2O, the latter neutralizing the reactive
species generated.
Full reversal of Redox reactions requires that as much electrical current enters
(during the negative or cathodal phase of the stimulation pulse) as exits (during the
positive or anodal phase of the stimulation pulse) the brain. Biphasic stimulation pulses
supply the means of establishing and maintaining this equality. Less current exiting
than entering suggests an incomplete reversal of the Redox potential. In such instances
residual toxic reaction products damage brain tissue.
To ensure that as much current exits as enters the brain, the two phases of the stim-
ulation must be balanced. The area under the curve, which is the amount of electri-
cal charge delivered in coulombs (the current multiplied by the duration of the pulse),
must be equal for the two phases (Figure 8.5). The actual waveform of the stimulation
pulse need not be the exact same shape for the cathodal and anodal phase, but for both
phases the area under the curve must be the same.
Cathodal (negative current) Anodal (negative current)
A B C
A* B* C*
FIGURE 8.5 Schematic representation of charge balance biphasic waveforms, A, B, and C. Each

component delivers the same charge. The dark gray squares (and triangle) share the same
charge but differ in the amount of current (height above or below the zero-current horizontal
line) and duration (length). The product of the current and duration, however, is the same for
all waveforms, of which A*, B*, and C* are constructions. In A*, the second phase duplicates
the initial, or cathodal, phase. The second phase is displaced below the zero-current horizon-
tal line. In B*, the displaced second phase also is rotated 90 o. In C*, two triangles represent
division of the anodal phase in B. These triangles are rearranged in time.
Some systems for intraoperative stimulation through microelectrodes may eschew

biphasic pulses in favor of a monophasic cathodal pulse (negative current). This latter
system typically delivers an amount of electric current that is extremely small, mere
microamps in current. The volume of tissue activated by it is likewise extremely small,
a mere 250 µm in diameter. Any tissue injury associated with brief pulses is therefore
likely to be exceedingly small. (Macrostimulation, which often delivers milliamps of
current, presents the opposite case.) Brain tissues, moreover, reduce toxic free radical
formation by certain mechanisms inherent to them.
The brain may suffer other injuries from electrical stimulation. Brain tissue offers
resistance to electric current passing through it. This resistance generates heat. Current
may also electrolyze water into hydrogen and oxygen to form gas bubbles that damage
brain tissue. Stimulation may excite muscle tissue of the arteries and arterioles and thus
affect circulation, or it may adversely affect glial tissue that supports the brain’s neurons.
Current density—the amount of electrical charges divided by the surface area—
determines the generally accepted safety limit for electrical stimulation. A typical max-
imal current density is 30 microcoulombs/cm2/phase, where cm2 is the surface area of
the electrical contact and phase (cathodal or anodal) is calculated by multiplying the
current (or the voltage divided by the impedance when voltage is the control param-
eter) and pulse duration or width. Coulombs are the standard measure for the total
charge introduced to the brain.
PURPOSE OF MICROSTIMULATION AND MACROSTIMULATION
Microstimulation and macrostimulation within intraoperative neurophysiological

monitoring serves two purposes: (1) they allow one to predict subsequent clinical
response to DBS in terms of therapeutic benefit and obstacles (potential side effects)
and (2) they enable one to identify regional anatomy.
Use of stimulation to prognosticate subsequently therapeutic results must
assume adequate simulation of the DBS effect. Typically, recreating the stan-
dard stimulation frequency and pulse width parameters is the default. However,
the critical issue is the spatial distribution of the current density, which deter-
mines the volume of tissue activated. The spatial distribution of the current den-
sity depends on the electrode size and geometry. Any attempt to replicate the DBS
effect with microstimulation will likely fail, because, as this author has learned from
his own experience, the small volume of tissue activated by a microelectrode or
semi-microelectrode tip militates against an accurate prediction of the clinical DBS
effect. Intraoperative macrostimulation through the DBS lead should enable one
to predict clinical response. However, even this can be problematic, as the operat-
ing room rarely affords sufficient time to test all relevant combinations of electrode
configurations and stimulation parameters. Further complicating the prognostica-
tion are differences and changes over time in the impedance when constant voltage
stimulation is used compared to constant current. (The importance of this distinc-
tion is discussed below.)
This author identifies the regional anatomy near the microelectrode site primar-
ily by use of microstimulation, which complements microelectrode recordings made
at the site. The threshold to an effect offers one a rough indication of the distance
between the electrode tip and the structure that produces the effects following stim-
ulation. An electrode whose tip exposure is approximately 20 microns and whose
impedance is 0.6 mOhms (mΩ) stimulates a muscle contraction at 90 microamps
when the electrode tip is approximately 250 µm. This is evidenced by the experience
of subsequently advancing the electrode and finding it detects a paucity of extracel-
lular action potentials, which indicates that it has penetrated the posterior limb of the
internal capsule.
The ability to switch between recording and stimulation at any point in the micro-
electrode trajectory facilitates comparison of the microelectrode recordings with
the effects of microstimulation. The advantage in this concern falls on the side of
platinum-iridium electrodes. Microstimulation proves particularly useful at sites of
low neuronal density and few extracellular action potentials. For example, microstimu-
lation of a muscular contraction in the absence of extracellular action potentials sug-
gests that the electrode has entered the posterior limb of the internal capsule. Similarly,
production of phosphenes in a region absent of extracellular action potentials suggests
that the electrode has entered the vicinity of the optic tract. One notes that in both
examples the absence of response to microstimulation remains inconclusive; it may
indicate a brain structure incapable of mediating a response to stimulation, or it may
indicate an insufficient volume of tissue activation.
The ventral intermediate nucleus of the thalamus presents a special case in terms of
stimulation performed for the purpose of identifying the homuncular representation at
the stimulation site. Paresthesias produced by microstimulation may not collocate with
the homuncular representation, because the former is related to stimulation of fibers
passaging through the local homuncular representation in transit to another represen-
tation elsewhere (Figure 8.6).
Leg
Amp
Concordant
Arm
Leg
Amp
Arm
Discordant
FIGURE 8.6 Possible mechanism of discordant and concordant paresthesias in response to

stimulation. In the case of concordant responses, microelectrodes record activity changes
correlated with movement of the arm but not the leg. Microstimulation (represented by the
lightning bolts) at the same site activates local neurons (white cartoons) whose excitation
manifests as paresthesias of the arm. In the discordant response, the microelectrode con-
tinues to record changes in neuronal activity with movement of the arm but not the leg.
Microstimulation activates axons as the electrode passes them on its way to neurons of the
leg representation. Stimulation of these axons causes paresthesias referred to the leg.
NEURONAL SYSTEMS ACTIVATION
The effects of stimulation must ultimately reach the effector organ in order to provoke
a behavioral response. Stimulation of the basal ganglia–thalamic-cortical system, for
example, must ultimately reach and affect the muscles that mediate the behavior. In the
case of subthalamic nucleus DBS, the minimum pathway may lie in two activations set
in motion: (1) antidromic activation of motor cortical neurons that project to the sub-
thalamic nucleus and (2) orthodromic activation through axon collaterals of the anti-
dromically activated motor cortex neurons that descend to the lower motor neuron in
the spinal cord and brain stem. Activation of the lower motor neurons in the spinal cord
and brain stem modulate activities of the motor units (the muscle fibers innervated by a
single lower motor neuron) to produce the behavior. The number of neuronal elements
activated likely determines the degree of the behavior affected. A sufficient degree of acti-
vation of neuronal elements at the site must be achieved in order to activate the neuronal
elements directly responsible for the behavior. In order to produce a motor effect, for
example, the sufficient stimulation of motor units at the electrode site must be achieved.
The above description of the minimum pathway is not to suggest that this path-
way mediates the therapeutic effect of DBS. Indeed, there are many other path-
ways that connect the site of stimulation with the lower motor neurons, probably
via the motor cortex. These are far more than the direct, indirect, and hyperdirect
pathways typically described for the basal ganglia–thalamic-cortical system (Figure
8.7). Conceptualization of the basal ganglia–thalamic-cortical system as networks
of recurrent neuronal oscillators of different length and consequently different fun-
damental frequencies provides a cogent theory of how motor units are controlled to
produce behaviors. There are multiple levels of orchestration of motor unit activi-
ties from the recurrent order of motor units to dynamic synergistic coordination of
different muscle groups, and each may be associated with specific sets of oscillators
within the basal ganglia-thalamic-cortical system (Montgomery 2013). The theory
provides an explanation of the effects of different DBS frequencies and the orchestra-
tion of motor unit activities.
Activation of motor units must be mediated by synapses at each point in the pathway
between the stimulation site and the effector site. Because individual synapses are highly
inefficient, this need for mediation presents a problem. Less than 10% of postsynaptic
potentials generate subsequent action potentials (cerebellar climbing fibers and Purkinje
cells stand as notable exceptions in this regard). Neuronal activations provoke a sufficient
response in at least two general ways: (1) spatial summation and (2) temporal summation.
Postsynaptic potentials (depolarization potentials are considered here) spread
over the neuronal membrane from the synaptic site and grow smaller in amplitude
A B C
MC PT MC PT MC PT
VL GPi GPe VL GPi GPe VL GPi GPe
STN STN STN
D E F
MC PT MC PT MC 5 PT
1
9 2
4 6
10
3
VL GPi GPe VL GPi GPe
VL GPi GPe
STN STN
STN 7
FIGURE 8.7 Schematic of a subset of nuclei in the basal ganglia and the ventrolateral thala-
mus (VL) and motor cortex (MC) that make up the basal ganglia–thalamic-cortical system.
The basal ganglia structures include the putamen (PT), globus pallidus externa (GPe), glo-
bus pallidus interna (GPi), and subthalamic nucleus (STN). Figure A shows just some of the
potential oscillators. Figure B shows what is referred to as the direct pathway. Figure C shows
the hyperdirect pathway, and D shows the indirect pathway. Also shown is the VL-MC oscil-
lator (E). Figure F shows another potential oscillator that is much longer, involving at least
two “passes” through the basal ganglia-thalamic-cortical system. Source: Reproduced from
Montgomery (2013).
as they spread in a manner closely similar to the action of a ripple caused by a pebble
tossed in water. Should two or more synaptic events occur at the same time and in
close enough proximity to each other, their respective depolarization may combine
and thereby generate an action potential in the postsynaptic neuron (Figure 8.8). In
many cases, a degree of spatial summation is assured by the fact that a given axon may
branch hundreds of times to produce many synapses, which, because they are driven
by the same orthodromic action potential, generate potentials nearly simultaneously.
Deep Brain Stimulation lead contacts and other large electrodes may simultaneously
generate an extremely large number of axonal activations and thus magnify spatial
summation.
Temporal summation offers another mechanism for amplifying the response
(Figure 8.9). Postsynaptic potentials decay over time. Should a second postsynaptic
potential occur before the potential preceding it sufficiently decays, there may occur a
summing of the two postsynaptic potentials to increase the probability of driving the
postsynaptic neuron and ultimately producing a behavioral effect. High-frequency
stimulation is capable of activations occurring in sufficiently rapid sequence to
increase the probability of a behavioral event. Experience with microstimulation in
the cerebellum of nonhuman primates suggests that an optimal stimulation frequency
t1 t2 t3
FIGURE 8.8 Schematic representation of spatial summation. Two simultaneous synapses pro-

voke adjacent postsynaptic potentials at time t1. Each synapse by itself cannot sufficiently
generate an action potential. The postsynaptic potentials spread over the neuronal membrane
(t 2). At time t 3, overlapping waves depolarize the neuronal membrane sufficient to generate an
action potential (indicated by the lightning bolt).
is on the order of 300 pps (Schultz, Montgomery, Marine 1979), which is certainly
well above the stimulation frequencies used in DBS, but which is usually available
for microstimulation. This author typically uses a stimulation frequency of 300 pps
and a pulse width of 100 microseconds. For monophasic stimulation he uses cathodal
or negative current. For biphasic stimulation he uses a cathodal (negative current)
initial phase.
AP 1
B AP 2
AP 3
A
+
Amp 0
AP 3 –
Threshold
PSP 1 PSP 2
FIGURE 8.9 Schematic representation of the effects of temporal summation to produce an

action potential. An intracellular microelectrode (A) in the axon hillock (or action potential
initiating segment) records the neuronal membrane potential of neuron B. Two action poten-
tials descend the presynaptic axon. The arrival of the first action potential (AP 1) provokes a
postsynaptic potential (PSP 1) that causes a depolarization, albeit depolarization insufficient
to reach the threshold for generation of an action potential in neuron B. A second action
potential (AP 2) follows close on heels of PSP 1, arriving before the first, which alone is insuf-
ficient for generating an action potential, decays too much. Added to the membrane potential
depolarized by the first postsynaptic potential, the second potential produces effects suffi-
cient to depolarize beyond the threshold and generate an action potential (AP 3) in neuron B.
CONSTANT CURRENT VERSUS CONSTANT VOLTAGE
Most stimulators are designed to supply constant voltage or constant current. As

described in c hapter 3, the electrical current is a function of voltage and impedance. As
described in chapter 4, electrical charge—the amount of current over a fixed period of
time—plays an important role in determining the neuronal response. A constant-voltage
stimulator is thus at a disadvantage in situations where the impedance changes. There
occurs a change in electrical current and consequent neuronal response should imped-
ance change but voltage remain constant. Constant-current stimulators maintain a
constant electrical current during stimulation that helps to ensure a relatively constant
delivery of electrical current (Figure 8.10). The constant-current stimulator adjusts to
any change in impedance and voltage.
Decreasing current Increasing current
Constant
Voltage
Stimulation
Increasing Decreasing
Constant current Impedance Constant current
Constant
Current
Stimulation
Increasing Decreasing
Impedance
FIGURE 8.10 Schematic representation of the effects of impedance on stimulation current

when constant voltage is used in place of constant-current stimulation. The impedance is
represented by the hill over which the automobile travels. The height of the hill represents
the impedance. The automobile’s accelerator controls the voltage. For the constant-voltage
stimulation (constant, even pressure on the accelerator), impedance determines the amount
of current. Increase in impedance is analogous to the automobile’s uphill progress. In the
constant-voltage situation, velocity (current) diminishes as the automobile travels uphill
(increased impedance) and increases as it travels downhill (reduced impedance). In the case
of constant-current stimulation, the motorist increases pressure on accelerator (increases the
voltage) in order to maintain the automobile’s speed (current) as it travels uphill. The motorist
then reduces pressure on the accelerator (reduces voltage) as the automobile travels downhill
(reduction in impedance) in order to continue at same speed (current). Source: Reproduced with
To the effects of impedance on constant voltage are added the effects of capaci-
tive reactance (see chapter 3). Formed by the electrode-tissue interface, negative
charge accumulates on the capacitor and begins to oppose electrical charge deposited
on the initial capacitor surface. This phenomenon is known as capacitive reactance.
Capacitive reactance opposes delivery of electrical current to the brain. From this
opposition there results a distorted stimulation pulse current waveform (Figure 8.11).
The area under the current waveform determines the net electrical charges introduced
to the brain. For the same stimulation pulse waveform, the constant-voltage stimula-
tor delivers lower electric charge. The constant-current stimulator accounts for the
capacitive reactances by increasing the voltage and thereby producing a current wave-
form that delivers greater electric charge to the brain.
Constant-voltage Voltage
1
A – – – – – Opposing
– – – electrostatic charge
– – –
– – – – – Current
–
– – D
B – – 2 –
– –
– –
– – – – 2
– 1
Constant-current
C – – E
–
– – – – – –
–– – – – – –
– – –
Electrode surface Brain Tissue
FIGURE 8.11 Schematic representation of the effects of capacitive reactants on the stimulus

pulse waveform. A biphasic stimulus pulse introduced to the brain describes the waveform
appearing in E. In the case of the constant-voltage stimulator, the actual current waveform (the
key determinant of a neuronal response) becomes distorted, as it appears in E. One observes
that in the constant-voltage stimulation electric charge accumulates on the electrode-tissue
interface (A) and acts as a capacitor (see chapter 3). As the charge accumulates (B), the elec-
trostatic charges begin to resist further introduction of charge. This opposition reduces cur-
rent flow through the electrode-tissue interface and distorts the current-stimulus waveform.
Constant-current stimulation, however, produces no such phenomenon. As the electrostatic
charge accumulates on the electrode-tissue interface and begins to resist further accumula-
tion, the constant-current stimulation increases the voltage to counter the effect and to assure
constant current flow into the brain. Source: Reproduced with permission from Montgomery (2010).
THE RETURN PATH
Because electricity flows from one point to another, current entering the brain at one point
through the active electrode exits the brain through another point, known as the return
path. Just as the electrical characteristics of the stimulating contact at the entry point of
electrical current into the brain is critically important, so too are the characteristics of the
electrical contact that removes electrical current from the brain. Just as the electrical safety
of the stimulation entering the brain is determined by the microcoulombs/cm2/phase
through the active electrode, the return-path electrode is subject to the same conditions.
The current source or entry point typically determines microcoulombs and phase, but the
electrode mediating the return path determines the surface area, as measured in cm2. If
the surface area of the return path is relatively small compared with the source contact, the
current density through the return path may be larger and possibly exceed safety limits.
Current introduced by monophasic microstimulation through microelectrodes
and semi-microelectrodes follows a return path other than the electrode tip, which
has an extremely small surface area. Current flowing during the positive current phase
of biphasic stimulation lies through the microelectrode or semi-microelectrode. As
discussed above, the amounts of current and extremely small volume of potentially
affected tissue will likely lack clinical significance. Clinical experience, however, has
yet to establish the validity of this likelihood.
Bipolar biphasic stimulation through the DBS lead may create a stimulation involv-
ing three cathodes that features an initial cathodal phase of the biphasic pulse in which
three contacts and a single anode supply the source current. Use of three cathodes tri-
ples the surface area, each cathode representing roughly one-third of the total current
flow. The presence of a single anode, however, reduces the surface area by two-thirds.
Current through the anode, meanwhile, triples and thus risks exceeding the safety lim-
its. In monopolar recording where the implanted pulse generator case is the anode, the
extremely large surface area relative to the source contacts greatly reduces the current
densities through the anode.
MACROSTIMULATION THROUGH THE DBS LEAD
As discussed above, macrostimulation through the DBS lead introduced in intraopera-

tive neurophysiological monitoring serves primarily to predict the clinical responses
during subsequent therapeutic use. These responses include not only various symp-
toms and signs related to the disabilities produced by the underlying diseases but also
any side effects that may limit the therapeutic use.
Many DBS centers simulate the electrode configurations and stimulation param-
eters reported in the literature to represent the most common use, often by referring to
the mean or median values. The electrode configuration relates to the patterns of active
cathodes (negative contacts) and anodes (positive contacts). The stimulation param-
eters refer to the stimulation voltage or current, pulse width, and frequency. Though
this simulation may seem reasonable, it is not advisable.
The situation involving a 130-pps median stimulation frequency presents a useful
example. In this situation, approximately 50% of patients require, by definition, stimu-
lation greater than 130 pps. Yet there is no way that an intraoperative neurophysiolo-
gist may know beforehand whether her patient belongs to this contingent. Her patient
requiring stimulation in excess of 130 pps may in fact benefit from stimulation of 160
pps. Also, it may occur that the patient will experience limiting side effects at some
stimulation rate below the beneficial one—at 145 pps, for instance. Should the intraop-
erative neurophysiologist limits DBS macrostimulation to 130 pps, she may discharge
her patient from the operating room without gaining any sense of the patient’s ability
to tolerate wider DBS parameters, which subsequent therapy might make necessary.
This eventuality the intraoperative neurophysiologist may anticipate by stimulation at
higher parameters and consequently accommodate by moving the DBS leads in such
a manner as to establish a therapeutic window, that is, a set of stimulation parameters
below bottom of the window in which stimulation fails to produce benefit and above
the top of the window in which side effects are generated.
This author’s approach is to begin DBS with a DBS frequency of 160 pps and a pulse
width of 120 microseconds. Because current commercially available devices make
monopolar DBS test stimulation difficult, this author uses a wide bipolar configuration,
which provides him with the greatest electrical charge density of the largest volume
(Montgomery 2010). Modifications of this approach to specific DBS targets receive
detailed discussion in the relevant chapters.
STIMULATION SAFETY
Discussed above were such specific aspects of stimulation safety as current density in
microcoulombs/cm2/phase, charge-balanced biphasic stimulation, and heating issues. The
discussion of safety below specifically addresses the electronics used to deliver stimulation.
The same concerns attending electrical isolation and grounding in the case of neu-
ronal activity recordings (chapter 5) apply to stimulation. Isolating the patient from AC
power lines and other sources of potentially unlimited electric current remains of para-
mount importance. Because they contain a finite amount of current, battery-powered
stimulating devices recommend themselves for use. Issuing from the device connected
to the AC power lines, stimulation pulse waveforms drive stimulation to the brain by a
second device that shares no electrical continuity with the first, and thus the patient is
isolated (see chapter 5).
The same concerns attending grounding and ground loop currents in devices
related to electrical recordings also apply to electrical stimulation (see chapter 5).
Because access to the earth ground may differ for the patient and the various devices
in use, ground loop currents may not flow to the ground. They may instead follow an
unanticipated route that lies through the patient’s body. Biomedical safety technicians
check electrical equipment used in the operating room for such ground loop or leak-
age current by comparing the amount of current going into a device to the amount of
current leaving the device through the intended path. Lower-than-expected current
exiting via the intended path suggests that some portion of this current has exited via
an unintended one.
SUMMARY
Virtually every DBS lead–implantation surgery uses some form of electrical stimu-
lation to help identify the optimal target. Stimulation in the operating room as an
approach to predicting the effectiveness of postoperative therapeutic DBS encounters
difficulty in such instances as when the electrode in use in the operating room is differ-
ent from the one that is to be used in therapy. These simulations depend on electrode
configurations and, more crucially, stimulation parameters believed typical of thera-
peutic DBS. Another approach involves optimizing electrical stimulation as a means of
defining regional anatomy around the electrode, which uses different mechanisms and
stimulation parameters.
Analogous to those principles that guide microelectrode and semi-microelectrode
recordings, the basic principles underlying electrical stimulation include those that
govern the biophysics of action potential generation. Electrical stimulation gener-
ates ions by introducing electrons. Though applying electrical potentials to brain
tissue generates potentially toxic reactive species, reversing current flow through
electrodes remedies this effect. Electrical stimulation thus requires charge-balanced
biphasic pulses.
Many of the concerns related to electrical safety for recording neuronal activities
are comparable to electrical safety issues related to electrical stimulation. Critical to
safe electrical stimulation are the isolation of the patient from sources of constant cur-
rent and any ground loop current that may pass through her.
REFERENCES
Montgomery EB, Jr.: Neurophysiology. Parkinson’s Disease, 5th ed. R. Pahwa and K.E. Lyons. Boca
Raton, CRC Press, Taylor & Francis Group, 2013: 25–280.
Schultz W, Montgomery EB, Jr., and Marine R: Proximal limb movements in response to microstimulation
of primate dentate and interpositus nuclei mediated by brainstem structures. Brain 102:127–146, 1979.
/ / / 9 / / / THE SUBTHALAMIC NUCLEUS
INTRODUCTION
An intraoperative neurophysiologist must undertake to implant the DBS lead in the

vicinity of the sensorimotor region of the subthalamic nucleus. This follows the pre-
sumption that placement within the sensorimotor region of the subthalamic nucleus
maximizes benefit and minimizes any adverse effects. Interestingly, no one specific
target exists for DBS. Other authors have described sites in the dorsal lateral region of
the subthalamic nucleus (the sensorimotor region) and the lenticular fasciculus (one
of the outflow paths from the globus pallidus interna to the thalamus), which is located
just above the subthalamic nucleus. High variance in most outcome measures—the
Unified Parkinson Disease Rating Scales (UPDRS), for example—makes it difficult to
conduct Evidence-Based Medicine level 1 clinical trials to establish the optimal target.
Such trials will therefore likely not be made any time soon. The most common target,
the dorsal lateral subthalamic nucleus, is the focus of this chapter.
The intraoperative neurophysiologist faces the challenge of using neurophysiologi-
cal properties to identify a target that predicts subsequent clinical response. Targeting
clinical efficacy is problematic in this regard, though less so than many of the potential
side effects that may be intolerable to patients or that may limit DBS efficacy. Relatively
straightforward, some adverse effects can be attributed to specific inadvertently stimu-
lated anatomical structures surrounding the target. The medial lemniscus, for example,
ascends immediately posteriorly to the subthalamic nucleus. Placing the DBS lead too
posteriorly and in proximity to the medial lemniscus risks producing paresthesias that
patients find intolerable and that limit DBS. Some adverse effects, such as depression
and impulsivity, are more difficult to relate to specific anatomy. These effects are more
likely if the DBS lead is implanted medially and ventrally in the subthalamic nucleus.
The subthalamic nucleus is relatively small, possessing a mean dimension of
1.2 × 6 × 3 mm and a long axis parallel to the rostrocaudal axis, and is flattened at
159
its periphery (Yelnik and Percheron 1979). The sensorimotor region within the sub-
thalamic nucleus is yet smaller. Unlike the ventral intermediate nucleus of the thala-
mus, the globus pallidus interna, and other larger targets, identifying specific regions
within the motor homunculus generally is not an issue.
CRITERIA FOR A PHYSIOLOGICALLY DEFINED OR SYMPTOMATICALLY

DEFINED OPTIMAL TARGET
Though it admits of some difference of opinion, general consensus holds that, accord-
ing to microelectrode or semi-microelectrode recordings and simulation, the definition
of the physiologically defined optimal target rests on the following three criteria: (1) a
sensorimotor subthalamic nucleus of at least 5 mm in size, (2) a lack of adverse effects
with microstimulation or macrostimulation, and (3) benefits that follow stimulation.
Each criterion requires explanation.
The criterion for at least 5 mm of sensorimotor subthalamic nucleus rests on ana-
tomical considerations. Because the size of the subthalamic nucleus is relatively small
and the electrodes’ typical trajectory is tangential to the long axis of the subthalamic
nucleus, moving the electrode from a trajectory at least 5 mm in length is likely to result
in shallower penetration of the subthalamic nucleus, which lessens the amount of sub-
thalamic nucleus that can be affected by DBS.
A rather straightforward requirement attends microstimulation and macrostimula-
tion: They must produce no adverse side effects. Microstimulation conveys more precise
estimates of the distance between the microstimulation site and the anatomical struc-
ture producing tonic muscle contraction or some other adverse effect. Though micro-
stimulation (and some types of macrostimulation) may produce no adverse effects, the
relatively smaller volume obviating it, there is no guarantee that DBS macrostimula-
tion will do likewise. Generally, however, if microstimulation produces adverse effects,
macrostimulation will produce them as well.
A second requirement—that DBS relieve symptoms—presents some difficulty. An
insufficient volume of tissue may be activated, particularly if the microelectrode tip,
semi-microelectrode tip, or indifferent electrical contact introduces the stimulation.
A micro-subthalamotomy effect may mask improvement with subsequent stimulation.
In those situations, the primary consideration turns on avoidance of adverse effects.
One must therefore test a range of symptoms. Muscle tone or rigidity, for example, is
the symptom that frequently shows the most improvement.
An intraoperative neurophysiologist who eschews use of microelectrodes or
semi-microelectrodes recordings or stimulation in favor of macrostimulation through
9. The Subthalamic Nucleus / / 161
the DBS leads looks to symptomatic response as a criterion for accurate placement in
addition to avoiding adverse effects that might limit subsequent therapeutic use of DBS.
Transient adverse symptoms produced by test DBS should not be considered adverse
events. A micro-subthalamotomy effect may complicate demonstration of symptom-
atic improvement. The experience of the author demonstrates that, at times, the micro-
electrode recordings are the only indication in the operating room that subsequent
therapeutic stimulation has a reasonable chance at success.
Yet another criterion during intraoperative test stimulation through the DBS lead
is the ability to demonstrate a reasonable therapeutic window—the difference between
the threshold at which symptoms improve and the threshold at which DBS macrostim-
ulation produces persistent adverse effects. This window defines the operating range
for postoperative therapeutic stimulation. A narrow window will make problematic
subsequent postoperative care.
The worst-case scenario in the context of postoperative DBS care should be antici-
pated intraoperatively. It is possible that the patient may require stimulation parameters
higher than what is typical, and consequently the highest stimulation parameters that
might reasonably occur in the postoperative setting should be tested intraoperatively.
Consider the scenario in which a patient has a micro-subthalamotomy effect preventing
intraoperative prognostication of postoperative benefit and only the most commonly
used stimulation strengths are tested, such as 130 pps, 60 μs pulse width, and 3.5 volts or
2 milliamps. In the outpatient clinic, once the micro-subthalamotomy benefit has abated,
the patient needs much higher stimulation parameters to receive benefit but cannot tol-
erate those parameters because of adverse effects. Had this been determined intraopera-
tively, the physician could have modified the DBS lead location to improve tolerance. To
this end, he uses wide bipolar configurations for intraoperative DBS macrostimulation.
Frequency for these configurations is 160 pps, and their pulse width is 120 μs.
Testing with the mean or median stimulation parameters only, as reported in the
literature, is problematic. The median and mean (assuming that the data distribution is
normal) means that half of patients treated required higher stimulation parameters. In
these patients, demonstration of tolerance at the median or mean stimulation param-
eters does not mean the patient will either benefit or tolerate higher stimulation param-
eters should they prove necessary. Raising the threshold of the median by some number
of standard deviations may prove effective, but data for this is generally unavailable.
The intraoperative neurophysiologist may choose the constant stimulation voltage
or current she wishes to use, even in cases in which she knows and has accounted for
the impedances. As a consequence of capacitive reactance, constant-voltage stimula-
tion alters the independent waveform (see chapter 8). A critical factor in stimulating
neural elements (see chapter 8), the amount of electrical charge is related to the area
under the stimulus pulse waveform. The same constant-voltage stimulation waveform
thus delivers fewer electrical charges than does a constant-current stimulator.
REGIONAL ANATOMY OF THE ELECTRODE TRAJECTORY
For all forms of intraoperative neurophysiological monitoring, an understanding of

the regional anatomy around the entire electrode trajectory is critically important. For
microelectrode and semi-microelectrode recordings, study must be made of the entire
trajectory. Typically, image-guided navigation targets the bottom of the anterior-lateral
subthalamic nucleus, and when this is confirmed by other means, such as microelec-
trode recordings, this author places the distal or ventral edge of the most distal or ven-
tral DBS lead contact at the bottom of the subthalamic nucleus. During microelectrode
and semi-microelectrode recordings, analysis of recording sites above the subthalamic
nucleus is important for identifying the regional anatomy. Recordings made well above
the subthalamic nucleus in the thalamus, for example, aid estimation of the relative
position of the trajectory in the anterior-posterior direction (discussed in detail below).
Consequently, the recording trajectories typically begin approximately 25 mm above
the bottom of the subthalamic nucleus.
Competing needs—that of obtaining sufficient information from recordings and
that of halting the guide cannula well above the target—create a tension that justi-
fies beginning the microelectrode or semi-microelectrode trajectory approximately
25 mm above the bottom of the subthalamic nucleus. A sharp microelectrode or
semi-microelectrode necessarily traverses a greater portion of brain, therefore, and the
risk of intracerebral hematoma increases accordingly. A blunt-tipped stylus within the
guide cannula, on the other hand, tends to push tissue to the side. However, placement
through the guide cannula ending 25 mm above the target forces the DBS lead to travel
25 mm. The latter thus risks being deflected from its intended course.
Stimulation introduced solely through the DBS lead permits use of a longer can-
nula. The intraoperative neurophysiologist must ensure, however, that she halts the
cannula’s progress above the most dorsal contact in the DBS lead so as not to interfere
with intraoperative DBS testing. Though a longer cannula reduces risk of deflection, it
presents problems when it comes time to withdraw it, because the surgeon must leave
sufficient clearance for easily fixing the DBS lead in place at the level of the burr hole
in the skull. Should she elect to use a microelectrode or semi-microelectrode, she may
replace a cannula that ends 25 mm above the target for one that ends immediately above
the dorsal-most contact. In doing so she mitigates the risk of deflection. This, however,
necessitates removing one cannula and replacing it with another. A brain shift occur-
ring between the removal of the first cannula and the insertion of the second introduces
the risk of misplacing the DBS lead.
Determination of regional anatomy depends on the angle of entry in the sagittal
and coronal plane. The need to avoid the lateral ventricles influences selection of the
entry point in the angle in the coronal plane (Figure 9.1). Conventional thinking holds
that traversal of the ependymal surface may deflect the DBS lead. A shallow trajec-
tory in the coronal plane, which is related to an extremely lateral entry point, does not
appear to significantly affect the subthalamic nucleus. A shallow trajectory in the coro-
nal plane in other structures may present a problem. An extremely shallow trajectory
Dorsal
Medial Lateral
Ventral
STN
Red nucleus B
Fascicles of the C
oculomotor
ul f
complex
ps o
e
ca mb
al li
Fp 4.0
rn or
te ri
in ste
Po
FIGURE 9.1 Schematic representation of the effects of the angle of trajectory in the coronal
plane 4 mm anterior to the midpoint of the line connecting the anterior and posterior commis-
sures (AC-PC line). An extremely lateral entry point to avoid the lateral ventricles may result
in an excessively shallow trajectory in the coronal plane whose upper portions may pass
through the posterior limb of the internal capsule (A). The lack of neuronal activity detected
as the electrode tips pass through the posterior limb of the internal capsule the intraoperative
neurophysiologist may confuse for intracerebral hematoma. Also, if the DBS lead follows an
excessively shallow trajectory in the coronal plane, the probability increases that the ventral
contacts will enter the medial subthalamic nucleus (B). Stimulation through these contacts
is more likely to produce cognitive and psychological complications. The ventral contacts
may also approach the fascicles of the oculomotor nerve (C), which, if stimulated, skews the
patient’s eyes and prompts her complaints of diplopia. Medial stimulation may cause ataxia
because it may affect the fibers of the brachium conjunctivum, a structure that relays neuro-
nal activity from the deep cerebellar nuclei to the thalamus passing through the red nucleus.
Source: Modified from Schaltenbrand and Wahren (1977).
could result in an electrode’s traversal of the posterior limb of the internal capsule, a
site at which microelectrode and semi-microelectrode recordings likely will not detect
neuronal activity. This lack of detected activity the intraoperative neurophysiologist
may mistake for intracerebral hematoma. Also, though the DBS lead’s ventral contacts
may enter the subthalamic nucleus, its dorsal-most electrodes may lie sufficiently close
to the posterior limb of the internal capsule to increase the risk of tonic contraction.
Because the DBS moves medially as it descends, such placement also increases the risk
of a stimulating current’s propagation to the medial subthalamic nucleus. Such prop-
agation may produce cognitive or psychological problems, skewed eye deviation as a
consequence of stimulation of the nerve fascicles of the oculomotor complex, for exam-
ple, or ataxia as a consequence of propagation to the brachium conjunctivum through
the red nucleus. Any such side effects limit the options for postoperative DBS program-
ming. This author prefers a highly verticle trajectory in the coronal plane.
The angle of entry in the sagittal plane is critically important. In most cases suffi-
ciently anterior to the coronal suture, the entry point avoids passage through the motor
cortex and thus reduces the risk of tissue damage and hemiplegia. If the entry point and
angle of trajectory are excessively shallow and anterior, however, the electrodes may
traverse the putamen and globus pallidus (Figure 9.2), particularly in such cases brain
shift may occur posteriorly as a result of gravity or intracranial air. Routine microelec-
trode or semi-microelectrode recording may not allow one to differentiate the medial
globus pallidus interna from the subthalamic nucleus. Evidence derived from sophis-
ticated analysis suggests that the globus pallidus interna’s mean discharge frequency
is lower than that of the subthalamic nucleus and that its neuronal density is greater.
Typical microelectrode recordings, however, may not differentiate the two.
Locating the substantia nigra pars reticulata helps to support the notion that neurons
immediately dorsal to the electrode site belong to the subthalamic nucleus. Yet there do
exist trajectories through the subthalamic nucleus that may not traverse the substantia
nigra pars reticulata (Figure 9.3). Recordings in the substantia nigra pars reticulata may
be absent, but one should not conclude from this absence that the DBS lead has missed the
subthalamic nucleus, particularly considering that laterally one may find the subthalamic
nucleus but pass lateral to the substantial nigra pars reticulata. The absence of demon-
strated activity in the substantia nigra pars reticulata leaves the intraoperative neurophys-
iologist unsure whether the trajectory lies through the subthalamic nucleus or the medial
globus pallidus interna. The medial globus pallidus interna may be encountered unin-
tentionally if there is marked brain shift consequent to air entering the skull and expand-
ing in volume as it warms to body temperature. Tension pneumocephalus indicative of
Dorsal
Posterior Anterior
Ventral B
D A
GPi
STN
ul f
ps b o
C
e
ca m
al r li
rn io
te r
in oste
17.0 P
FIGURE 9.2 Schematic representation in the sagittal plane (17 mm lateral to the AC-PC line) of
the effects of the angle of trajectory. An extremely anterior entry point may result in a shallow
trajectory in the sagittal plane. The upper portions of the trajectory may pass through the pos-
terior limb of the internal capsule (C) or, worse, through the medial aspect of the globus pal-
lidus interna (B). In the latter case, microelectrode and semi-microelectrode recordings may
not differentiate the medial globus pallidus interna from the subthalamic nucleus, because
both structures are characterized by irregular neuronal activity, sustained high frequency,
high density at the site, and high density along the trajectory. DBS lead placement in an exces-
sively shallow trajectory in the sagittal plane increases the probability that posteriorly situ-
ated ventral contacts may effect the ascending medial lemniscus (not shown). Stimulation
of the ascending medial lemniscus resulting in paresthesias may limit DBS effectiveness.
Source: Modified from Schaltenbrand and Wahren (1977).
marked brain shift may register on intraoperative MRI or CT scans, and marked increase
in intracranial air may register on carefully performed intraoperative fluoroscopy.
Should the intraoperative neurophysiologist detect any substantial intracranial air,
she must consider aborting the surgery. Were it even possible to place the DBS lead in
an optimal position, resolution of the brain shift with reabsorption of the intracranial
air is likely to cause the DBS lead to migrate nonetheless, thus rendering ineffective
any postoperative therapy. Counterpoised to placing the DBS lead in the hope that any
migration would not be sufficient to reduce efficiency is the risk of complications aris-
ing from the second placement. Unfortunately, insufficient data exist to recommend a
specific course of action. The surgeon must prevent any intracranial air by using appro-
priate sealants to cover the dural defect.
Two findings suggest that the trajectory lies through the medial region of the glo-
bus pallidus interna rather than the subthalamic nucleus. This author has observed
that, on rare occasions, DBS leads implanted in the medial globus pallidus interna
Dorsal
Medial Lateral
Ventral
STN
ul f
ps o
r
e
ca mb
SN
al li
rn or
Fp 4.0
te ri
in ste
Po
FIGURE 9.3 Schematic representations depicting circumstances in which the microelectrode

or semi-microelectrode traverses the subthalamic nucleus but skirts the substantia nigra pars
reticulata, in the coronal plane 4 mm anterior to the midpoint of the AC-PC line. Either set of
circumstances leaves the intraoperative neurophysiologist unsure whether the trajectory lies
through the subthalamic nucleus or the medial globus pallidus interna. Source: Modified from
Schaltenbrand and Wahren (1977).
have led the patient to report “feeling weird,” yet no change of mood attended this
complaint. If stimulation produces tonic muscle contraction, the threshold decreases
as the electrode is moved posteriorly relative to the globus pallidus interna rather than
increasing, as it would if the electrode were in the subthalamic nucleus (Figure 9.2).
In the sagittal plane the DBS lead must pass between the posterior limb of the inter-
nal capsule and the ascending medial lemniscus. At 5 mm or so, the distance between
these two structures is relatively small. An excessively shallow DBS lead placement in
the sagittal plane may bring the upper (dorsal) contacts too close to the posterior limb of
the internal capsule and the lower (ventral) contacts too close to the medial lemniscus,
especially if the DBS leads with widely spaced contacts are used (biophysical properties
make use of such leads preferable [Montgomery 2010]).
Though somatotopic organization is normally configured in the subthalamic
nucleus with the leg medial and arm and head more lateral, this information is not
particularly useful for two reasons: (1) In the event that the leg area is encountered
on microelectrode recordings, one may be tempted to move the DBS lead laterally.
However, because of the relatively small size of the subthalamic nucleus and the
fact that the minimal change in distance for a new trajectory is 2 mm, repositioning
it that distance for a new trajectory may result in an excessively lateral placement
that is moreover too close to the posterior limb of the internal capsule; and (2) in
Parkinson’s disease, the somatotopy breaks down in such a way that an individual
neuron may respond to rotations of multiple joints in both upper and lower extremi-
ties, and this result may thus mislead the intraoperative neurophysiologist influenced
by localized sensorimotor driving.
REGIONAL PHYSIOLOGY OF THE ELECTRODE TRAJECTORY
Anterior Thalamus
Neuronal activity is typically irregular, transient, low frequency, low density within the
site, and low density within the trajectory (Figure 9.4). Generally, no sensorimotor driv-
ing is apparent. If the trajectory is posterior, however, neuronal activity from the ventral
lateral posterior oralis may be detected. The activity in the ventral lateral posterior ora-
lis is typically irregular, persistent, low to moderate frequency, low to moderate density
within the site, and low to moderate density in the trajectory (Figure 9.5). Though senso-
rimotor driven, the neuronal activity responds more readily to active joint rotations than
it does to passive joint rotations.
Posterior Limb of the Internal Capsule
A trajectory in the coronal plane may run parallel to the posterior limb of the internal
capsule. It is thus possible that most, if not all, of the trajectory lies in the posterior
limb of the internal capsule. Though neuronal activity is absent, one must consider
0.05 s
FIGURE 9.4 Representative example of a low-frequency and low-density microelectrode

recording site. There are two extracellular action potentials of only one amplitude.
0.5 s
FIGURE 9.5 Representative examples of moderate-frequency and moderate-density activity

within the microelectrode recording site. Though more than five sites are evident, their activ-
ity falls short of nearly continuous. In addition, there are at least three sets of extracellular
action potentials of different amplitudes, indicated by the arrows.
other reasons for the lack of neuronal activity. These include electrode failure, record-
ing system failure, or intracerebral hematoma. While microstimulation or macrostim-
ulation may produce tonic contraction indicative of the posterior limb of the internal
capsule, there is the possibility that too small a volume of activated tissue means that
the absence of any observable muscle contraction. In the absence of a muscle contrac-
tion, one cannot establish that the trajectory lies in the posterior limb.
Zona Incerta
Beneath the anterior thalamus and above the subthalamic nucleus lies the zona
incerta. This region consists of axons arriving from the globus pallidus interna
and elsewhere, and few neurons. Since the neurons are so few, neuronal activity is
rarely detected. When it is detected, it is irregular, transient, low density within the
site, and very low density within the trajectory. These qualities suggest that the site
encountered is indeed the zona incerta. Burst neurons may occasionally be detected.
Microstimulation or macrostimulation in this region may improve the symptoms of
Parkinson’s disease, an outcome likely owing to the fact that stimulating the lenticular
fasciculus serves to carry output fibers from the globus pallidus interna to the thala-
mus and thus may be analogous to globus pallidus interna DBS.
Lateral Subthalamic Nucleus
Neuronal activity is typically irregular, persistent, high frequency, high density within
the trajectory, high density within the trajectory consistent with location of the lat-
eral subthalamic nucleus, and robustly sensorimotor-driven by passive joint rotation.
Microstimulation and macrostimulation may improve symptoms of Parkinson’s disease.
FIGURE 9.6 Representative examples of high-frequency and high-density microelectrode

recording site. Evident are nearly continuous extracellular action potentials and at least six
sets of extracellular action potentials of different amplitudes.
Such improvement may not be seen, however, because the relatively small volume of
activated tissue and lack of response to microstimulation does not predict an absence of
response to DBS clinically (Figure 9.6).
Medial Subthalamic Nucleus
Neuronal activity is typically irregular, persistent, moderate to high frequency, moder-

ate to high density within the trajectory, moderate density within the trajectory con-
sistent with location of the medial subthalamic nucleus, and at best only moderately
sensorimotor-driven by passive joint rotation. Microstimulation and macrostimulation
typically do not improve symptoms of Parkinson’s disease. Some patients report expe-
riencing a sudden onset of depression or a feeling of impending doom with stimulation
in this region, which cessation of stimulation resolves.
Substantia Nigra Pars Reticulata
Neuronal activity is typically regular, sustained, high frequency, low density within a
site, and moderate density in the trajectory consistent with location of the substantia
nigra pars reticulata (Figure 9.7).
FIGURE 9.7 Representative example of a microelectrode recording in the substantia nigra pars

reticulata. As can be seen, the discharges are regular, high frequency, and sustained. There is
only a single amplitude of the neuronal action potentials indicating a low density within the
recording site.
RECONSTRUCTING THE REGIONAL ANATOMY IN

CONFORMITY TO THE PHYSIOLOGY
Intraoperative neurophysiological monitoring requires one to reconstruct the regional

anatomy around the trajectory in conformity to neurophysiological findings, both record-
ing and microstimulation. This author uses microstimulation with high frequencies—300
pps, for example—to drive a physiological response, in order to take advantage of temporal
summation (see c hapter 8) rather than attempting to mimic the clinical effects of DBS.
Mapping the anatomy according to the physiology proves difficult for two rea-
sons: (1) differences in regional anatomy among patients and (2) variation of actual
trajectories within and between patients. Either reason may owe to idiosyncracies of a
particular patient’s anatomy or to vagaries of the surgical procedure—the stereotactic
frame or apparatus’s orientation vis-à-vis a patient’s unique anatomy, for example, or pre-
cautions taken (if indeed any are) to prevent intracranial air and subsequent brain shift.
Intraoperative neurophysiologists can visualize a patient’s regional anatomy by call-
ing to mind representations from anatomy atlases. These remembered representations
they then morph and rotate in their minds until they match the electrophysiological
findings. Neurophysiologists can acquire this skill by frequently consulting stereo-
tactic human atlases and by attempting to predict the electrophysiological findings
as a trial-and-error method of learning. This author has created a three-dimensional
atlas by tracing lines from an atlas on transparencies, which he mounted on stack-
able Plexiglas panes in order to represent the anatomies. These allowed him to rotate
the three-dimensional images in order to map electrophysiological results on a spe-
cific anatomical location and orientation. After using these representations for a time,
this author became able to picture them in his mind. In addition, the probe’s view in
image-guided navigation also provides a sense of the regional anatomy around the
trajectory. For this reason it is important that intraoperative neurophysiologists be
involved in the image-guided navigation or, at the least, review the navigation.
Some intraoperative neurophysiologists find such mental imaging, morphing, and
rotations difficult, finding it easier to use language syntax or some other set of rules
instead. Descriptions of various trajectories may supply these rules. Yet deducing these
rules proves difficult, because the number of possible trajectories is very large. Syntax
in the form of a decision tree mitigates this difficulty to some extent (Appendix A).
Frequent use of the algorithms simultaneously with stereotactic atlases has enabled
some intraoperative neurophysiologists to intuit visual representations.
The decision tree rests on a logic that divides the microelectrode trajectory into an
upper segment and lower segment. In the case of the subthalamic nucleus, the upper
trajectory typically corresponds to the anterior thalamus. The lower trajectory corre-
sponds to the structures below the thalamus, including the zona incerta, the posterior
limb of the internal capsule, the medial and lateral subthalamic nucleus and the sub-
stantia nigra pars reticulata.
The angle of trajectory is such that the trajectory must often pass through sev-
eral structures. Lacking in the upper trajectory are sensorimotor-driven neurons
and any unique discharge characteristic save that of low frequency. Importantly,
recording in the dorsal tier demonstrates proper functioning of microelectrode and
semi-microelectrode systems. Neurons of the dorsal teir also show an absence of sen-
sorimotor driving in every case except that of the ventral posterior oral thalamus (dis-
cussed above).
These algorithms are generalizations and approximations; actual physiological and
anatomical correlates may vary greatly, depending on the spatial orientation of the
trajectories to the patient’s specific anatomy. Though a patient’s regional anatomy cor-
relates closely to the line connecting the anterior commissure to the posterior com-
missure (AC-PC line), the specific orientation of the AC-PC line to the stereotactic
apparatus may vary greatly according to the six degrees of spatial freedom (translation
in the three axes and rotations about the three axes in the Cartesian space). The respon-
sibility of the use of this heuristic thus falls on the physician and healthcare professional
who must exercise her own judgment in caring for her patient.
SUMMARY
The unique characteristics of the sensorimotor region lying in the lateral subthalamic
nucleus allow for relatively robust identification. Readers must bear in mind that many
times neuronal activities of the subthalamic nucleus prove difficult to distinguish from
the medial globus pallidus interna. Signs that the DBS has likely entered the medial
globus pallidus interna include a lower threshold to tonic muscle contraction as trajec-
tories are repositioned posteriorly and a patient’s report of an unusual feeling that no
change in mood attends.
REFERENCES
Yelnik J and Percheron G. Subthalamic neurons in primates: a quantitative and comparative analysis.
Neuroscience 4(11): 1717–1743, 1979.
/ / / 10 / / / THE GLOBUS PALLIDUS
INTERNA NUCLEUS
INTRODUCTION
Intraoperative neurophysiological monitoring of the globus pallidus interna nucleus aids

the surgeon in implanting the Deep Brain Stimulation (DBS) lead in the vicinity of the
globus pallidus interna. The implantation site is selected with a view to maximizing relief
of a patient’s disorder and minimizing any adverse effects. An intraoperative neurophysi-
ologist faces the challenge of using neurophysiological properties to identify a target that is
predictive of subsequent clinical response—a task which requires not only that she target
to establish efficacy but also to prevent intolerable side effects that prevent optimal DBS
stimulation. Some relatively straightforward adverse effects can be attributed to specific
anatomical structures surrounding the target. Because the posterior limb of the internal
capsule, for example, contains corticobulbar and corticospinal tracts that descend imme-
diately posterior and ventral to the globus pallidus interna, an excessively posterior and
ventral DBS lead placement in this structure may produce intolerable tonic contractions
that may limit DBS effectiveness. Further, the optic tract lies below the globus pallidus
interna, and inadvertent stimulation may result in phosphenes or other visual distur-
bances. Consequently, globus pallidus interna targeting necessitates the specific identifi-
cation of the anterior border of the posterior limb and the dorsal border of the optic tract.
Other adverse effects, such as when patients describe having odd feelings, which are more
difficult to relate to specific anatomy, become likelier with a medial DBS lead placement.
Larger than the subthalamic nucleus, the sensorimotor region of the globus pallidus
interna is the primary target. Its larger size, however, presents a problem: The effective
volume of tissue activated by electrodes on typical DBS leads is so small that affect-
ing the entire homuncular representation may prove impossible. An intraoperative
neurophysiologist must use a patient’s symptoms to guide her in determining which
17 2
10. The Globus Pallidus Interna Nucleus / / 173
homuncular representation has priority. Cervical dystonia, for example, signals to her
that she must target the head homuncular representation. Patients with generalized
dystonia, in whom the problem is particularly acute, may require an additional DBS
lead placed in the globus pallidus interna. This author thus favors DBS leads whose
contacts are more widely spaced, because they activate a greater volume of tissue than
do contacts whose leads are closer together (Montgomery 2010).
For DBS of the globus pallidus interna the angle of entry is important. Excessively
shallow DBS lead placement in the sagittal plane may situate the upper contacts ante-
rior to the sensorimotor region and thus reduce efficacy. Similarly, excessively shallow
placement in the coronal plane may send the DBS lead through the entire homuncular
representation. Consequently, only a few contacts enter the homuncular representa-
tion in the area most critical for relieving the patient’s worst symptoms.
CRITERIA FOR PHYSIOLOGICALLY DEFINED OR

SYMPTOMATICALLY DEFINED OPTIMAL TARGET
As discussed above, a preeminent criterion requires that the intraoperative neurophysi-

ologist must place the DBS lead in the homuncular representation most consistent with
the patient’s worst symptoms. The sensorimotor region of the globus pallidus interna
abuts the anterior border of the posterior limb of the internal capsule. Consequently,
the optimal trajectory in the homuncular representation lies sufficiently anterior to pre-
vent the volume of activated tissue from extending posteriorly to the posterior limb of
the internal capsule. Similarly, the most ventral contact of the DBS lead must be suf-
ficiently far from the optic tract to prevent visual disturbances.
The optimal distance between the DBS lead and the anterior border depends on the
anticipated stimulation intensities, which in turn depend on the disorder being treated.
In Parkinson’s disease, for example, 3.5 volts delivered by a constant-voltage stimula-
tor, or 2.5 mA delivered by a constant-current stimulator, represents typical stimula-
tion. This voltage/current corresponds to a distance of approximately 2 to 3 mm. The
intraoperative neurophysiologist must therefore place the DBS 2 to 3 mm anterior to
the anterior border of the internal capsule’s posterior limb. Some patients with dystonia
may require twice the stimulation voltage/current. In such cases, the intraoperative
neurophysiologist must place the DBS lead approximately 4 mm anterior to the anterior
border of the internal capsule’s posterior limb. She does so to avoid a DBS lead place-
ment that lies too close to the internal capsule’s posterior limb, because from such a
placement result tonic muscle contraction and other adverse effects that limit effective
DBS. At the same time, she thus faces the significant challenge of avoiding a DBS lead
placement too distant anteriorly from the internal capsule’s posterior limb, which as a
consequence places fewer contacts in the sensorimotor region and decreases efficacy.
The above discussion makes clear that, rather than the sensorimotor region of the
globus pallidus interna, the critical target in globus pallidus interna DBS is the ante-
rior border of the internal capsule’s posterior limb. The intraoperative neurophysiolo-
gist must locate the anterior border of the internal capsule’s posterior limb in order to
avoid it. If the intraoperative physiologist bases her trajectory solely on the appropri-
ate homuncular representation, she risks an excessively anterior or posterior DBS lead
placement. The situation with globus pallidus interna DBS is therefore different from
the situation with subthalamic nucleus DBS.
As with the situation of avoiding the posterior limb of the internal capsule, the depth
of the optimal DBS trajectory relative to the location of the optic tract also depends on
the patient’s diagnosis and age. Excessively ventral (deep) DBS lead placements risk
stimulating the optic tract and producing visual hallucinations of distorted images or
bright spots known as phosphenes. Again, the bottom of the DBS lead must rest 2 mm
to 3 mm above the optic tract in patients with Parkinson’s disease and 4 mm above in
patients with dystonia.
Children present a special case. The US Food and Drug Administration (FDA)
has approved DBS for dystonia in children aged 7 years or older. The FDA based its
approval on the fact that by age 7 years, a child’s head has reach 90% of its anticipated
adult size in circumference. Yet the DBS lead may nonetheless migrate (drift upward)
as the child’s head grows an additional 10%. This author therefore places the distal edge
of the second deepest, or distal, contact at the bottom of the optimal DBS trajectory.
The deepest, or distal-most, contact, meanwhile, may migrate to the optimal position as
the child’s head grows. Certain severe situations have made necessary DBS surgery for
children younger than 7 years, for example to treat serious disorders, such as dystonic
storm, as an “off-label” use of an FDA approved device. In such cases, the intraopera-
tive neurophysiologist must carefully consider DBS lead placement depth. Similarly,
such cases require a longer extension wire connecting the DBS lead to the stimulator
in order to account for continued growth of the body. The child’s body size dictates
DBS implanted pulse generator placement. Also, a smaller rechargeable implanted
pulse generator, which this author prefers, decreases risk of skin erosion and repeated
implanted pulse generator replacements.
The requirement that microstimulation and macrostimulation produce no adverse
effects is rather straightforward. Microstimulation permits more precise estimates of
the distance between the microstimulation site and the anatomical structure responsi-
ble for tonic muscle contraction or other adverse effects. Though with microstimulation
and some forms of macrostimulation one may observe an absence of adverse effects,
she must refrain from concluding that therapeutic macrostimulation will not produce
them. If microstimulation produces adverse effects, then macrostimulation will likely
also produce them.
The requirement of some symptomatic benefit with stimulation is problematic for
two reasons. First, stimulation may activate an insufficient volume of tissue, particularly
if it is introduced through the microelectrode or semi-microelectrode tip, or through
the indifferent electrical contact in bipolar microelectrodes and semi-microelectrodes.
Second, mere insertion of the electrode (more typically the DBS lead) without stimula-
tion may produce a micropallidotomy effect, in which improvement masks any benefit
to subsequent stimulation. The intraoperative neurophysiologist must also test a range
of symptoms. Muscle tone or rigidity, for example, is a symptom that most frequently
demonstrates improvement.
Intraoperative neurophysiologists who prefer macrostimulation through DBS leads
to microelectrodes or semi-microelectrodes must base their criteria on the ability to
demonstrate the absence of adverse effects. Surgeons using general anesthesia must
bear in mind that, though macrostimulation may demonstrate tonic contraction, the
patient will be incapable of reporting phosphenes or other visual disturbances associ-
ated with DBS lead placement excessively close to the optic tract. As discussed above,
the intraoperative neurophysiologist must avoid adverse effects that may limit subse-
quent therapeutic DBS.
An intraoperative physiologist must satisfy the additional criterion of demonstrat-
ing a reasonable therapeutic window as defined by the difference in the threshold at
which symptoms improve and the threshold at which persistent adverse effects are pro-
duced when conducting intraoperative DBS testing. One approach often used is to select
stimulation parameters, such as frequency, voltage/current (with constant-voltage or
constant-current stimulators), and pulse width reflective of those typically reported in
the literature. This is inadvisable, as the actual stimulation parameters needed for post-
operative clinical efficacy may be substantially different. Consider the situation where a
patient experiences a significant micropallidotomy effect, which prevents intraoperative
DBS test stimulation to determine efficacy. The intraoperative neurophysiologist con-
ducts test stimulation only using 3 volts or 2.5 milliamps, in the case of constant-voltage
or constant-current simulation respectively. Postoperatively, it is found that the patient
requires 4 volts or 3 milliamps to achieve benefit but experiences limiting side effects.
Had the intraoperative neurophysiologist realized the potential that higher stimula-
tion strengths would be necessary, she could have tested at these higher strengths and if
adverse effects resulted, she could reposition the DBS lead accordingly. Consequently,
this author stimulates at 160 pulses per second with a pulse width of 120 μs and with
voltage or currents up to 4 and 3 milliamps respectively, in the case of Parkinson’s dis-
ease. For dystonia, voltages and currents are tested up to at least 5 volts and often to at
least 8 volts or at least 3.5 milliamps and up to 4.5 milliamps. Bear in mind the caveat
that this author has not the extensive experience with constant-current stimulation in
patients with dystonia to makes these recommendations with confidence.
All forms of intraoperative neurophysiological monitoring require that an intraoperative

neurophysiologist understand the regional anatomy around the entire electrode trajectory.
For microelectrode and semi-microelectrode recordings, she must study the entire trajec-
tory, beginning approximately 25 mm above the bottom of the globus pallidus interna as
revealed by image-guided targeting. Recordings in the putamen made well above the glo-
bus pallidus interna, for example, may be of aid in estimating the relative position of the tra-
jectory in the anterior-posterior and medial-lateral direction (discussed in detail below).
The rationale for beginning the microelectrode or semi-microelectrode trajectory
approximately 25 mm above the bottom of the globus pallidus interna nucleus rests
on three reasons: (1) It allows the intraoperative neurophysiologist to obtain sufficient
information from recordings; (2) it allows the intraoperative neurophysiologist to ele-
vate the guide cannula end well above the surface of the skull, which keeps the DBS lead
in sight and under control prior to cannula removal; and (3) it minimizes risk of intra-
cerebral hematoma caused by sharp microelectrode or semi-microelectrode tips (the
stylus in the guide the cannula has a blunt tip that tends to push aside tissue rather than
pierce it) that would accompany starting the microelectrode or semi-microelectrode
further above the target.
An intraoperative neurophysiologist opting for macrostimulation through the DBS
lead may use a longer cannula, but she must take care to halt the cannula above the
dorsal-most DBS lead contact. Though a longer cannula reduces risk of deflection, in
some systems it may prove difficult to withdraw it in such a way as to leave sufficient
clearance for fixing the DBS lead in place in the burr hole in the skull.
The angle of entry in the sagittal and coronal plane figures crucially in determin-
ing the regional anatomy. For angles lying in the coronal plane, the entry point is often
selected with a view to avoiding the lateral ventricles, because the ependymal surface
may deflect the cannula, electrode, or DBS lead as it exits the ventricle. A DBS lead
trajectory that is too shallow in the coronal plane risks placing the more ventral con-
tacts too medial and the upper contacts too lateral, depending on the anterior-posterior
plane (Figure 10.1).
in
Dorsal
cp.
Anterior Rt.po
e
.i.
Cp
Medial Lateral A La
.p
.l Put
P.l
Posterior .e
Ventral P.m
.i
Lop
P.m
La.
l
p.li
cm
. .a Cl.sst
An d .l
p An
C B
B .sII
cm
sA
II p.A
D
Leg
Arm
Head
FIGURE 10.1 Schematic representations of the effects of placing a DBS lead in a trajectory

that is too shallow in the coronal plane. Figure A is a coronal section 5.5 mm anterior to the
midpoint of a line connecting the anterior and posterior commissures (AC-PC line). Figure
B shows the sensorimotor region as a cylinder that runs from anterior and dorsal to poste-
rior and ventral. Thus a DBS lead entering from a shallow angle in the coronal plane could
have contacts that extend beyond the sensorimotor region into the medial globus pallidus
interna (C) or have only the ventral contacts in the sensorimotor region. Source: Modified from
The typical angle of entry in the sagittal plane lies sufficiently anterior to the
coronal suture, away from the motor cortex. An excessively anterior entry point and
excessively shallow trajectory angle may situate the DBS lead’s upper contacts ante-
rior to the sensorimotor region of the globus pallidus interna (Figure 10.2), particu-
larly if there occurs any posterior brain shift as a result of gravity or intracranial air.
Lack of changes in neuronal activity related to sensorimotor testing suggests that
a posterior shift has occurred. Intraoperative neurophysiologists who forgo use of
microelectrode or semi-microelectrode recordings, however, do not have this evi-
dence available to them.
Two findings suggest that the trajectory lies through the medial region of the glo-
bus pallidus interna rather than through the lateral region, where sensorimotor neu-
rons predominate. This author has observed DBS leads implanted in the medial globus
pallidus interna rather than the subthalamic nucleus consequent to posterior brain
displacement caused by excessive air entering the intracranial vault and expanding
(Figure 10.3). Stimulation in the region subsequently identified as the medial globus
pallidus interna often produced an effect that “felt weird” according to patients’ subjec-
tive reports, but no change in mood accompanied it.
D.o.e A
Cp.i.g
Rt
z.o La
.p.
l Cp
.i.
.i.p
a
V.o.p
Cp
P.l
La e
P.m
.p
.m
.
B
La.p.li
P.m
.i
Cm.a
An.
l B
Str. pd. i
GPI
II A. tb
Sensori-
motor
Dorsal
Posterior Anterior
Ventral
FIGURE 10.2 Schematic representation of the effects of an excessively shallow trajectory in

the sagittal plane. Figure A shows a sagittal section 20 mm lateral to the AC-PC line. The sen-
sorimotor region is posterior in the globus pallidus interna. As can be seen, a DBS lead too
shallow in the sagittal plane could have the dorsal contacts in the nonmotor region, thereby
reducing efficacy. Source: Modified from Schaltenbrand and Wahren (1977).
A C
B e
d
FIGURE 10.3 MRI scan depicting trajectories through the medial globus pallidus interna during
surgery to implant the subthalamic nucleus. Axial MRI (A) and reconstructed sagittal (B) and
coronal (C) planes containing the DBS lead. The tip of the DBS lead in A appears to lie at the
posterior edge of the posterior limb of the internal capsule. In the coronal plane (C), there
appear the DBS lead and remnants of two posterior DBS lead trajectories (d and e). The lead
appears to lie between the posterior limb of the internal capsule with medial globus pallidus
interna. The upper contacts appear to lie within the medial globus pallidus interna.
in
Dorsal
cp.
Anterior
g
Rt.po
e
.i.
Cp
Medial Lateral A La
.p Put
.l
P.l
La
Posterior
. p.
.e
Ventral P.m
m
.i
La.p
P.m
.l
La.p
cm
.li
.a Cl.sst
pd .l
An B
B .sII
cm sA
II pA
Putamen
Globus pallidus externa
Border cell area
Globus pallidus interna
Optic tract
Corticobulbar and corticospinal
tract
FIGURE 10.4 Regional anatomy around the globus pallidus interna. Schematic representations
of the coronal plane 5.5 mm anterior to the midpoint of a line connecting the anterior and
posterior commissures (AC-PC line) (A). Figure B shows the globus pallidus interna posterior
and medial to the globus pallidus externa and separated by a region of border cells. The puta-
men is anterior and dorsal to the globus pallidus interna. The corticobulbar and corticospinal
tracts, in the posterior limb of the internal capsule, are posterior while the optic tract is ventral
to the globus pallidus externa. Source: Modified from Schaltenbrand and Wahren (1977).
Regional anatomy in the vicinity of the lateral globus pallidus interna is shown
in Figure 10.4. In the coronal plane, the posterior limb of the internal capsule lies
medially to the globus pallidus interna. The globus pallidus externa lies laterally to
the globus pallidus interna, the putamen lies dorsally, and the optic tract ventrally.
In the sagittal plane, the posterior limb of the internal capsule runs posteriorly to the
globus pallidus interna. The globus pallidus externa occupies an anterior position in
the sagittal plane, the putamen an anterior and superior position, and the optic tract a
ventral position. As the sagittal section shows most clearly, border cells surround the
globus pallidus interna.
REGIONAL NEURONAL PHYSIOLOGY
What follows is a description, elaborated from the perspective of microelectrode

recordings, of neurophysiological characteristics of the various anatomical structures
in the vicinity of the globus pallidus interna’s sensorimotor region. One may generalize
these findings to semi-microelectrode recordings, bearing in mind the latters’ inferior
ability to isolate individual neurons.
Putamen
Unless the trajectory is very medial, an intraoperative neurophysiologist typically first

encounters the putamen. Neuronal activities she encounters are of low density within
the segment of the trajectory that is within the putamen, few extracellular action
potentials of various amplitudes, and waveforms consistent with low density within
the recording site. When first encountered, the neuronal discharges are at a low rate
(low frequency) and are transient (Figure 10.5). Often, the intraoperative neurophysi-
ologist must therefore traverse the trajectory a few millimeters before she encounters
a site at which she may discern extracellular action potentials (low density within the
trajectory). The neuronal activity is irregular. The medial putamen neurons are typi-
cally not sensorimotor driven, unless the intraoperative neurophysiologist uses an
extremely shallow angle of entry and thus enters the putamen laterally.
Globus Pallidus Externa
Sustained high frequencies within a recording site and extracellular action potentials
of different amplitudes, when evident, indicate a high neuronal density at a recording
site. The relatively short distance between recording sites demonstrating extracellu-
lar action potentials (high density within the trajectory) renders the globus pallidus
externa readily discernible from the putamen above.
Though the neuronal activity is irregular, using microelectrode recordings one
may discern in Parkinson’s disease two relatively unique neuronal discharge pat-
terns. The high-frequency discharges of some neurons are interrupted by pauses
(Figure 10.6). Neurons of this kind are known as high-frequency-pause neurons.
Other discharges describe a low-frequency bursting pattern. These patterns also
help one to distinguish the globus pallidus externa from the globus pallidus interna
(discussed below). These patterns are often observed in patients with Parkinson’s
disease but may not be observed in patients with dystonia and other conditions. In
patients of the second sort, neuronal activities in the globus pallidus externa tend to
FIGURE 10.5 Representative example of transient neuronal activities. Neuronal activity begins

abruptly, then fades.
A 12 s
B 2s
C 2s
FIGURE 10.6 Representative examples of high-frequency-pause neurons indicative of the

globus pallidus interna in Parkinson’s disease (possibly unobservable in the globus pallidus
interna of patients with dystonia). Figures A and B are from the same neuronal recording sites
but cover different time scales. Figure C is another example from a different recording site.
be high frequency, irregular, high density within a recording site, and high density
within the trajectory. They moreover tend to lack high-frequency-pause neurons and
low-frequency bursting. Distinguishing the globus pallidus externa from the globus
pallidus interna in patients with dystonia may thus prove difficult.
Border Cells
Border cells form a shell surrounding the globus pallidus interna. One may find them
between the globus pallidus externa and the globus pallidus interna, behind the globus
pallidus interna, and beneath the globus pallidus interna. The recording sites within the
border cell regions are typically sustained moderate frequency with moderate regular-
ity, low density within the recording site, and low density within the trajectory. The
trajectory through the border cell region is small (Figure 10.7).
Globus Pallidus Interna
Extracellular action potentials generally emit sustained high frequencies within a record-
ing site and multiple extracellular action potentials of different amplitudes. The fact that
they are detectable suggests a high neuronal density within a recording site. The relatively
short distance between recording sites indicates extracellular action potentials (high
Border cell
10 s
B Nonborder cell
3.5 s
FIGURE 10.7 Representative examples of two microelectrode recordings sites (A and B) con-
taining border cells. In A, the border cells are indicated by the downward arrows. In B, nearly
all extracellular action potentials are from border cells. C shows an expanded few extracel-
lular action potentials of a border cell compared to a cell of another kind. The much broader
extracellular action potential in the border cell causes it to emit a low-pitch sound.
density within the trajectory). The neuronal activity is irregular (Figure 10.8). In the case
of Parkinson’s disease, stimulation may improve symptoms; in hyperkinetic syndromes, it
may reduce them. Stimulation, however, is unlikely to produce any symptomatic effect in
dystonia. Microelectrode and semi-microelectrode stimulation may not produce any ben-
efit, because the volume of tissue activation may be too small. Failure to produce a symp-
tomatic benefit is no indication that the electrode has failed to reach an optimal position.
The sensorimotor homunculus is distributed in the lateral globus pallidus interna, a
structure in which the lower extremity representation is relatively anterior, dorsal, and
medial and the head representation posterior, ventral, and lateral. The upper extremity rep-
resentation lies between the lower extremity and the head representation. The typical neu-
rons encountered as one traverses the medial portion of the globus pallidus interna may
not be sensorimotor related, particularly if the trajectory is shallow in the coronal plane.
Optic Tract
The small amplitude and brief duration of extracellular action potentials within the
optic tract make discrimination by use of microelectrode recordings difficult. As
the intraoperative neurophysiologist approaches the optic tract, however, she may
Regular
Irregular
FIGURE 10.8 An example of the irregular activity characteristic of the globus pallidus interna.
An example of a regular high-frequency neuron (substantia nigra pars reticulata) is shown for
comparison.
hear the extracellular action potentials, which produce a hiss. Photic stimulation
may produce a modulation of the hiss that correlates to the photic stimulation pulse.
Microelectrode and semi-microelectrode stimulation in close proximity to the optic
tract can produce phosphenes or other visual distortions.
Because this region contains few neurons (if any), neuronal activities in terms of extra-
cellular action potentials are generally indiscernible. Stimulation of the posterior limb
or in the vicinity of the internal capsule may produce muscle contraction by activating
the corticobulbar and corticospinal pathways. In the case of stimulation through the
microelectrode or semi-microelectrode, however, the volume of activated tissue may
be insufficient to produce an observable response.
RECONSTRUCTING REGIONAL ANATOMY ACCORDING TO PHYSIOLOGY
Key to intraoperative neurophysiological monitoring is reconstructing, according to

neurophysiological findings, regional anatomy around the trajectory. This is a compli-
cated task, because each patient’s regional anatomy is different, and actual trajectories
within and among patients vary. Differences and variance may owe to the idiosyncratic
anatomy of an individual patient or to vagaries of the surgical procedure—a particular
stereotactic frame or apparatus’s orientation relative to the patient’s unique anatomy,

for example. Consequently, it is important for the intraoperative neurophysiologist to
be involved in or at least review the image-guided navigation. The “probe’s view” can
help the intraoperative neurophysiologist to anticipate the findings during the micro-
electrode or semi-microelectrode recordings.
Some intraoperative neurophysiologists develop the ability to visualize regional
anatomy. Their visualizations they base on images recalled from anatomy atlases.
They then visually morph and rotate the mental image in such a way as to bring it
into correspondence with electrophysiological findings. Many intraoperative neu-
rophysiologists acquire this skill by frequent use of human stereotactic atlases and
by their trial-and-error attempts at predicting electrophysiological findings. This
author fashioned a three-dimensional atlas by tracing lines from an atlas onto trans-
parencies. These transparencies he mounted on Plexiglas in order to stack them to
create three-dimensional anatomical representations. Rotating them enabled him
to map the electrophysiological results to a specific location and orientation in the
three-dimensional images. In time, however, he became able simply to imagine them.
Intraoperative neurophysiologists for whom such mental imaging, morphing, and
rotation prove difficult find it easier to rely instead on language syntax or some other
set of rules. Descriptions of various trajectories may supply these rules. Yet deducing
these rules proves difficult, because the number of possible trajectories is extremely
large. Syntax in the form of a decision tree greatly mitigates this difficulty (Appendix
C). Frequent use of the algorithms simultaneously with stereotactic atlases has enabled
some intraoperative neurophysiologists to intuit visual representations.
The decision tree rests on a logic that divides the microelectrode trajectory into
an upper segment and lower segment. The lower segment is divided into an initial,
middle, and bottom portion for reasons explained in the algorithms (see Appendix
C, Figure 10.9). In the case of the globus pallidus interna, the upper trajectory cor-
responds to the putamen. The lower trajectory corresponds to structures below the
putamen, including the globus palldius externa, globus palldius interna, the posterior
limb of the internal capsule, and the optic tract.
The angle of trajectory may be such that several structures are traversed. Lacking in the
upper trajectory are sensorimotor drive and any unique discharge characteristic save that
of low frequency. Importantly, recording in the putamen typically helps to demonstrate
proper functioning of microelectrode and semi-microelectrode systems. The neurons also
show an absence of sensorimotor driving except if the angle of entry is very shallow in the
coronal plane due to a very lateral entry such that the very lateral putamen is traversed.
Dorsal
Medial Lateral
Ventral
Internal
capsule
Upper trajectory Low frequency, transient, irregular, low density
P High frequency, sustained, irregular, high density

t High frequency-pause or low frequency burst
Initial lower trajectory In Parkinson’s disease in upper region, moderate
GPe frequency, transient, regular, moderate density in
GPi
Middle lower trajectory lower region
High frequency, sustained, irregular, high density

Bottom lower trajectory Border cell
area
High frequency, sustained, regular, low density in upper
Optic region, related to light flashes or phosphenes with
tract stimulation in the lower region
FIGURE 10.9 Schematic representation of an electrode’s regional divisions. The image is of a

coronal section 5.5 mm anterior to the midpoint of the AC-PC line. The trajectory is schemati-
cally represented to illustrate the heuristic for interpretation. The trajectory is divided into an
upper and lower trajectory with the lower trajectory further divided into an initial, middle, and
bottom trajectories. The characteristics of neuronal activities in each of these segments help
to define the regional anatomy around the trajectory (see Appendix C). Source: Modified from
These algorithms in the appendix are generalizations and approximations; actual

physiological and anatomical correlates may vary greatly depending on the spatial ori-
entation of the trajectories to the patient’s specific anatomy. Though a patient’s regional
anatomy correlates closely to the line connecting the anterior commissure to the posterior
commissure (AC-PC line), the specific orientation of the AC-PC line to the stereotactic
apparatus may vary greatly according to the six degrees of spatial freedom (translation in
the three axes and rotations about the three axes in the Cartesian space). The responsibil-
ity of the use of these heuristics in the appendices thus falls on the physician and health-
care professional who must exercise her own judgment in caring for her patient.
SUMMARY
The unique characteristics of the sensorimotor region lying in the globus pallidus
interna allow for relatively robust identification. Readers must bear in mind that many
times neuronal activities of the globus pallidus interna prove difficult to distinguish
from those of the medial subthalamic nucleus. Signs that the DBS has likely entered the
medial globus pallidus interna include a lower threshold to tonic muscle contraction as
trajectories are repositioned posteriorly and a patient’s report of an unusual feeling that
no change in mood attends.
REFERENCES
/ / / 11 / / / THE VENTRAL INTERMEDIATE
NUCLEUS OF THE THALAMUS
INTRODUCTION
Benefiting the patient while minimizing any adverse effects requires that the intraop-
erative neurophysiologist place the DBS lead in a location in the vicinity of the ven-
tral intermediate nucleus. The intraoperative neurophysiologist faces the challenge of
using neurophysiological properties to identify a target that predicts subsequent clini-
cal response. Meeting this challenge requires that the intraoperative neurophysiolo-
gist target for the purpose of not only determining efficacy but also preventing effects
potentially adverse to DBS efficacy (adverse in the sense that patients experience intol-
erable side effects that limit therapy). Relatively straightforward adverse effects can be
attributed to specific anatomical structures surrounding the target. The posterior limb
of the internal capsule, for example, contains corticobulbar and corticospinal tracts
that descend immediately laterally and ventrally to the ventral intermediate nucleus.
Excessively lateral and ventral placement the DBS lead risks producing intolerable
tonic contractions that limit DBS’s effectiveness. Some adverse effects—feelings of
depression that patients report experiencing, for example—are more difficult to relate
to specific anatomy.
The primary target, the ventral intermediate nucleus, has a particularly prob-
lematic anatomical distribution: rather large in the dorsal-ventral and medial-lateral
dimensions and relatively narrow in the anterior-posterior direction (Figure 11.1). The
nucleus’s small size in the anterior-posterior direction means that critical importance
attaches to the angle of trajectory in the sagittal plane. An excessively shallow angle
in the sagittal plane sends the DBS lead tangentially through the ventral intermedi-
ate nucleus of the thalamus, and the lead trajectory establishes relatively few stimu-
lation contacts therein. Along such a trajectory, the ventral-most contacts also move
187
DBS lead
Vc-deep Vim Po
s
int terio
ern r li
Vc-tactile al c mb
aps
ule
Vop
Medial
lemniscus STN
Sagittal plane
Dorsal 16.0 mm lateral

to the AC-PC
Posterior Anterior line
Ventral
FIGURE 11.1 The anatomy of the ventral intermediate thalamus presents significant targeting
challenges. The figure is a coronal section 16 mm lateral to the line connecting the anterior
and posterior commissures (AC-PC line). The target for thalamic DBS is the ventral interme-
diate nucleus (Vim), which is the cerebellar relay nucleus. Just anterior to the ventral inter-
mediate thalamus is the ventral lateral posterior oralis nucleus, and posterior is the ventral
caudal nucleus. The ventral caudal thalamus is divided into two sections. The anterior section
receives deep sensory inputs from, for example, joint capsule receptors. The posterior sec-
tion of the ventral caudal thalamus receives superficial sensations, for example, from cuta-
neous receptors. The small size in the anterior-posterior direction means that the angle of
trajectory in the sagittal plane is critical. An excessively shallow trajectory in the sagittal plane
sends the DBS lead couple tangentially through the ventral intermediate nucleus of the thala-
mus, and the lead couple establishes relatively few stimulation contacts therein. Along such
a trajectory, the ventral-most contacts also move significantly posterior toward the tactile
region of the ventral intermediate thalamus. This positioning may produce paresthesias that
limit DBS’s efficacy. Source: Modified from Schaltenbrand and Wahren (1977).
significantly posterior toward the tactile region of the ventral caudal thalamus. This
latter repositioning may produce paresthesias that limit DBS’s efficacy. It is thus impor-
tant fully to traverse the thalamus in the dorsal-ventral direction as vertically as pos-
sible while avoiding penetration of the motor areas of the cerebral cortex.
The larger size in the medial-lateral dimension coincides with the homuncular
representation, which is layered like half an onion, the central core corresponding to
the head, a half-sphere around the core corresponding to the upper extremity and a
half-sphere around the upper extremity representation corresponding to the lower
extremity (Figure 11.2). Proper positioning of the DBS lead in the medial-lateral direc-
tion allows intraoperative neurophysiologist to target the appropriate homuncular rep-
resentation. She must avoid the head homuncular representation in order to reduce risk
11. The Ventral Intermediate Nucleus of the Thalamus / / 189
CORONAL SOMATOTOPIC ARRANGEMENT

AXIAL SOMATOTOPIC ARRANGEMENT
FIGURE 11.2 Coronal (left) and axial (right) section showing the homuncular organization in
the ventral intermediate thalamus. As can be seen, the homunculus is organized from medial,
head representation, to lateral, lower-extremity representation, with the upper-extremity
representation between. In the coronal plane the lower-extremity representation curves
beneath the upper-extremity representation, which curves beneath the head representation.
Source: Modified from Hassler in Schaltenbrand and Wahren (1977).
to speech, language, and swallowing. For example, a neurophysiologist who locates in

the upper extremity neuronal activity responsive to muscle palpation may think this
sufficient and may wrongly infer that the trajectory is satisfactory. Yet had the neuro-
physiologist continued to record throughout the electrode’s descent, she might have
encountered neuronal activity related to opening and closing and jaw and palpation of
the masseter muscle—activity suggestive of an excessively medial trajectory.
The effective volume of tissue activation through electrodes on typical DBS leads does
not obviate the possibility of affecting the entire homuncular representation. The intra-
operative neurophysiologist must therefore prioritize various homuncular representations
according the patient’s symptoms, focusing on those associated with the most incapacitat-
ing tremor. Because dysarthria, subcortical aphasia, and dysphagia are significant com-
plications, the head representation medially in the ventral intermediate nucleus is best
avoided. The intraoperative neurophysiologist must determine whether upper-extremity
symptoms take priority over lower-extremity symptoms or vice versa. Within the homun-
cular representation of the upper extremity, the neurophysiologist must prioritize the
proximal versus distal upper extremity, for tremor that predominates proximal or distally
respectively, depending on which creates the greater disability.
The angle of entry carries equal importance for DBS of the ventral intermediate
nucleus, subthalamic nucleus, and globus pallidus interna. An excessively shallow
placement in the sagittal plane may send the DBS lead tangentially through the long
axis of the ventral intermediate nucleus. Such a trajectory situates the lower contacts
too close to the ventral caudal nucleus of the thalamus, which if stimulated produces
paresthesias that may limit DBS efficacy, and the upper contacts in the ventral lateral
posterior oralis thalamus, where stimulation is less effective (Figure 11.4). Similarly, an
excessively shallow angle of entry in the coronal plane may result in the deepest con-
tacts projecting medially in the head representation, where stimulation increases the
risks of subcortical aphasia, dysarthria, and dysphagia, and the upper contacts reach-
ing the leg representation, stimulation of which is less effective in cases where the most
disabling symptom is upper-extremity tremor (Figure 11.2).
CRITERIA FOR A PHYSIOLOGICALLY DEFINED OR SYMPTOMATICALLY

DEFINED OPTIMAL TARGET
A first criterion is that the DBS lead be placed in the homuncular representation consistent
with the patient’s most incapacitating symptoms. Extremely narrow in the anterior-pos-
terior direction, the ventral intermediate nucleus presents two significant risks: (1) exces-
sively anterior and therefore ineffective DBS lead placement and (2) excessively posterior
DBS lead placement, which produces limited therapeutic effects and intolerable paresthe-
sias. In order to be effective, DBS lead placement must be anterior enough so as to not
produce paresthesias with stimulation intensities that may be necessary for symptomatic
control. The initial intent is to identify the anterior border of the tactile (superficial sensa-
tions) region of the ventral caudal nucleus and placing the DBS lead 2 mm to 3 mm anterior
to the anterior border of the tactile ventral caudal thalamus. Consequently, this author will
target the typical image-guided navigation coordinates relative to the line connecting the
anterior and posterior commissures (AC-PC line) for the ventral intermediate thalamus
but start the initial microelectrode trajectory 2 mm posteriorly to identify the anterior bor-
der of the tactile ventral caudal thalamus more quickly.
This approach resembles the approach to recording in globus pallidus interna DBS,
which involves finding the anterior border of the posterior limb of the internal cap-
sule and placing the DBS lead a fixed distance anterior to the anterior border of the
internal capsule. Critical to the first approach is identifying the anterior border of the
tactile region of the ventral caudal thalamus. One may claim that it is sufficient merely
to find the region of the thalamus responsive to muscle palpation, because this respon-
siveness, presumably mediated by muscle spindles, suggests involvement of the ventral
intermediate thalamus. A problem, however, attends this approach. Relying solely on

muscle palpation responses does not offer sufficient resolution in the anterior-posterior
direction to ensure sufficient distance from the tactile region of the ventral intermedi-
ate thalamus.
The depth of the optimal DBS trajectory also depends on the patient’s age. DBS
leads placed too ventrally (deep) risk stimulating the corticobulbar and corticospi-
nal tract within the posterior limb of the internal capsule and producing tonic muscle
contractions. Though the US Food and Drug Administration (FDA) has not approved
ventral intermediate nucleus DBS in children (where it has for primary dystonia), the
absence of approval should not be construed as disapproval. Indeed, ventral inter-
mediate nucleus DBS in children is a reasonable “off-label” use of an FDA-approved
device although the necessity of thalamic DBS in child age 7 would be exceptionally
rare. In the appropriate situation, ventral intermediate nucleus DBS is also reasonable.
The FDA and most institutional review boards (IRBs), moreover, have provisions for
compassionate use—for ventral intermediate nucleus DBS in children, for example.
For children, consideration of continued skull and body growth is necessary, because
increased growth of the skull could displace the DBS lead, and increased body length
could damage the DBS lead or extension wires.
By 7 years of age, a child’s head circumference has grown to 93% of its anticipated adult
measure (based on the mean). The DBS lead therefore may migrate with continued head
growth. To address this concern, this author would place the distal edge of the second deep-
est (the distal rather than the distal-most contact) at the bottom of the optimal DBS trajec-
tory. With continued growth, the distal-most contact may drift to the optimal position.
Children younger than 7 years have undergone DBS surgery. In these critical situations,
the depth of DBS lead placement must be carefully considered. (Similarly, the neurosur-
geon must use a longer extension wire that connects the DBS lead to the implanted pulse
generator in order to accommodate the body’s continued growth. The child’s body size
determines placement of the DBS stimulator. This author would prefer smaller recharge-
able stimulators, which decrease repeated replacements and the risk of skin erosion).
The ways of minimizing the risk of adverse effects using microstimulation or
macrostimulation are relatively straightforward. Microstimulation, however, conveys
more precise estimates of the distance between the site and the anatomical struc-
ture producing tonic muscle contraction, paresthesias, or some other adverse effect.
Microstimulation of this relatively smaller volume of tissue may produce no adverse
effects, but an absence of adverse effects does not necessarily imply that DBS is inca-
pable of producing them. If microstimulation does produce them, DBS is likely to
produce them as well.
Problems attend producing symptomatic benefit. First, the volume of tissue acti-
vation may be insufficient, particularly if stimulation is done through the microelec-
trode or semi-microelectrode tip or through the indifferent electrical contact in bipolar
microelectrode and semi-microelectrodes. Second, there may occur a microthalamot-
omy effect, in which improvement associated solely with initial physical penetration
of the thalamus masks any benefit to subsequent stimulation. In the case of significant
microthalamotomy effect, the only indication of optimal placement may be from the
microelectrode or semi-microelectrode recordings.
For those intraoperative neurophysiologists who do not use microelectrodes or
semi-microelectrodes but rather macrostimulation through the DBS leads, the criteria are
based primarily on the ability to demonstrate the absence of adverse effects. This has par-
ticular significance for those surgeons who use general anesthesia. Though macrostimu-
lation under anesthesia may produce tonic contraction indicative of spread of stimulation
current to the posterior limb of the internal capsule, a patient will be unable to report
paresthesias associated with excessively posterior DBS lead placement. Avoiding adverse
effects that might limit subsequent therapeutic use of DBS becomes the primary concern.
Even in patients who remain awake during DBS, paresthesias produced by macro-
stimulation cannot provide reliable information about the DBS lead location in the
medial-lateral direction. One cannot know whether the somatotopic representation of
any paresthesias produced is concordant (reflective of the homuncular representation
at the site of stimulation) or discordant in the sense that the paresthesias are related to
stimulation of fibers in passage (Figure 11.4).
There applies an additional criterion: demonstrating a reasonable therapeutic window,
that is, the difference between the threshold at which symptoms improve and that which
produces persistent adverse effects. When testing for adverse effects, this author prefers
using a minimum of 4 volts or 3.5 milliamps, because many patients require higher stimu-
lation voltages or currents and it would be important to know that such higher stimula-
tion strength would be possible should it prove necessary. Also, it is important that there
be a sufficient difference between the minimum threshold necessary to provide a benefit
and the threshold that produces persistent side effects in order to provide those with the
responsibility for postoperative management sufficient “working room.”
In intraoperative DBS testing one cannot merely default to stimulation with the
“average” stimulation parameters typically reported in the literature. Assuming that
the effective stimulation voltage, for example, is normally distributed around a mean
of 3 volts, at least half of patients will require more than 3 volts and thus, failure to
improve with intraoperative DBS test stimulation limited to 3 volts may drive an incor-
rect conclusion. The same concepts apply to testing with respect to production of
adverse effects.
All forms of intraoperative neurophysiological monitoring require that one under-

stand the regional anatomy around the entire electrode trajectory. Microelectrode and
semi-microelectrode recordings require that one study the entire trajectory that begins
approximately 25 mm above the bottom of the ventral intermediate nucleus and that
image-guided targeting had located. Recordings in the dorsal thalamus made above
the ventral intermediate nucleus, for example, aid demonstration of properly operating
recording systems.
This author typically begins the microelectrode recordings approximately 25 mm
above the target based on image-guided navigation. The target is the estimated bottom of
the thalamus. The rationale for beginning the microelectrode or semi-microelectrode tra-
jectory approximately 25 mm above the bottom of ventral intermediate nucleus rests on the
tension obtaining between the need for gathering sufficient information during recordings
and the risk of causing intracerebral hematoma with the sharp microelectrode—whereas
the blunt tip of the stylus housed in the guide cannula tends to push tissue aside as it travels
along the trajectory. Similarly, ending the guide cannula higher in the trajectory increases
the risk that the DBS lead may be deflected from the proper trajectory. A neurophysiologist
may use a longer cannula to introduce macrostimulation through the DBS lead, provided
the cannula remains above the lead’s dorsal-most contact. The longer cannula reduces the
risk of deflection, but it may prove difficult to withdraw while leaving clearance sufficient
for easily fixing the DBS lead in place in the burr hole to the skull.
Determining the regional anatomy depends on the angle of entry in the sagittal and
coronal plane. When the angle lies in the coronal plane, neurophysiologists often select
an entry point that avoids the lateral ventricles. Conventional belief holds that as the
electrode or lead encounters the ependymal surface on exiting the ventricle, the elec-
trode and lead may be deflected from its proper course. In certain cases, avoiding the
lateral ventricle may prove impossible—in patients whose ventricles are enlarged, for
example. Too shallow a trajectory in the coronal plane may result in the electrical con-
tacts of the DBS lead spanning the homuncular representation with the ventral-most
and ventral contacts lying within or too close to the head representation. Such place-
ment might increase the risk of speech, language, and swallowing problems.
The angle of entry in the sagittal plane is critically important. A sufficiently anterior
entry point to the coronal suture avoids traversal of the motor cortex. An excessively
anterior entry point, however, establishes an excessively shallow trajectory angle, which
may place the upper contacts of the DBS lead anterior to the ventral intermediate nucleus
and the more ventral contact posterior and thus risk intolerable paresthesias that would
limit therapy (Figure 11.3). Indication of a trajectory that is too shallow in the sagittal
Leg
Amp
Concordant
Arm
Leg
Amp
Arm
Discordant
FIGURE 11.3 Possible mechanism of discordant and concordant paresthesias in response

to stimulation. In the case of concordant responses, microelectrodes recorded activity
changes correlated with movement of the arm but not with movement of the leg. Thus, there
is relative certainty that the microelectrode is within the arm homuncular representation.
Microstimulation (represented by the spark images) at the same site activates local neurons
(white cartoons of neurons). This activation patients experience as paresthesias of the arm. In
the discordant response, the microelectrode continues to record changes in neuronal activity
with movement of the arm but not with movement of the leg. Hence, the microelectrode is
within the arm homuncular representation. However, microstimulation activates axons that
pass through the site as they project to the leg representation. Stimulation of these axons as
they pass through the arm representation causes paresthesias referred to the leg.
DBS lead
Vc-deep Vim
Vc-tactile
Vop
Sagittal plane
Dorsal
16.0 mm lateral
Posterior Anterior to the AC-PC
line
Ventral
FIGURE 11.4 Schematic representation of a trajectory whose excessive shallowness in the
coronal plane owes to an excessively anterior entry point. The dorsal (upper) contacts lie in
the ventral lateral posterior oralis thalamus, and the ventral (lower) contacts lie in the anterior
ventral caudal thalamus. Because there are few contacts in the DBS target, the ventral inter-
mediate thalamus, there is a significant risk that DBS may be ineffective. Source: Modified from
plane lies in demonstrating that multiple, for example, more than three, divisions of the
thalamus lie in the trajectory. These may include combinations of the ventral lateral pos-
terior oralis, ventral intermediate, anterior (deep sensation) ventral caudal, and poste-
rior (tactile) ventral caudal thalamus. Once demonstrated, the surgeon may be able to
change the angle of trajectory in the sagittal plane. Avoidance of such trajectories is not
available to those intraoperative neurophysiologists who eschew use of microelectrode or
semi-microelectrode recordings.
REGIONAL NEURONAL PHYSIOLOGY
The following describes the neurophysiological characteristics of the various anatomi-

cal structures in the vicinity of the ventral intermediate nucleus from the perspective of
microelectrode recordings. One may generalize these findings to the semi-microelectrode,
understanding that there obtains a crucial difference, namely, that semi-microelectrode
recordings are less capable than microelectrode recordings of isolating individual neu-
rons on which identification of the homunculus rests, depending on the spatial resolution
attendant of the selection of a particular semi-microelectrode.
Dorsal Thalamus
Intraoperative neurophysiologists typically encounter first the dorsal tier of thalamic

nuclei. Few neurons generate the neuronal activity encountered, as evidenced by few
extracellular action potentials of various amplitudes and waveforms at a single recording
site. The recording sites are consequently low-density. Transient are these neurons, which
often discharge at a low rate when first encountered and with low frequency (Figure 11.5).
The intraoperative neurophysiologist must often traverse the trajectory for a few millime-
ters before encountering a site at which she discerns extracellular action potentials (low
density within the trajectory), whose irregular neuronal activity is their signature.
FIGURE 11.5 Representative example of transient neuronal activities. Neuronal activity arises

relatively abruptly, then fades.
Ventral Intermediate Nucleus
Though more persistent than dorsal tier nuclei are the extracellular action potentials of the
ventral teier. The activity at the recording site is low-frequency, that is, it contains few detect-
able extracellular action potentials. The recording site is of low density, meaning that few
different amplitudes are seen. The neurophysiologist may have to traverse a few millimeters
before encountering another site at which she discerns extracellular action potentials (low
density within the trajectory). Sensory stimulation changes neuronal extracellular action
potential activities. The neurons stimulated typically respond to muscle spinal activation,
as demonstrated by responses to muscle palpation, and to joint rotations that stretch the
muscle. The absence of responses to muscle palpation, however, suggests that the neurons
responding to joint rotation do not reside in the ventral intermediate nucleus.
In layers like half an onion, the homuncular somatotopic organization features a
central core that represents the head, an inner layer that represents the upper extrem-
ity, and an outer layer that represents the lower extremity (Figure 11.2). Establishing
the homuncular representation requires that one first traverse the entire ventral inter-
mediate nucleus. Penetrations exclusively in the upper region of the ventral interme-
diate nucleus may record only neurons responding to lower extremity stimulation.
From this may follow the false conclusion that the homuncular representation is con-
fined to the lower extremity. Deeper penetrations likely would encounter neurons
that respond to the upper extremity. Further penetration may reach the homuncular
somatotopic organization’s central core, where neurons respond to jaw opening and
closing and tongue protrusion and retractions, corresponding to the head representa-
tion. When placed in this trajectory, the DBS lead introduces stimulation that may
affect the head representation and increase the risk of subcortical aphasia, dysarthria,
and dysphagia. This risk persists even if the neurophysiologist ceases the microelec-
trode recording upon finding neurons responsive to upper-extremity stimulation.
The ventral intermediate nucleus is the cerebellar relay nucleus, which conveys infor-
mation from the cerebellum to the cortex. Interestingly, in patients with Parkinson’s
disease it appears more effective to stimulate the ventral intermediate nucleus than the
ventral lateral posterior oralis thalamus, which is the basal ganglia relay nucleus.
Deep Sensation (Anterior Ventral Caudal Thalamus)
The anterior ventral caudal thalamus responds to joint capsule receptors. Light touch
or muscle palpation that at the same time does not rotate the joint consequently will not
activate it. The somatotopic homunculus resembles the homunculus associated with
the ventral intermediate nucleus. The extracellular action potentials are low frequency,
irregular, low density within the site, and low density within the trajectory. The anterior
deep sensation ventral caudal thalamus is posterior to the ventral intermediate nucleus
and anterior to the tactile region of the ventral caudal thalamus. At relatively higher
currents microstimulation in this region may produce paresthesias. Again, the distri-
bution of paresthesias does not predict the homuncular representation, because the
paresthesias may be concordant or discordant (Figure 11.4).
Tactile Sensation (Posterior Ventral Caudal Thalamus)
The posterior ventral caudal thalamus responds to light touch but can be unintentionally
activated during joint rotation and muscle palpation if the receptive field of the posterior
ventral caudal thalamic neuron is accidentally activated. The somatotopic homuncu-
lus is similar to the homunculus associated with the ventral intermediate nucleus. The
extracellular action potentials are low frequency, irregular, low density within the site,
and low density within the trajectory. The posterior tactile ventral caudal thalamus is
posterior to the deep sensation ventral caudal nucleus. At relatively low currents micro-
stimulation in the tactile ventral caudal thalamus may produce paresthesias. Again, the
distribution of the paresthesias does not predict the homuncular representation, because
the paresthesias may be concordant or discordant (Figure 11.4).
Ventral Lateral Posterior Oralis Thalamus
Immediately anterior to the ventral intermediate nucleus lies the ventral lateral poste-
rior oralis thalamus. More persistent than the ventral intermediate nucleus, extracel-
lular action potentials generate low- to moderate-frequency activity at a recording site
of few detectable extracellular action potentials of various amplitudes (low density
within the recording site). The neurophysiologist may have to traverse a few millimeters
(fewer, however, than those she must traverse in the ventral intermediate nucleus) before
encountering another site in which extracellular action potentials can be discerned (low
density within the trajectory). Neuronal extracellular action potential alters activities
with joint receptor sensory stimulation. These neurons appear to respond more read-
ily with voluntary movements initiated by the patient (active rather than passive joint
rotations)—a good reason for having the patient remain awake during recording.
This region contains few if any neurons. Neuronal activities in the form of extracellular
action potentials within the axons generally are indiscernible. Stimulation in or near
the posterior limb of the internal capsule may activate the corticobulbar and cortico-
spinal pathways and thus produce muscle contraction. In the case of stimulation via the
microelectrode or semi-microelectrode, however, the volume of tissue activation may
be insufficient to produce an observable response. Consequently, the lack of a response
is not evidence that the trajectory is not in the posterior limb of the internal capsule.
RECONSTRUCTING THE REGIONAL ANATOMY BASED ON THE PHYSIOLOGY
It is critical to reconstruct the regional anatomy in the vicinity of the electrode trajectories
in order to have the highest probability of optimal placement. Reasons include: (1) imag-
ing cannot differentiate the ventral intermediate nucleus of the thalamus from other
ventral tier thalamic nuclei, (2) imaging cannot differentiate the homuncular repre-
sentation, (3) intraoperative brain shift risks making targeting based on preoperative
imaging very problematic, and (4) modest brain shifts can be accommodated by elec-
trophysiological mapping. However, image-guided navigation, particularly the “probe’s
view” can be every helpful in informing the intraoperative neurophysiologist as to what
she may expect. Consequently, the intraoperative neurophysiologist is advantaged by
participating in or, at the least, reviewing the image-guided navigation.
Reconstructing the regional anatomy around the trajectory according to neurophys-
iological findings relies on recordings of neuronal characteristics, of which this author
favors microelectrode recordings of individual neuronal action potentials, and stimu-
lation, whether microstimulation or macrostimulation. Macrostimulation to establish
the regional anatomy is made problematic by reduced spatial resolution and the higher
probability of producing a microthalamotomy effect that would obscure observations
of efficacy for localization. Microstimulation, for the purposes of identifying functional
anatomy, is best done using principles to increase the probability of driving a physiologi-
cal response. This author uses high frequencies—300 pps, for example—during micro-
electrode stimulation to take advantage of temporal summation, which helps to ensure
effective synaptic transmission initiated by stimulation across multiple synapses.
Intraoperative neurophysiologists can visualize a patient’s regional anatomy by call-
ing to mind representations from anatomy atlases. These remembered representations
they then morph and rotate in their minds until they match the electrophysiological
findings. Neurophysiologists can acquire this skill by frequently consulting stereo-
tactic human atlases and by attempting to predict the electrophysiological findings
as a trial-and-error method of learning. This author has created a three-dimensional
atlas by tracing lines from an atlas onto transparencies, which he mounted on stack-
able Plexiglas panes in order to represent the anatomies. These allowed him to rotate
the three-dimensional images in order to map electrophysiological results on a specific

anatomical location and orientation. After using these representations for a time, this
author became able to visualize them in his mind.
Some intraoperative neurophysiologists find such mental imaging, morphing, and
rotations difficult, finding it easier to use language syntax or some other set of rules
instead. Descriptions of various trajectories may supply these rules. Yet deducing these
rules proves difficult, because the number of possible trajectories is extremely large.
Syntax in the form of a decision tree greatly mitigates this difficulty (Appendix B).
The decision tree rests on a logic that divides the microelectrode trajectory into an
upper segment and lower segment. In the thalamus, the upper trajectory corresponds
to the latter’s dorsal tier and the lower trajectory corresponds to that portion of the ven-
tral tier that contains the ventral intermediate thalamus, the target of DBS, the tactile
region of the ventral caudal thalamus (to be avoided), and the anterior ventral caudal
and ventral lateral posterior oralis thalami.
The trajectory’s angle may be such that the DBS lead traveling along it passes
through several ventral tier thalamic nuclei. The lower trajectory is thus divided
into three segments—initial, middle, and bottom. Changes in the characteristics of
encountered neuronal activities determine these segments’ scales, which are more-
over relative to each other. Lacking in the upper trajectory are sensorimotor drive and
any unique discharge characteristic save that of transient, low frequency, low density
within the recording site, and low density within the upper trajectory. Importantly,
recording in the dorsal tier demonstrates proper functioning of microelectrode and
semi-microelectrode systems.
The algorithms provided in Appendix B are generalizations and approximations;
actual physiological and anatomical correlates may vary greatly, depending on the spa-
tial orientation of the trajectories to the patient’s specific anatomy. Though a patient’s
regional anatomy correlates closely to the line connecting the anterior commissure
to the posterior commissure (AC-PC line), the specific orientation of the AC-PC line
to the stereotactic apparatus may vary greatly according to the six degrees of spatial
freedom (translation in the three axes and rotations about the three axes in Cartesian
space). The responsibility of the use of this heuristic thus falls on the physician and
healthcare professional, who must exercise her own judgment in caring for her patient.
SUMMARY
Unlike the case of DBS for the subthalamic nucleus, targeting in the ventral intermedi-
ate thalamus depends on identifying the structure the neurophysiologist must avoid
stimulating, namely, the tactile region of the ventral caudal thalamus. Any stimulation
current that spreads to the ventral caudal thalamus may produce intolerable paresthe-
sias and consequently limit DBS’s effectiveness. The intraoperative neurophysiologist
must identify the anterior border of the tactile region of the ventral caudal thalamus.
Merely locating neurons responsive to muscle palpation proves insufficient, as the tra-
jectory may be posterior such that the tactile region of the ventral caudal thalamus is
stimulated, thus complicating postoperative DBS therapy.
The narrow length of the ventral intermediate thalamus in the anterior-posterior
direction poses a challenge. Excessively shallow trajectories in the sagittal plane risk
the trajectory traversing the short axis, thus risking the ventral-most and ventral con-
tacts too posterior and too close to the tactile region of the ventral caudal thalamus. At
the same time the dorsal or dorsal-most contacts may be in the ventral lateral poste-
rior oralis thalamus, stimulation of which may be less effective. Posing a further chal-
lenge—particularly if the trajectory pursues too shallow an angle in the coronal plane,
as might happen should the surgeon attempt to avoid a large lateral ventricle—is the
relatively long length in the medial-lateral direction relative to its corresponding com-
plex homuncular representation. Thus, the DBS may lead to stimulation of the head
region, as it is medial, and thereby increase the risk of speech, language, and swallowing
difficulties.
REFERENCES
Hassler R: Architectonic organization of the thalamic nuclei. Atlas for Stereotaxy of the Human Brain.
G. Schaltenbrand and W. Wahren. Stuttgart, Thieme, 1977.
/ / / 12 / / / CLINICAL
ASSESSMENTS
DURING INTRAOPERATIVE
NEUROPHYSIOLOGICAL
MONITORING
INTRODUCTION
The importance of monitoring clinical response to stimulation during intraoperative

neurophysiological monitoring becomes particularly evident during intraoperative
testing by use of the Deep Brain Stimulation (DBS) lead. Evaluating improvement in
symptoms is important, as is evaluating complications or adverse effects. Intraoperative
clinical assessments consist of two components: (1) appropriate evaluation of the
treated disease’s symptoms with a view to assuring postoperative therapeutic effective-
ness and (2) clinical assessment of the stimulation’s potential adverse effects (these
typically indicate that the stimulation current has spread to unintended structures).
By considering the nature of potential effects adverse to subsequent DBS effectiveness,
one may gain insight into anatomy around the stimulation site. This insight aids one
in determining whether the DBS lead occupies a reasonable position. It also helps one
to identify, for the purpose of postoperative DBS programming, those contacts which
must be avoided because they are likeliest to produce side effects.
The quality of clinical examination determines the quality of the information gained
from intraoperative clinical assessments. Some symptoms of the disorders treated may
respond to stimulation in the operating room, albeit in a manner that fails to predict
subsequent beneficial or adverse effects. Propagation of the stimulation current to the
corticospinal tract, for example, may reduce tremor and thus create the impression of
symptomatic benefit. The involvement of the corticospinal tract, however, limits ben-
efit, because it is likely to interfere with normal use of the limb. One must establish
201
that tremor reduction owes to the therapeutic mechanisms, known or unknown, rather
than involvement of the corticospinal tract. In order to do so, one must distinguish the
first sort of reduction from the second.
CLINICAL ASSESSMENTS BY DISEASE
This section reviews the clinical assessments of the symptoms of diseases targeted
for DBS before proceeding to a discussion of the clinical assessments specific to the
DBS anatomical or physiological target that is particularly relevant to assessment of
adverse effects. DBS of more than one target improves symptoms of various disorders.
Stimulating both the globus pallidus interna and the subthalamic nucleus relieves a
wide range of symptoms related to Parkinson’s disease. Conventional wisdom holds
that DBS of the thalamus is effective solely for tremor. This wisdom rests, however,
on findings from studies in which subjects were selected primarily according to the
presence of tremor and comparatively less attention was paid to bradykinesia and other
motor symptoms. Whether ventral intermediate thalamic DBS would help with symp-
toms other than tremor therefore remains unknown.
Clinical Evaluation of Parkinson’s Disease
Deep Brain Stimulation fulfills its primary purpose when it relieves the symptoms and
disabilities characteristic of Parkinson’s disease without producing adverse effects.
Numerous and varied, symptoms and signs of Parkinson’s disease include motor symp-
toms—slow movement and others—and nonmotor symptoms: depression, cognitive
decline, and so on. Though current use of DBS is directed primarily toward alleviating
motor symptoms, a physician may detect, during routine DBS testing in the operating
room, other potential nonmotor complications: depression and impulse control problems.
The primary motor symptoms of Parkinson’s disease include bradykinesia (slow-
ness of movement), akinesia (absence of movement), tremor, rigidity (resistance to
passive joint rotations), and postural and gait abnormalities. These last do not lend
themselves to testing in the operating room. As the primary purpose of microelectrode
and semi-microelectrode recordings is to establish the regional anatomy around the
trajectory and because the volume of tissue activated with microstimulation may be
insufficient to produce an improvement in the motoric symptoms, the responses to
microstimulation tend to be assessed qualitatively (Appendix D and E). Intraoperative
test stimulation through the DBS lead primarily is to determine efficacy and tolerance.
With respect to efficacy, quantitative assessments are performed. A physician may
12. Clinical Assessments During Intraoperative Neurophysiological Monitoring / / 203
quantify these scores by use of the Movement Disorders Society-Unified Parkinson’s

Disease Rating Scales Part III (Goetz et al. 2008). Clinicians may record scores by use
of the form available in Appendix G and on the Greenville Neuromodulation Center
website (www.grnneuromod.com).
Bradykinesia may affect limb movement, speech, respiration, and nearly every move-
ment. Movements tested to assess bradykinesia in the operating room include rapid
repetitive finger tapping—the index finger on the thumb, for example. The physician or
intraoperative neurophysiologist notes this activity’s frequency (speed) and amplitude
and determines whether there is a reduction in amplitude or halting of the movement
as the effort continues. Unique to Parkinson’s disease is the finger-tapping amplitude,
which tends to decrease as it continues. This decrease in amplitude aids the intraopera-
tive neurophysiologist in distinguishing bradykinesia owing to Parkinson’s disease from
similar symptoms owing to inadvertent stimulation of the corticospinal tract.
Two phenomena—the relative lack of finger-tapping amplitude reduction with cor-
ticospinal tract stimulation, and frank muscle contractions with progressive increase
the stimulation voltage or current—may allow the intraoperative neurophysiolo-
gist to identify slowness of movement as owing to Parkinson’s disease. Patients with
Parkinson’s disease also make normal individuated finger movements. A intraopera-
tive neurophysiologist tests for this by having the patient tap in succession each finger
to the thumb. Patients with Parkinson’s disease move independently all fingers except
the ulnar-most two. Corticospinal tract involvement diminishes the independence of
the finger movements: All fingers tend to move together (Figure 12.1). However, this
observation needs further testing.
Bradykinetic finger-tapping quantities correspond to the following scale: a grade of
0 = normal; a grade of 1 = slight slowing with disintegration of the normal rhythm (one or
two possible interruptions or diminishing amplitude after approximately 10 taps); a grade
of 2 = three to five interruptions or prolonged arrest, mild slowing, or amplitude reduction
midway in a 10-tap sequence; a grade of 3 = more than five interruptions or at least one pro-
longed arrest, moderate slowing, or decrease amplitude following the commencement of
tapping; a grade of 4 = finger tapping proves extremely difficult or impossible to perform.
One may assess bradykinesia in a hand-opening and -closing task according to a
similar rating scale, which consists of the following values: a grade of 0 = normal; a
grade of 1 = slight slowing with disintegration of a normal rhythm (one or two interrup-
tions or diminishing amplitude as the task approaches completion); a grade of 2 = three
to five interruptions during tapping, prolonged arrest, mild slowing, or amplitude
reduction mid-task; a grade of 3 = more than five interruptions, at least one prolonged
arrest, moderate slowing, or decreased amplitude immediately following the task’s
Bradykinesia Corticospinal tract stimulation

individuated movements indiviuated movements
retained lost
FIGURE 12.1 Comparison of effects on finger tapping by Parkinson’s disease and stimulation

of the corticospinal tract, respectively. Finger tapping performed by patients with Parkinson’s
disease may be slow in frequency and diminished in amplitude. Finger tapping by patients
experiencing stimulation of the corticospinal tract may also be slow in frequency and dimin-
ished in amplitude. Patients with Parkinson’s disease, however, are able to make individu-
ated movements such that the each finger moves relatively independently, whereas patients
experiencing stimulation of the corticospinal tract are only able to move their fingers together.
commencement; a grade of 4 = finger tapping proves extremely difficult or impossible

for the patient to perform.
The importance of testing both finger tapping and hand opening and closing owes to
different degrees of sensitivity to DBS. Experience has shown this author that though hand
opening and closing is more responsive to DBS, finger-tapping response more effectively
predicts subsequent therapeutic effect. (This observation requires additional study.)
Typical of Parkinson’s disease, tremor during rest is assessed intraoperatively as
the patient lies still and has her upper extremities supported against gravity. Tremor
amplitude determines ratings that correspond to the following scale: a grade of 0 = an
absence of tremor; a grade of 1 = an amplitude less than 1 cm; a grade of 2 = an ampli-
tude greater than 1 cm but less than 3 cm; a grade of 3 = an amplitude greater than 3 cm
but less than or equal to 10 cm; a grade of 4 = an amplitude greater than 10 cm.
Patients displaying little tremor while at rest may display pronounced tremor
when holding their extremity in a particular position. Intraoperative assessment in
such instances involves testing DBS effects on the postural tremor resulting from the
patient’s assuming that position. Rating proceeds according to the same grading scale
as that listed for resting tremor.
The examiner assesses rigidity by noting the amount of resistance she encounters when
rotating various joints on the patient’s body. She typically holds the patient’s hand as she
rotates the patient’s wrist and simultaneously pronates and supinates the forearm and flexes
Assessing rigidity
FIGURE 12.2 Schematic representation of assessing muscle tone in the upper extremity in the
clinic, which can be applied in the operating room. The examiner holds the patient’s hand and
rotates the wrist, pronates and supinates the forearm, and flexes and extends at the elbow
simultaneously. The rotations are not rhythmic in order to avoid the patient’s anticipation of
the rotations and thereby assisting the examiner.
and extends the elbow (Figure 12.2). The examiner must take care that the rotations do not
follow the same pattern. Rather, she must make these rotations as varied and random as
possible. Should the patient anticipate the rotations, the patient may unknowingly begin to
assist the examiner’s manipulations, resulting in a misleading reduction in resistance. The
reader may discover this effect of anticipation for herself by rotating the limb of a person
with normal tone in the manner described above. The degree of resistance encountered in
this instance rates a 0 according to the Movement Disorders Society–Unified Parkinson’s
Disease Rating Scales Part III. The reader may then rotate her own arm, using her other
arm to do so. She will notice that her own arm resists far less than another person’s, because,
unbeknownst to her, the rotated arm yields to the arm rotating it. The examiner might thus
be tempted to rate her own resistance -1. To avoid this, normal rigidity scores a 1 in order
to reserve 0 for lower tone. This is important because intraoperative DBS may reduce resis-
tance to a level less than that which is normally encountered in normal subjects.
A 0 rating goes to any patient mustering resistance similar to that which the examiner
felt when rotating her own limb that is less than normal. A score of 1 indicates that the
patient offers normal resistance. A score of 2 indicates that the patient offers increased
resistance but remains capable of a full range of motion. A score of 3 indicates that the
patient offers greater resistance and remains capable of a full range of motion, albeit at
considerably greater effort. A score of 4 indicates that the patient is incapable of a full
range of motion.
Clinical Evaluation of Essential Tremor and Cerebellar Outflow Tremor
The examiner assesses tremor under specific conditions. These conditions include
the patient’s remaining at rest, maintaining a particular posture, touching a finger to
her chin, and lifting a cup to her lips (Figure 12.3). Though bringing a cup to the lips
may prove difficult, patients fitted with a stereotactic frame are nonetheless asked to
perform this task, known as the cup task. Magnitudes of tremor under each condition
correspond to the following rating scale: a grade of 0 = absence of tremor; a grade of
1 = tremor amplitude of 1 cm or less; a grade of 2 = tremor amplitude of more than 1 cm
but less than 4 cm; a grade of 3 = tremor amplitude of more than 4 cm that does not
prevent the patient from completing the task; a grade of 4 = tremor of such pronounced
amplitude that it prevents the patient from performing the task. The examiner may hold
a measuring device or her index finger 4 cm apart from her thumb as a way of measuring
tremor amplitude (Figure 12.4).
The examiner assesses resting tremor by having the patient lie quietly with the
extremities supported against gravity. The examiner assesses postural tremor by having
the patient hold the upper extremity outstretched. The examiner assesses action tremor
as the patient first holds the upper extremity outstretched then touches the index finger
on the hand of the outstretched extremity to the chin. (Touching the chin as opposed
to the nose allows the patient to avoid affecting the stereotactic frame or approaching
the surgical field.) The patient performs this task several times. In the cup task, the
patient reaches for a cup—a disposable specimen kind such as one finds in an operating
room—held out to her by the examiner. The patient takes the cup from the examiner
and brings it to her lips as if to drink from it. Diagnosticians may record scores by use
of the form available in Appendix H and on the Greenville Neuromodulation Center
website (www.grnneuromod.com).
The measures of tremor appear to vary in terms of their degree of sensitivity to
DBS. Rest tremor appears the most responsive, and cup task–associated tremor the
least. Though it is important to test rest tremor in order to discern some effect of DBS,
Rest Postural
A B C
Finger-
Nose
Cup
FIGURE 12.3 Demonstration in the clinic for the types of tremor that may be assessed in the
operating room. Note on the finger-nose task, the examiner should vary finger target posi-
tions during the performance (different target positions in A and C). Patients in a stereotactic
frame may not be able to bring the cup to the lips but should try to approximate the lips. Care
must be observed least the patient contaminate a sterile field.
FIGURE 12.4 Demonstration in the clinic of how to quantitate the amplitude of tremor, in this
case postural tremor. On the panel to the left a centimeter ruler can be used, this author finds
a transparent ruler most helpful. Alternatively, as demonstrated in the panel to the right, the
examiner can space his fingers to a specific distance with which to compare to the tremor
amplitude.
the examiner must bear in mind that reduction of cup task–associated tremor indi-
cates probably the most significant functional improvement. This author’s experience
includes patients whose intraoperative response to DBS testing was disappointing but
whose condition improved with postoperative management.
Cerebellar outflow tremors, such as those owing to multiple sclerosis, present a chal-
lenge. In this situation DBS would be considered a standard and accepted “off-label”
use of an FDA-approved device. It appears that tremor in distal-most musculature
responds better to DBS than does tremor in proximal musculature.
As it is for Parkinson’s disease–associated tremor, it is important to distinguish
between tremor suppression attributable to the therapeutic mechanisms of action of
DBS and suppression attributable to stimulation current’s having spread to the cortico-
spinal tract (see discussion above).
Clinical Evaluation of Dystonia
A host of challenges greet the examiner when she attempts to evaluate clinical response
to intraoperative DBS. Many dystonic symptoms may fail to respond to intraoperative
DBS, because their improvement often requires months of stimulation. Tremor and
some other phasic symptoms of dystonia, however, may demonstrate an acute response
to intraoperative DBS. In such cases, an examiner may evaluate the tremor according
to the ratings described for Essential tremor and cerebellar outflow tremor. In other
instances, a phasic dystonic symptom may appear as hyperkinesias, as it does in chorea.
An examiner may evaluate these symptoms according to the ratings for hyperkinetic
syndromes listed below.
An examiner may assess persistent posturing, a static feature of dystonia, by esti-

mating, as a percentage of the normal range of motion in the direction of the depar-
ture from neutral position at rest due to dystonia (Figure 12.5). As the normal upper
extremity remains at rest, the wrist is extended to approximately 10 degrees and has
a normal range of motion of 70 degrees in extension and 75 degrees in flexion. The
metacarpal—phalangeal joints are flexed to an angle of approximately 30 degrees
and have a normal range of motion of 45 and 90 degrees, in extension and flexion
respectively. The proximal interphalangeal are flexed to an angle of approximately 30
degrees and has a normal range of motion of 70 and 175 degrees, in extension and flex-
ion respectively. The wrist is pronated to an angle between 30 degrees and 60 degrees
and has a normal range of motion of 70 degrees. The elbow is flexed approximately 30
degrees from full extension and has a normal range of motion of 70 and 80 degrees in
pronation and supination, respectively. Neutral position (while supine) in the lower
extremity is typically at 180 degrees at the hip and has a normal range of motion of
100 degrees. Neutral position (while supine) of the knee is 180 degrees and has a nor-
mal range of motion of 150 degrees. The ankle is laterally rotated to approximately
30 degrees from the vertical and has a normal range of motion of 30 and 45 degrees
in extension and flexion respectively. According to this author’s definition, neutral
Torticollis 100 Score as % or range of motion
Normal range
of motion
Normal or neutral Deviation from

normal
A B C
FIGURE 12.5 Schematic representation of estimating the degree of departure from neutral

position in torticollis in cervical dystonia. First, the neutral position must be determined,
which in this case is when the head is straight ahead (A). The patient’s position at rest (the
patient not attempting to straighten the head) is determined. The angle of departure is divided
by the normal range of motion in the direction the head is turned. In example B, the deviation
is slightly more than 50%, resulting in a grade of 3. In example C, the deviation is more than
75%, resulting in a grade of 4.
position of the head is directed and extended such that the eyes are centered in the
orbit in order to direct the gaze straight ahead. The normal range of motion for lateral
rotation is 80 degrees in each direction. For neck flexion and extension the range of
motion is 50 and 60 degrees respectively. For neck lateral flexion is 30 degrees in each
direction. Schematic examples are shown in Figure 12.5, Figure 12.6, and Figure 12.7.
Translating the degrees of departure from neutral position is as follows: 0 = no depar-
ture; 1 = < 25%; 2 = between 25% and 50%; 3 = > 50% but ≤ 75%; and 4 = >75%. In
the operating room, the examiner can observe her own neutral position and range of
motion with which to compare with the patient by imitating the patient’s position and
subsequently estimating the grade of departure from normal (assuming that the exam-
iner is normal). Intraoperative neurophysiologists and clinicians may record scores
by use of the form available in Appendix I and on the Greenville Neuromodulation
Center website (www.grnneuromod.com).
Stimulation of the corticospinal tract may affect the assessment of dystonia. The
examiner must distinguish the former from the latter if the latter appears to have wors-
ened. Stimulation of the corticospinal tract is evident when frank tonic contraction of
muscles originally unaffected by the dystonia results from increasing the stimulation
voltage or current. In the event that the tonic posturing owes to stimulation of the pos-
terior limb of the internal capsule, reexamination with stimulation switched off ought
to improve it.
Wrist 100 Score as % or range of motion
Normal range
of motion
Normal or neutral
A
Deviation from
normal
FIGURE 12.6 Schematic representation of estimating the degree of departure from neutral

position of the wrist in dystonia. First, the neutral position must be determined, which in this
case is when the wrist is at approximately 10 degrees of extension. The patient’s position at
rest (the patient not attempting to straighten the wrist) is determined. The angle of departure
is divided by the normal range of motion in the direction the wrist is turned. In example A, the
deviation is slightly less than 50%, resulting in a grade of 2. In example B, the deviation is more
than 75%, resulting in a grade of 4.
A B C D E
F G
45°
20°
FIGURE 12.7 Schematic representation of range of motion of the ankle. The normal range of
ankle inversion and eversion are shown in A and B with the angles estimated in C and D. If the
patient’s neutral position was as illustrated in E, the deviation from neutral clearly is greater
than 75%, resulting in a grade of 4. F and G show the normal range of motion with ankle plan-
tarflexion and dorsiflexion, respectively.
Clinical Evaluation of Hyperkinetic Disorders
Characteristic of hyperkinetic disorders are the following four features: (1) the portion

of the body affected, (2) the speed of the involuntary movement, (3) the coarseness of
the movement, and (4) the degree of stereotype. For example, athethosis, which tends
to affect more distal musculature, is more graceful; whereas chorea, which tends to
affect proximal musculature, is jerkier.
For purposes of intraoperatively assessing response to test stimulation, more impor-
tant than the above mentioned precise characteristics is the degree of change in the char-
acteristics with stimulation. This author considers changes in amplitude and frequency
of involuntary movements to be most important. Again, the range of motion through
joints allows one to estimate the degree of movements. Most examiners, however, rely
on a qualitative scale ranging from 0 (no involuntary movements) to 4 (the worst possi-
ble, or maximal involuntary movements). Though this qualitative assessment is reason-
able, its scale lacks anchoring as well as fixed intervals, the critical issue being the degree
of change under two conditions: stimulation and no stimulation. Examiners using the
qualitative scale must take care that they maintain their subjective reference points
(anchors) and intervals. As discussed above, the effects of inadvertent stimulation of the
corticospinal tract need to be assessed.
This author uses the following scale: 0 = no dyskinesia; 1 = slight dyskinesia, which
does not interfere with the patient’s attempting a volitional task that utilizes the joint
rotations involved by the dyskinesia; 2 = the dyskinesia does interfere but can be over-
come without great difficulty; 3 = the dyskinesia does interfere but can be overcome
with great difficulty; 4 = the dyskinesia does interfere and cannot be overcome.
Diagnosticians may record scores by use of the form available in Appendix K and on
the Greenville Neuromodulation Center website (www.grnneuromod.com).
Clinical Evaluation of Tic Disorders
Deep Brain Stimulation for tic disorders such as Tourette’s syndrome is considered a
standard and accepted “off-label” use of an FDA-approved device. These disorders are
assessed according to the tics’ frequency and location. A patient may have multiple tics.
There are multiple rating scales for tic disorders such as Tourette’s syndrome. Most are
retrospective and developed primarily to assess the degree to which the symptoms
interfere with the quality of life. Generally, they are not suited to the operating room.
This author suggests rating based on the rapidity of the tics such that 0 = no tics; 1 = <3
tics per minute; 2 = 3–6 tics per minute; 3 = 7–10 tics per minute; and 4 = > 10 tics per
minute. intraoperative neurophysiologists and clinicians may record scores by use of
the form available in Appendix J and on the Greenville Neuromodulation Center web-
site (www.grnneuromod.com).
CLINICAL ASSESSMENT OF CORTICOSPINAL AND

CORTICOBULBAR STIMULATION
The corticospinal and corticobulbar tracts, within the posterior limb of the internal
capsule, may be stimulated as a consequence of DBS of the globus pallidus interna, the
thalamus, and the subthalamic nucleus. The effects of stimulation of the corticospinal
and corticobulbar tracts will therefore be discussed independently.
The corticobulbar fibers innervate the lower motor neurons in the brainstem that
subsequently innervate muscles of the face, the tongue and pharynx, and the extra-
ocular muscles of the eyes. Typically, stimulation of the corticobulbar fibers effects a
conjugate, rather than disconjugate gaze owing to involvement of the frontopontine
fibers that originate in the frontal eye fields and descend to the centers in the pons for
conjugate horizontal eye movements (Figure 12.8).
Though the nature and severity of the symptoms produced is related to the inten-
sity of stimulation, the mechanisms’ nonlinearity complicates the relationship. One
of the earliest symptoms described by patients is a “funny feeling,” which is localized
somewhere on the body. Though many patients report the sensation as one of being
pulled, others are unable to describe it in such terms. Care must therefore be taken to
assure that the “funny feelings” do not owe to the medial lemniscus’s involvement in
FIGURE 12.8 Schematic representation of possible effects of stimulation on eye movements.

Figure A demonstrates disconjugate gaze. The patient is instructed to look at the examiner’s
finger. As can be seen, one eye is able to do so, but the other eye is being pulled by the medial
rectus muscle due to spread of stimulation to the fascicles of the oculomotor nerve. In B,
again, the patient is instructed to look to the examiner’s finger, but now both eyes are devi-
ated away owing to stimulation of the frontopontine fibers that descend in the corticobulbar
tract. Figures C and D demonstrate testing for disconjugate gaze by examining the reflection
of a focused beam of light, such as from a flashlight. In C, the reflection falls on the same point
(homologous) on the iris consistent with normal gaze. In figure D, the reflection falls on non-
homologous points suggesting disconjugate gaze.
the stimulation in the case of subthalamic nucleus DBS, or the posterior or tactile ven-
tral caudal thalamus in the case of ventral intermediate nucleus DBS.
Increasing stimulus intensity allows one to distinguish a “funny feeling” owing to
corticobulbar or corticospinal tract involvement from a funny feeling owing to involve-
ment of the medial lemniscus or posterior tactile ventral caudal thalamus. At some
point contraction of the muscle may become evident; thus, it is important to increase
the intensity within reasonable limits to assure that the “funny feeling” has neither a
corticobulbar nor corticospinal origin.
An early sign of underlying muscle contraction, “dimpling” of the skin surface aids
in distinguishing muscle-pulling from corticobulbar or corticospinal tract activation
from dystonia, because, at relatively lower intensities, not all the muscles contract in
response to stimulation of the internal capsule’s posterior limb. By shining a beam of
light across the patient’s skin surface and looking for shadows, one may observe such
dimpling (Figure 12.9).
Reduced tremor and worsening dexterity also indicate corticobulbar or cortico-
spinal tract involvement prior to demonstrating frank tonic contractions. The latter is
most apparent in finger movements and speech. As discussed above, the loss of manual
dexterity manifests as a reduction in the individuated finger movements (Figure 12.1).
Speech may be slurred. Patients are often best able to judge the quality of their speech,
and thus they ought to be asked whether it sounds normal to them (speech and lan-
guage is discussed in greater detail below).
FIGURE 12.9 A subtle sign of stimulation of the corticobulbar or corticospinal tracts may be

contraction of small groups of muscle fibers causing a dimpling of the skin that may be diffi-
cult to see, particularly if a diffuse light is shown on the muscle (A). The dimpling may be seen
if a focused beam of light is shown tangentially on the surface of the muscle (B).
Assessing patients with facial dystonia for facial contractions owing to stimulation
of the corticobulbar tract within the posterior limb of the internal capsule may prove
difficult. (Patients with Meige’s syndrome are notable examples.) Stimulation appears
to affect sufficiently muscles on one side of the patient’s face. Facial dystonias, on the
other hand, often affect muscles bilaterally.
CLINICAL ASSESSMENT OF SPEECH AND LANGUAGE
There are a number of ways in which speech may be affected by DBS, and consequently
each must be assessed through careful examination. Effects include aphonia, inability to
produce any sound; word-finding difficulty; changes in prosody, the melodic changes in
intonation; and dysarthria, slurring of words. It is important to have the patient speak
phrases that are not routinely used, such as the patient’s name and to use phrases that
stress a range of speech articulators. This author has patients say “Today is a lovely day,”
“British constitution,” and “Methodist episcopal.” Patients often are better judges of
effects on speech, so asking the patient’s impressions of her speech is helpful.
Stimulation of the posterior limb of the internal capsule may affect speech and lan-
guage. Speech abnormalities may owe to the stimulation current propagation to the
corticobulbar fibers within the posterior limb of the internal capsule or they may owe
to other mechanisms intrinsic or extrinsic to the targeted structure. Continuing to
increase the stimulus intensities to a reasonable extent may demonstrate observable
muscle contractions of the face and head.
In the case where speech and language involvement is not associated with frank
facial muscle contractions at reasonable stimulation intensities, speech and language
involvement may relate to stimulation of the head homuncular representation within
the ventral intermediate thalamus or globus pallidus interna (although speech and
language functions typically are less affected by DBS of the globus pallidus interna).
For example, moving the DBS lead anteriorly or medially may improve the speech and
language in the case of globus pallidus interna DBS. However, moving the lead laterally
may be helpful in the case of ventral intermediate nucleus DBS. The subthalamic nucle-
us’s relatively small size also makes lead repositioning to alleviate effects on speech and
language difficult. There are cases in which it is difficult to attribute speech and lan-
guage involvement to a specific position. In such cases, lead repositioning fails to avoid
effects on speech and language.
Bilateral DBS occurrence may significantly increase the risk of speech and language
impairment. In the operating room, this risk may go overlooked, particularly during test
stimulation of the newly implanted DBS lead while the already implanted pulse genera-
tor is turned off. The risk may only become apparent with use of both DBS systems in
the postoperative clinic.
CLINICAL EVALUATION OF DBS TARGET
Bearing on the matter are two extremely important issues. First, one must establish
a clinical response’s exact nature. As discussed above, reduction of tremor with stim-
ulus propagation to the corticospinal tract within the posterior limb of the internal
capsule should not be construed as a therapeutic effect. However, this recommenda-
tion is somewhat problematic. It is indeed possible, even probable, that antidromic
activation of cortical projections to, or in the vicinity of, the subthalamic nucleus may
play a therapeutic role. Yet it may be a matter of degree. This author proposed that
antidromic activation of cortical projections to the vicinity may be therapeutic, how-
ever, with less than 10% of the DBS pulses resulting in an antidromic action poten-
tial. Antidromic activation through branch points may be a major source of failure to
propagate the antidromic action potential. It is not clear that cortical neurons project-
ing to the vicinity of the subthalamic nucleus also project to the lower motor neurons.
Activation of corticobulbar and corticospinal tract fibers may be more efficient, lead-
ing to tetanic activations of lower motor neurons.
Second, one must consider electrode configuration—the combination of active
anodes (positive contacts) and cathodes (negative contacts)—and stimulation param-
eters: frequency, pulse width, and voltage or current. As discussed in chapters 9–11, test-
ing with the typical or average settings may potentially lead to problems postoperatively.
Some patients may require higher stimulation intensities for benefit, and a subset of
these may experience adverse effects at stimulation intensities lower than that required
for benefit but higher than the average. One must therefore see to it that stimulation
occurs at the highest possible setting that may reasonably be encountered postopera-
tively. This will vary according to the DBS target selected and the disorder treated.
A critical concept, the therapeutic window denotes the difference between

the stimulation intensities that produce benefit—which, depending on whether a
constant-voltage pulse generator or constant-current pulse generator is in use, is mea-
sured in terms of voltage or current—and the stimulation intensities that produce
sustained adverse effects. The postoperative DBS programmer must work within this
window. When there arises the need for a beneficial stimulation intensity greater than
that which produces adverse effects, an impracticable situation results. One would like
a wide separation between the lower intensities that produce benefit and the higher
intensities that cause sustained adverse effects. If he uses a constant-voltage pulse gen-
erator, this author attempts to provide a therapeutic window of at least 2 volts. If he uses
a constant-current pulse generator, he attempts to provide a therapeutic window of at
least 2 milliamps.
This author has frequently observed that at times intraoperative DBS test stimula-
tion does not reflect a patient’s postoperative experience. Some patients enjoying little
intraoperative clinical benefit may do well postoperatively, particularly when micro-
electrode monitoring suggests an optimal DBS lead trajectory. This improvement may
owe, among other reasons, to significant changes in electrical impedance as a result of
acute changes in the tissue around the electrodes associated with the acute trauma of
lead insertion. Acute and significant stress to the patient in the operating room may
blunt any clinical benefit. Finally, the generally greater flexibility of postoperative pro-
gramming allows discovery of more optimal programming of electrode configurations
and stimulation parameters than is possible in the operating room
Clinical Evaluation During Subthalamic Nucleus Stimulation
In the vicinity of the lateral subthalamic nucleus one finds the following major struc-
tures: the corticospinal tract, the medial lemniscus, the nerve roots of the oculomo-
tor neuron, and the medial subthalamic nucleus. Note: Because DBS in these two
regions of the subthalamic nucleus produces starkly different effects, a distinction is
made between the lateral and medial subthalamic nucleus. The general consensus is
that effective DBS activates neuronal elements in the lateral and dorsal region of the
subthalamic nucleus.
The corticospinal tract borders the subthalamic nucleus laterally, anteriorly, and
ventrally. At sufficient currents, stimulation propagation to the corticospinal tract may
cause tonic muscle contraction. The examiner must bear in mind that at lower stimula-
tion currents, frank tonic muscle contractions may escape her notice. She may notice
instead a reduction in tremor that she may mistake for a therapeutic effect. Were the
examiner to end her assessments upon gaining this false impression, the patient may
later experience tonic muscle contractions intolerable enough to limit the effective-
ness of the DBS. Should the examiner note a reduction in tremor, she must continue to
increase the stimulation voltage/current in order to determine whether she will observe
tonic contraction at higher stimulation voltages or currents.
Similarly, stimulation current propagating to the corticospinal tract may appear
to worsen bradykinesia. The examiner thus risks drawing an incorrect inference. For
example, she may observe that bradykinesia improves at lower stimulation voltages/
currents and worsens at higher stimulation voltages/currents. From this she concludes
that at the higher voltages/currents stimulation spreads to the corticospinal tract. Her
observations may moreover lead her to the erroneous inference that the DBS lead
is too close to the corticospinal tract. In fact, what she has observed is the putative
“U-shaped” curve of clinical response (Montgomery 2010). Such instances require that
one continue to increase the stimulation voltage/current to observe whether frank tonic
muscle contractions occur with an increase of 1 volt or 1 mA, with constant-voltage or
constant-current stimulation, respectively, a result which suggests that the DBS lead
lies inordinately close to the corticospinal tract.
Though a motor homunculus lies within the lateral subthalamic nucleus, the local-
izing value of the pattern of muscles contracting is limited. One cannot know whether
the corticospinal and corticobulbar fibers originating within or projecting to the
homuncular representation within the subthalamic nucleus were activated and there-
fore concordant with the pattern of muscle contractions. Nor can she know whether
fibers in passage actually relate to a different homuncular origin or termination and
therefore give rise to a discordant response.
When different contacts on the DBS lead are in use, thresholds to tonic contrac-
tion has some localizing value. A greater threshold implies a greater distance from the
surface of the cathodal (negative) contact and the corticobulbar and corticospinal tract
fibers. For example, the lowest threshold at the ventral-most contact and progressively
increasing thresholds with more dorsal contacts suggests that the long axis of the DBS
lead is not parallel to the long axis of the corticobulbar and corticospinal tract. This
pattern of thresholds might suggest a DBS lead that is too deep (ventral) or too lateral
(if the lead is not too shallow in the coronal plane) if tonic muscle contractions are con-
tributing to a narrow or inverted therapeutic window (Figure 12.10). In this case, the
only way to differentiate a trajectory that is too ventral from one that is too lateral is to
consult microelectrode and semi-microelectrode recordings. For example, the record-
ings of action potentials may demonstrate that the subthalamic nucleus was entered
relatively high compared with the image-guided navigation target at the bottom of the
Dorsal
Medial Lateral
Ventral
le
le
psu
psu
A B
l ca
l ca
rn a
rn a
in te
in te
limb
limb
STN STN
rio r
rio r
te
te
Pos
Pos
Fp 4.0 Fp 4.0
FIGURE 12.10 Schematic representation of situations in which there are low-threshold tonic

contractions when the ventral-most contact is used as the cathodal (negative) contact and
a higher threshold is seen when the dorsal-most contact is used as the cathode. A and B
represent coronal sections approximately 4 mm anterior to the midpoint of the line con-
necting the anterior and posterior commissures (AC-PD line midpoint). As can be seen, the
ventral-most contact is closer to the posterior limb of the internal capsule compared with
the dorsal-most contact. Consequently, the threshold to tonic muscle contraction will be less
with the ventral-most contact compared with the dorsal-most contact when either is used as
the cathode. A represents the situation where the DBS lead is too ventral (deep) and can be
retracted while still leaving a sufficient number of contacts within the subthalamic nucleus
(STN). B represents the situation where the DBS lead is too lateral and, consequently, just
raising the DBS lead is likely to remove contacts from within the STN. Note, drawings are not
to scale. Source: Modified from Schaltenbrand and Wahren (1977).
subthalamic nucleus. Additional clues suggesting that the trajectory is too lateral are
the bottom of the thalamus and the width of the zona incerta, which tend to be higher
and wider, respectively, in more lateral trajectories.
Similar thresholds to tonic muscle contraction at all the contacts suggests that the
long axis of the DBS lead lies parallel to the long axis of the corticobulbar and corti-
cospinal tract fibers such that each contact is equidistant to the posterior limb of the
internal capsule. Typically, this occurs when the angle in the sagittal plane shows the
trajectory as excessively anterior (Figure 12.11). However, if the trajectory is shallow in
the coronal plane, the long axis of the contacts in the lead will run parallel to the poste-
rior limb of the internal capsule (Figure 12.11).
Paresthesias produced by inadvertent stimulation of the medial lemniscus poste-
rior to the subthalamic nucleus may indicate an excessively posterior DBS lead place-
ment (Figure 12.12). However, one must distinguish paresthesias associated with
stimulation of the medial lemniscus from sensations produced by subtle mild muscle
Dorsal Dorsal
Posterior Anterior Medial Lateral
Ventral Ventral
ule
A B
ule
a ps
aps
al c
al c
tern
tern
b in
b in
r lim
STN STN
r lim
erio
erio
t
Pos
t
Fascicles of the
Pos
17.0 oculomotor
nerve Fp 4.0
FIGURE 12.11 Schematic representation of situations in which there are low-threshold tonic

contractions when the ventral-most contact is used as the cathodal (negative) contact and the
same approximate threshold is seen when the dorsal-most contact is used as the cathode.
A represents a sagittal section approximately 17 mm lateral to the midpoint of the line con-
necting the anterior and posterior commissures (AC-PC line midpoint). B represents s coronal
section approximately 4 mm anterior to the AC-PC line midpoint. As can be seen, the distance
between ventral-most contact and the posterior limb of the internal capsule is the same as that
between the dorsal-most contact and the posterior limb of the internal capsule. Consequently,
the threshold to tonic muscle contraction will be same with the ventral-most contact compared
with the dorsal-most contact when either is used as the cathode. Section A represents the situ-
ation where the DBS lead is too anterior in the subthalamic nucleus (STN). Section B represents
the situation where the DBS lead is too lateral when the trajectory is shallow in the coronal
plane. Note, drawings are not to scale. Source: Modified from Schaltenbrand and Wahren (1977).
contraction owing to stimulation of the corticospinal tract. Again, increasing the stim-
ulation voltage/current will help one to distinguish stimulation of the medial lemnis-
cus from stimulation of the corticobulbar–corticospinal tract in that stimulation of the
latter may produce muscle contraction.
Stimulation may affect speech, causing it to be slurred, nonfluent (stuttering, for
example), or arrested (see discussion above).
Stimulation in the vicinity of the subthalamic nucleus may impair eye movements.
Absence of parallel movement of the eyes indicates disconjugate gaze disturbance
(Figure 12.8). Patients experiencing this disturbance report diplopia, double vision
caused by the visual image’s falling on a retinal region of one eye that is nonhomolo-
gous with the retinal region of the other eye. The examiner may detect disconjugate
gaze disturbance by having the patient focus on the tip of the examiner’s finger or a
small spot of light and asking the patient to state the number of objects she perceives.
Subtle disconjugate gaze may be detected by observing the reflection of the spot of light
Dorsal
Posterior Anterior
Ventral
ule
aps
al c
tern
Tactile
b in
Vc
r lim
Medial
erio
lemniscus
t
STN
Pos
17.0
FIGURE 12.12 Schematic representation of situations in which there are low-threshold pares-

thesias when the ventral-most contact is used as the cathodal (negative) contact. The figure
represents a sagittal section approximately 17 mm lateral to the midpoint of the line con-
necting the anterior and posterior commissures (AC-PC line midpoint). As can be seen, the
ventral-most contact is close to the medial lemniscus. Note, drawing is not to scale. Source:
Modified from Schaltenbrand and Wahren (1977).
on the cornea. In disconjugate gaze a reflection does not fall on homologous regions
of the cornea (Figure 12.8). The patient must focus on both extremes of lateral gaze
in order to place maximum stress on the extraocular muscles innervated by the third
cranial nerve (oculomotor nerve), the structure that stimulation in the vicinity of the
subthalamic nucleus is most likely to affect. Disconjugate gaze disturbance and diplo-
pia suggest an inordinately medial DBS lead placement (Figure 12.11).
Though a patient with conjugate eye deviation may not report diplopia, she will
betray a lateral eye gaze preference, her eyes directed to the right or left most of the
time. Conjugate eye deviation and lateral eye preference suggest that electrical stimula-
tion had spread through the corticobulbar fibers within the posterior limb of the inter-
nal capsule to the fibers descending from the frontal eye fields. These findings suggest
an excessively lateral, anterior, or ventral DBS lead placement. Tonic muscle contrac-
tion following further increases in stimulation voltage/current supports the inference
that electrical stimulation has spread to the corticobulbar and corticospinal tracts.
Intraoperative stimulation may influence mood and affect. Patients may experience
sudden feelings of depression or euphoria, for example, which suggests that electrical
stimulation had spread ventrally to the limbic or medially to the frontal regions of the
subthalamic nucleus. When present, these symptoms suggest an excessively ventral or
medial DBS lead placement.
Clinical Evaluation During Globus Pallidus Interna Stimulation
In the vicinity of the sensorimotor region of the globus pallidus interna one finds the
following major structures: the nonmotor regions of the globus pallidus interna, the
corticospinal and corticobulbar tracts in the posterior limb of the internal capsule, and
the optic tract. Though there may be no other symptoms referable to their stimulation,
the exception being “feeling weird,” the nonmotor regions will likely fail to benefit from
stimulation. Lack of clinical benefit in the absence of any other adverse effect suggests
an excessively anterior DBS lead placement.
Corticospinal and corticobulbar tract involvement may be associated with exces-
sively posterior stimulation and is suggested by tonic muscle contraction or conjugate
deviation of the eyes (rare in this author’s experience). At low stimulation voltage/
current, spread of stimulation to the corticospinal tracts may manifest as worsening of
bradykinesia in patients with Parkinson’s disease (see discussion above).
Inadvertent spread of electrical stimulation to the optic tract may disturb vision.
Patients experiencing this will report seeing flashes of light in the contralateral hemi-
field. Some patients, however, may report only distorted perception of objects viewed.
The presence of these symptoms suggests an excessively ventral DBS lead placement.
Clinical Evaluation During Ventral Intermediate Thalamic Stimulation
In the vicinity of the ventral intermediate nucleus of the thalamus, one finds the following
structures: the tactile region of the ventral caudal nucleus of the thalamus (situated posteri-
orly) and the corticospinal and corticobulbar tracts (situated laterally). Spread of electrical
stimulation to the ventral caudal nucleus of the thalamus may produce paresthesias, which
suggests an inordinately posterior DBS lead placement. The homuncular representation of
the paresthesias has no localizing value (see Chapter 9). One must distinguish paresthesias
owing to stimulation of the tactile ventral caudal thalamus from the sensation of muscle
pulling owing to stimulation of the corticobulbar and corticospinal tracts. Again, tonic
muscle contraction resulting from an increase in stimulation voltage/current to treat par-
esthesias suggests that the sensations are related to muscle pulling, and that therefore the
DBS lead is excessively lateral (see discussion above). Speech may also be affected, becom-
ing slurred, nonfluent (such as stuttering), or arrested (see discussion above).
Anterior to the ventral intermediate nucleus target in the thalamus lies the ventral
lateral posterior oralis. Medial to the ventral intermediate thalamus lies the medical
group of thalamic nuclei. Deep brain stimulation of these targets will likely fail to pro-
duce benefit, although, to the best of this author’s knowledge, no identifiable adverse
effects are associated with it. Lack of benefit in the absence of an identifiable adverse
effect thus suggests an excessively anterior or medial DBS lead.
ASSESSMENT OF POTENTIAL COMPLICATIONS
Though somewhat rare, the most common intraoperative neurophysiological monitoring

complication is intracerebral hemorrhage. Inability to record neuronal extracellular action
potentials in an otherwise apparently functioning electrode may indicate an intracerebral
hemorrhage. Blood on the tip of a withdrawn electrode reinforces this suspicion. Because
withdrawing the electrode may wipe blood from it, one must extend the electrode tip by
use of appropriate electrodes in order to express blood. One may also examine a guide can-
nula to observe any blood ascending or escaping it. Though neurosurgeons may vary their
response to hemorrhage, many allow the guide cannula to remain in the brain in order to
allow blood to escape and to prevent the hematoma’s expansion.
Cases in which microelectrode or semi-microelectrode recordings fail to demon-
strate neuronal extracellular action potentials require one to ensure that the patient
has not received sedation. This author has been in the unfortunate position of having to
request that a DBS surgery be aborted when microelectrode recordings, which appeared
to be functioning, failed to record action potentials in patients who, as this author later
learned, had been given midazolam and morphine. In subsequent similar situations in
which the effects of midazolam and morphine were chemically reversed, this author
observed a return of neuronal action potentials in microelectrode recordings.
The discussion that follows is not intended to give the intraoperative neurophysiolo-
gist sufficient knowledge to conduct the appropriate neurological assessments (unless the
intraoperative neurophysiologist is a neurologist or neurosurgeon). Rather, because the
intraoperative neurophysiologist may be in the position to first note a complication, the
discussion below is intended to increase the intraoperative neurophysiologist’s ability to
recognize quickly a possible complication and seek appropriate evaluation.
The clinical manifestations of an intracerebral hematoma depend on its size and
location, both of which determine which functional systems are affected and, conse-
quently, which symptoms and signs will manifest. Indeed, it is estimated from postop-
erative imaging that 1 in 10 patients will have a hematoma, although in only 1 in 10 of
those will the hematoma be symptomatic. In this author’s experience, most hematomas
occur in the depth of the trajectories, probably that segment traversed by the micro-
electrodes. The sharp point of the microelectrode is more likely to cut tissue as opposed
to pushing tissue aside as may be the case with the blunt tips of the guide cannula stylet
or the DBS lead.
Hematomas deep in the trajectories of the ventral intermediate thalamus and the
subthalamic nucleus are in close proximity to the reticular activating system, which
is important for supporting consciousness. Larger hematomas, particularly those that
push or displace significant brain tissue, may compromise the reticular activating
system and thus lead to a decreased level of consciousness. Distinction is necessary
between the level of consciousness and the content of consciousness. For example, the
presence of delusions or hallucinations in a patient that otherwise appears alert sug-
gests an alteration in the content of an otherwise normal level of consciousness. The
presence of delusions in an otherwise alert patient thus does not suggest physical dys-
function of the reticular activating system.
Sometimes occurring without observable physical injury, delusions or hallucina-
tions may be related to stress or withholding of medications before surgery. An intra-
operative neurophysiologist must remain mindful of this concern, particularly when
treating patients with Essential tremor or dystonia taking benzodiazepines. These
patients risk experiencing sudden withdrawal symptoms if given flumazenil to reverse
the actions of midazolam prior to surgery. For patients whose agitation may interfere
with completion of surgery, sedation with dexmedetomidine often allows continued
recordings of neuronal extracellular action potentials. This author recommends that
dexmedetomidine be made available during DBS lead placement surgery for these
contingencies.
Inability to maintain a level of alertness conducive to meaningful, if disturbed,
interactions with the environment or an intraoperative neurophysiologist indicates a
reduction in level of consciousness. Patients may appear asleep, but may rouse at hear-
ing their names called before appearing to fall asleep again. At this point it would be
important to demonstrate that the patient’s vital functions are intact, the patient is being
well oxygenated, and the patient has not received any medications capable of sedation.
Also, the patient is examined for any sign of neurological dysfunction that would sug-
gest brain injury (as will be discussed below). Inspection of the withdrawn electrode
for evidence of bleeding or examination for blood emerging from the guide cannula
is done. In the event of any evidence of neurological injury, the DBS surgery should
be aborted and the patient undergo CT and/or MRI scanning. The temptation to just
implant the DBS lead should be avoided as the hematoma likely will displace tissue and
consequently, the DBS lead may not be placed in an optimal position. Resolution of the
hematoma may cause further shift of the brain and subsequent lead migration.
Lying adjacent to the ventral intermediate thalamus, subthalamic nucleus, and glo-
bus pallidus interna, the corticobulbar and corticospinal tracts may suffer damage that
leads to weakness of the muscles of the side contralateral to the site of damage. In the
A Patient attempting to B
hold fingers spread
apart
Examiner attempting to
squeeze the fingers
together
FIGURE 12.13 Schematic representation of testing the strength of the dorsal interossei, which
function to abduct the fingers (A) and the ankle extensor (B).
patient able to follow motor commands, specific muscles can be tested for weakness.
However, sophistication in testing is required, as muscles may vary in their weakness.
For example, with subtle damage to the corticospinal tract, proximal muscles, such
as those about the shoulder or upper arm, may retain strength. Distal muscles, on the
other hand, are most sensitive to damage to the corticospinal tract. This author recom-
mends specifically testing the dorsal interossei muscles of the hand and ankle extensors
(Figure 12.13).
In patients whose compromised levels of consciousness render them incapable of
cooperating with strength testing, other subtle signs of weakness may be examined.
For example, a mild facial weakness may give raise to asymmetry of the mouth or loss
of wrinkles from one side of the forehead (Figure 12.14). External rotation of the lower
extremity may indicate weakness (Figure 12.14). Conjugate eye deviation may sug-
gest injury to the frontopontine fibers in the posterior limb of the internal capsule
(Figure 12.11).
Signs other than weakness may be produced with injury to the corticobulbar
and corticospinal tract. When present, these findings lend support to the suspicion
of corticobulbar and corticospinal tract injury. Both deep and superficial reflexes
may be altered with involvement of the corticobulbar and corticospinal tracts. Deep
reflexes may be exaggerated, but failure to demonstrate exaggerated reflexes is not
evidence of absence of corticobulbar and corticospinal tract injury. Deep reflexes
typically are mediated by applying a rapid stretch to a muscle tendon and then
observing a reflex contraction of the muscle of that tendon. Typically, the tendon of
the quadriceps muscle just below the knee cap (patella) is tapped, with the leg jerk-
ing out (Figure 12.15). Another reflex is to tap on the Achilles’ tendon and observe
the foot bending at the ankle (Figure 12.15).
A B
FIGURE 12.14 Schematic representation of asymmetric facial weakness (A). The patient is

lying quietly, but asymmetry of the mouth is still apparent, as noted by flattening of the naso-
labial fold (A). Alternatively, the patient may be subjected to a painful stimuli and the mouth
will move asymmetrically. While quite, the patient also may demonstrate weakness noted by
external rotation of the leg (arrow in B).
Unlike deep reflexes, the superficial reflexes are often reduced with injury to the
corticobulbar and corticospinal tracts. A superficial reflex relatively easy to test in the
operating room is the superficial abdominal reflex in which the abdomen is stimu-
lated, such as by scratching, and contraction of the abdominal muscles is looked for
(Figure 12.16). When present, the superficial reflex is intact and is evidence against
involvement of the corticospinal tract.
Because the subthalamic nucleus lies close to the oculomotor nerve and its
emerging nerve fascicles, an intracerebral hemorrhage in this region may cause dis-
conjugate gaze (Figure 12.11). The ventral intermediate thalamus is immediately
anterior to the tactile ventral caudal thalamus, which is the relay nucleus for sensory
fibers arising in the body on their way to the sensory cortex. A hematoma in this
region may result in decreased sensation to pin pricks to the skin surface on the
contralateral body.
A B
FIGURE 12.15 Schematic representation of knee or patella deep tendon reflex (A) and Achilles’
deep tendon reflex (B). A slight stretch is placed on the tendon by holding the limb in the
appropriate position. A sudden additional stretch is applied using the reflex hammer. The
normal result is a rapid contraction of the muscles that cause the knee to extend and the ankle
to flex, respectively. With lesions of the corticospinal tract, the reflex muscle contraction is
exaggerated. Comparison to the tendon reflexes of the other side of the body serves as a
reference.
FIGURE 12.16 Schematic representation of superficial abdominal reflex. In this case, the left
upper quadrant of the abdomen is “scratched” with a relatively sharp edge. Normally, the
abdominal muscles in the quadrant stimulated reflexively contract, which can be seen as
movement of the umbilicus (belly button). With lesions of the corticospinal tract, the reflex is
lost. Comparison to the superficial reflexes of the other side of the body serves as a reference.
HINTS ON CLINICAL EVALUATIONS
The purpose of DBS is to improve the disability of neurological disorders, in this

case particularly movement disorders, through postoperative programming of the
implanted pulse generator. Because postoperative programming may be complex, phy-
sicians and healthcare professionals need to be confident that the DBS lead was placed
in an optimal location. Lack of confidence may lead to premature abandonment of the
therapy or to drawing prematurely the conclusion that DBS lead revision surgery must
be performed. Providing that confidence is the responsibility of the surgeon and intra-
operative neurophysiologist.
One must increase stimulation until effects clear. As discussed above, multiple
mechanisms may produce a specific adverse effect detectable by clinical assessment
during intraoperative stimulation. Effective DBS of the target may relieve tremor but
may also cause inadvertent propagation to the corticospinal tract. All cases require
that one confidently understand the mechanisms. The examiner must consequently
increase, within safe limits, stimulation voltage/current until she clearly determines
the nature of the adverse effect. She must also repeat the test stimulation until she
becomes confident of the results.
It is insufficient to perform intraoperative DBS testing at electrode configurations
and stimulation parameters deemed “average” in the published literature, as there is
no way of knowing whether a particular patient may need greater than the “average”
stimulation in order to achieve optimal benefit. If one applies “average” stimulation and
subsequently finds that higher stimulation, which is associated with adverse effects, is
needed to improve symptoms, then DBS is unlikely to bring the patient much benefit.
One must permanently document the stimulation configuration and parameters
for, and all the responses to, each test condition. In order to save time and effort, the
examiner may be tempted cut corners in her documentation. Such economizing does
not create a problem in most cases. Cases involving extremely complicated responses,
however, benefit from thorough documentation, because it is essentially a record of
efforts and results. Also, in the event of suboptimal postoperative responses, one may
review the detailed response to intraoperative test stimulation and glean important
insights for improving outcomes in future patients.
SUMMARY
Careful clinical assessments in the operating room serve the following three necessary
purposes: (1) demonstrating efficacy against the symptoms associated with the disease
being treated, (2) understanding the nature of any adverse effects and then correcting
them, and (3) assessing the possibility of complications during the surgery. An under-
standing of the regional functional anatomy in the vicinity of the electrode trajectories
facilitates adequate and accurate clinical assessments.
REFERENCES
Goetz CG, Tille, BC, Shaftman SR, et al.: Movement Disorder Society-sponsored revision of the Unified
Parkinson’s Disease Rating Scale (MDS-UPDRS): scale presentation and clinimetric testing results.
Movement Disorders 23(15): 2129–2170, 2008.
/ / / 13 / / / CASES
INTRODUCTION
The cases presented here were selected for heuristic purposes. For that reason, several
of the cases suggest that the DBS lead was not placed in the optimal position, as it is
said that “one learns more from error than chaos” or when things go right all the time.
The reader should not infer that these complications are frequent. In capable hands,
they occur in less than 2% of cases. Readers should interpret each report with a view to
determining a proper course of action.
Included with the case studies are the microelectrode recordings and the DBS test-
ing results. The microelectrode recording reports describe the depth at which neu-
rons were recorded and when microelectrode stimulation was applied. The report also
describes the characteristics of the neuronal activity at each recording site, the defini-
tions of which are provided in chapter 9. Also, included in the report were the results
of sensorimotor testing and from microstimulation. The reports in most of the cases
were taken from the intraoperative monitoring forms, examples of which are shown in
Appendices D–F.
The results of macrostimulation through the DBS lead also are shown. The report
forms are specific for the structure stimulated and the patient’s disorder. Blank forms
are available in Appendices G–I and can be downloaded from http://www.greenvil-
leneuromodulationcenter.com/. The forms indicate the status of each contact, whether
negative (cathode), positive (anode), or off (electrode configurations). Note there are
a number of naming conventions associated with the various contacts. The forms also
indicate the stimulation parameters, any adverse effects from macrostimulation, and
the effects on specific symptoms. The effects on symptoms form is designed as a visual
analog scale with the vertical hash mark indicating the rating for those symptoms.
Descriptions of the rating scales used are provided in chapter 12.
227
CASE 1
The patient has bilateral Essential tremor involving the distal musculature. The bottom
of the ventral intermediate nucleus of the thalamus was the intended target. The DBS
microelectrode and macrostimulation reports for this case appear on successive pages
(Tables 13.1–13.3). The depths, reported in millimeters, are relative to the microdrive.
The depth of the target, as determined by image-guided surgical navigation, is 25 mm.
Image-guided navigation also determined the first trajectory. The results of the macro-
stimulation through the DBS lead are provided.
TABLE 13.1 Documentation of the microelectrode recordings from the first trajectory.
First Trajectory
Depth Activity Description Sensorimotor Driving Microstimulation Response
4.0 Injury current
4.5 Moderate frequency Not tested
Low density
Transient
4.9 Moderate frequency Not tested
Moderate density
transient
5.7 Injury current Not tested
7.5 Low frequency Not tested
Low density
Transient
Low density
Transient
11.5 Decrease in background activity Not tested
12.9 Increase in background activity Not tested
Low density
16.2 Low frequency Light touch in fingers
Moderate density
Interpretation of the first trajectory: At depths 4.0 to 15.7 mm were thought to reside
dorsal tier thalamic neurons. At depth 16.2 mm, an increase in the density within the
recording sites and along the trajectory suggested that the microelectrode had passed
into the ventral tier. Accomplishment of the first objective—that of identifying the
posterior region of the ventral caudal nucleus of the thalamus associated with tactile
stimulation—occurred at a depth of 16.2 mm.
The tactile region of the ventral caudal thalamic nucleus was encountered at the high
point of the microelectrode trajectory (Figure 13.1). This is suggestive of a very poste-
rior trajectory. A planned second trajectory, 3 mm anterior to the first trajectory, was
to be slightly anterior to the anterior border of the tactile region of the ventral caudal
A B C
Vc- Vim
deep
Vc- Vop
tactile
Dorsal
Posterior Anterior Sagittal plane
Ventral 16.0 lateral
FIGURE 13.1 Schematic representation of the interpretation of the first trajectory’s results.

Tactile responses were encountered at the high point of the trajectory corresponding to
microelectrode A. Typically, the DBS lead is placed 2 to 3 mm posterior to the anterior border
of tactile ventral caudal nucleus of the thalamus. Had the DBS lead been placed 2 or 3 mm
anterior to the first trajectory, which did determine the anterior border of the ventral caudal
thalamus, corresponding to trajectory B, the DBS lead still would have been too close to the
ventral caudal thalamus ventrally. Low-threshold paresthesias would probably have resulted,
preventing therapeutic DBS. For that reason a second microelectrode trajectory was placed
3 mm anteriorly to more precisely localize the anterior border of the ventral caudal thalamus.
Discovery of the tactile ventral caudal nucleus more ventrally provides a reasonable estimate
of the location of the anterior border of tactile ventral caudal thalamus. Upon such discovery,
the DBS lead could then be placed in the trajectory corresponding to microelectrode C.
13. Cases / / 231
nucleus. If so, then the anterior border of the tactile ventral caudal thalamus would
have been bracketed between the first and second trajectory and would have enabled
identification of the anterior border. Because the homunculus in the thalamus tends to
shift medially with large movements anteriorly, the second trajectory was placed 2 mm
medially. The results are shown in Table 13.2.
TABLE 13.2 Documentation of the microelectrode recordings from the second trajectory.
Second Trajectory: Moved 3 mm Anteriorly and 2 mm Laterally

Microstimulation
Depth Activity Description Sensorimotor Driving Response
12.4 Low frequency
Low density
Irregular
Transient
12.9 Low frequency
Low density
Irregular
Transient
13.9 Injury current Not tested
16.7 Decrease in background activity
18.2 Decreased in background activity
25.4 No effects to 90 microamps
Low density
Interpretation of the second trajectory: At depths of 12.4 to 13.9 mm were thought to

reside dorsal tier thalamic neurons, owing to the low frequencies and low densities at
the recording site and in the trajectory. At depths of 16.7 to 25.4 mm neither neurons
nor injury currents were encountered. There arose concern as to a possible hematoma
in the trajectory at depths 16.7 to 25.4 mm. Concern arose also as to possible electrode
failure. Repeated impedance checks, however, were stable. Recordings were therefore
made as the microelectrode was withdrawn. There was a paucity of neurons available
for recording during withdrawal the microelectrodes. At the topmost point of the tra-
jectory, however, a single neuron was encountered, indicating that the microelectrode
functioned properly.
Whether the trajectory was situated anterior to the anterior border of the tactile
region of the ventral caudal nucleus of the thalamus remained unknown, because the
lack of neuronal recordings in the bottom half of the trajectory did not permit such
determination. Again concern arose as to a possible hematoma. Upon inspection, how-
ever, neither the microelectrode, cannula, nor stylet showed any sign of blood. Because
the first trajectory had encountered neurons responding to tactile stimulation, a third
microelectrode trajectory was placed 3 mm anteriorly and 2 mm laterally. The results
are shown in Table 13.3.
13. Cases / / 233
TABLE 13.3 Documentation of the microelectrode recordings from the third trajectory.
Third Trajectory: Moved 3 mm Anteriorly and 2 mm Laterally

Microstimulation
Low density
Transient
Low density
Irregular
24.2* Moderate frequency Questionable response
Low density to wrist flexion
Irregular
25.3 Decrease in background activity
25.9 Decreased in background activity
28.5 Moderate frequency Lower extremity response to muscle
Low density palpation
Irregular
28.9 Moderate frequency Lower extremity response to muscle
Low density palpation
Irregular
30.1 Moderate frequency Possible background response to
Low density wrist flexion
32.0 Moderate frequency Shoulder and hip passive movement
Low density
32.6 Deep Brain Stimulation lead
placed at this depth
* brief neurological examination unremarkable

Interpretation of the third trajectory: The lack of neuronal recordings at depths of

10.8 to 24.2 mm produced concerns about a possible hematoma. A hematoma may have
caused the brain to shift and in so doing has created uncertainty as to the microelec-
trode’s location. The patient showed no symptoms of intracerebral hematoma, how-
ever, and the microelectrode, cannula, and stylet bore no traces of blood. The surgeon
elected to place the DBS electrode in the first trajectory, a course of action suggested
to him by image-guided navigation. Performed was test macrostimulation through the
DBS electrodes, the results of which appear in the next page.
INTRAOPERATIVE MONITORING –MACROELECTRODE
Date: _______________ Page __ of __ ; Penetration number ___
Time started: ____________ Time stopped: ___________
Electrode
selection
0 1 2 3 Case Pulse Rate Volts Paresthesias and/or motor Rest tremor Postural tremor Action tremor Cup task
width phenomena
Where and transient vs
sustained
N O O P O 90 160 0 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1 none 0 1 2 3 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2 Speech affected 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2.5 Speech affected 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
3.0 Speech more affected, 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4

facial muscle contraction
Moved lead up 4 mm 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
N O O P O 90 160 0 Normal speech 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1 Speech affected 0 1 2 3 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
O O N P O 90 160 2 Normal speech 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
3 Normal speech 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
4 Normal speech 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Moved lead 3 mm 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
posterior
N O O P O 90 160 0 Feeling “goofy” 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2 Feeling “goofy” 0 1 2 3 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2.5 Feeling “goofy” 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Range of electrode impedances 3299 to 3334 ohms

Interpretation of macrostimulation through DBS electrodes: The effects on speech

were a significant issue. The voltage was increased to determine whether these effects
were related to any possible spreading of electrical current to the corticobulbar axons
residing in the posterior limb of the internal capsule lateral and below the ventral cau-
dal nucleus of the thalamus. Stimulation at 3 volts produced facial muscle contraction
indicating spread of the stimulation current to the posterior limb of the internal capsule.
Particularly significant is the fact that the threshold was 3 volts, which is in the range
that might be used postoperatively. If the DBS lead were to be left in this position, the
production of tonic muscle contraction would have limited the options for postoperative
stimulation and probably limited therapeutic efficacy. The DBS lead was subsequently
withdrawn (moved dorsally) by 4 mm, but the low threshold to speech effects persisted.
The DBS lead was removed and repositioned 3 mm posteriorly, because in the axial
plane the posterior limb of the internal capsule moves laterally with movement poste-
riorly. No change in the thresholds affecting speech was observed. The DBS lead was
allowed to remain in this location.
Despite repositioning of the DBS lead some considerable distance from the original
trajectory, the nearly continual tonic contractions with low-voltage DBS suggests, in retro-
spect, that the lead approached extremely close to the posterior limb of the internal capsule.
Such a close approach would have been difficult had the lead been placed in the thalamus.
On the day following stimulation, a magnetic resonance imaging (MRI) scan was
performed (Figure 13.2). Observed was significant tension pneumocephaly, which
A B
e
FIGURE 13.2 Postoperative MRI scan performed approximately 24 hours following DBS tar-
geting the ventral intermediate nucleus of the thalamus. One observes in (A) that the final
position of the DBS lead (d) lies extremely anterior vis-à-vis the ventral intermediate nucleus
of the thalamus and abuts the posterior limb of the internal capsule. Speech impairment with
stimulation likely owes to the current’s spreading to the internal capsule. In (B) there appears
pneumocephalus (e). Evidence of the microelectrode trajectories (c) appears in the medial
globus pallidus interna.
13. Cases / / 237
displaced the brain posteriorly. Positioned too anteriorly, the DBS electrode abutted
the posterior edge of the posterior limb of the internal capsule. Evidence of the micro-
electrode penetrations appears in the medial globus pallidus interna.
Figure 13.3 shows the position of the DBS lead in the coronal plane approximately
24 hours after lead implantation. Figure 13.3B shows the coronal plane approximately
parallel to and containing the DBS lead, while A, a few millimeters anterior, approxi-
mates the microelectrode trajectories. In B, the lead appears to lie in the extreme medial
aspect of the globus pallidus interna and to traverse the caudate-putamen. The trajec-
tories of the microelectrodes appear to traverse the caudate-putamen and the internal
capsule as they proceed to the medial aspect of the globus pallidus interna. The record-
ing sites in the caudate-putamen are typically low frequency, low density, and transient.
Confusion thus arises from the fact that these same characteristics apply to the dorsal
tier thalamic neurons. The lengths of the microelectrode trajectory in which no neu-
rons were encountered likely owed not to hematomas but to the microelectrode tip’s
penetration of the internal capsule.
A B
d f
e
g
c
FIGURE 13.3 Coronal sections of the MRI scan performed approximately 24 hours follow-
ing DBS lead placement. B shows the plane nearly parallel to and containing the DBS lead.
A shows the plane a few millimeters anterior to the lead and approximating the trajectories of
the microelectrode recordings. In B, the lead traverses the caudate-putamen (f) en route to the
medial globus pallidus interna (g). The proximity of the lead to the internal capsule explains
the continual tonic muscle contractions that follow attempts at macrostimulation through the
DBS lead. A shows the approximate trajectories of the microelectrode recordings. The trajec-
tory directs the microelectrodes through the caudate-putamen (d) and the internal capsule
(e) before reaching its terminal point in the medial globus pallidus interna (c).
CASE 2
The patient has Essential tremor. Her tremor is bilateral and has progressed to caus-
ing significant disability. Medications either brought the patient no effective relief or
visited her with intolerable adverse effects. The microelectrode and macrostimulation
reports appear on successive pages. Reported in millimeters, the depths are relative to
the microdrive. The depth of the bottom of the ventral intermediate thalamus (target),
as determined by image-guided surgical navigation, is 25 mm relative to the micro-
drive. Image-guided navigation established the initial trajectory. Provided also are the
results of the macrostimulation through the DBS lead. The reader is invited to inter-
pret each report and determine the action she would take to address the issues therein
(Table 13.4).
First Trajectory
Microstimulation
4.9 Low frequency Not tested Not tested
Low density
Irregular
Low density
Irregular
10.4 Decrease in background Not tested Not tested
13.2 Low Not tested Not tested
frequency
Low density
Irregular
14.3 Low frequency Shoulder and ankle rotation Not tested
Low density Palpation of gastrocnemius
Irregular and deltoid muscles
15.0 Moderate frequency Hip flexion Not tested
Low density
transient
15.5 Moderate frequency Shoulder rotation Not tested
Low density Deltoid palpation
Transient
16.9 Low frequency No response Not tested
Low density
13. Cases / / 239
First Trajectory
Microstimulation
Low density
Low density
17.8 Moderate frequency Ankle rotation, no response Not tested
Low density to muscle palpation
Irregular
Low density
Transient
Low density
Transient
19.0 Low frequency Wrist rotation, no response Not tested
Low density to muscle palpation
Irregular
Low density
Irregular
21.2 Low frequency Possible response to light Not tested
Low density touch on leg
Irregular
Low density
Irregular
21.9 Low frequency Light touch on leg Not tested
Low density
Irregular
22.1 Low frequency Light touch on arm Not tested
Low density
Irregular
Interpretation of the first trajectory: Characteristic of the dorsal tier of the thalamus,

depths 4.9 to 14.3 contained low-frequency, low-density, transient neuronal activity.
In addition, the distances between the neurons encountered at these depths—termed
“low density in the trajectory”—were fairly large compared with deeper segments
of the trajectory. The neuronal activity detected at five sites produced an average of
approximately 0.5 neuronal sites per millimeter of trajectory.
Increased frequencies and densities of neuronal activities at 15.0 mm led the neu-
rophysiologist to suspect that she had encountered the ventral tier—a discovery that
subsequent recording sites confirmed and that increased to approximately 0.7 the aver-
age sites per millimeter. In the lower segment of the initial trajectory the electrode
encountered neurons that responded to muscle palpation. This discovery enabled the
neurophysiologist to identify recording sites in the ventral intermediate nucleus of the
thalamus. The electrode then encountered sites whose neurons responded to joint rota-
tions but did not respond to muscle palpation. These sites probably lie in the anterior
portion of the ventral caudal nucleus of the thalamus, a portion responsive to deep sen-
sations. Finally, the electrode encountered at the deepest sites neurons that responded
to light touch and that therefore were likely located in the tactile region of the ventral
caudal nucleus of the thalamus.
The homuncular representation encoded by the neuronal activities began to reveal
itself at depths 14.3 mm to 15.5 mm, the predominant features being the leg and proxi-
mal upper extremity in the ventral intermediate nucleus of the thalamus. Yet the trajec-
tory appears to pursue too lateral a course, as suggested by the patient’s considerable
distal upper extremity tremor.
This trajectory accomplished the goals of establishing the target homuncular rep-
resentation, the medial-lateral position of the physiologically defined optimal target
(slightly medial to the microelectrode trajectory), and the anterior border of the tactile
region of the ventral caudal nucleus of the thalamus, which was low in the trajectory
corresponding to electrode trajectory B in Figure 13.4. One could consider a second
microelectrode trajectory 2 mm anteriorly; however, given the depth at which the ante-
rior border of the tactile ventral caudal thalamus was encountered, a second trajectory
likely would not have demonstrated neuronal activities related to tactile stimulation,
perhaps corresponding to trajectory C in Figure 13.4. The inference would have been
that the anterior border of the tactile ventral caudal thalamus would lie somewhere
between the first microelectrode trajectory and the second if the second had been done.
13. Cases / / 241
A B C
Vc- Vim
deep
Vc- Vop
tactile
Dorsal
Posterior Anterior Sagittal plane
Ventral 16.0 lateral
FIGURE 13.4 Schematic representation of the interpretation of the results of the trajectory.

Tactile responses were encountered very low in the trajectory corresponding to microelec-
trode trajectory B. A second microelectrode trajectory could have been done, but likely would
not have demonstrated any neuronal activity related to tactile stimulation, perhaps corre-
sponding to microelectrode trajectory C.
The microelectrode recordings bear out another important consideration, namely, that
the microelectrode in the single trajectory tangentially traversed the ventral interme-
diate nucleus of the thalamus and anterior portion of the ventral caudal nucleus of the
thalamus. The electrode also passed into the posterior portion of the ventral caudal
nucleus of the thalamus as neuron activities changed with light touch stimulation.
Pursuit of the tangential trajectory carried a risk of placing many of the DBS lead’s
electrode contacts on one side or the other of the ventral intermediate nucleus of the
thalamus (Figure 13.5). In order to eliminate this risk, the neurophysiologist placed the
DBS lead 3 mm anterior to the microelectrode trajectory. This she did despite her pre-
viously encountering the tactile portion of the ventral caudal nucleus of the thalamus.
She thus did not know for certain whether, despite her precautions, she might place the
ventral-most contact too close to the tactile region of the ventral caudal nucleus of the
thalamus. A more anterior placement of the DBS lead might have also resulted in fewer
electrical contacts entering the ventral intermediate nucleus of the thalamus.
DBS lead
Vc-deep Vim
Vc-tactile
Vop
Sagittal plane

to the AC-PC
Posterior Anterior line
Ventral
FIGURE 13.5 Schematic representation of the consequences of a too-shallow placement of a
DBS lead, its angle relative to the vertical plane of the AC-PC coordinate system excessively
acute. The ventral-most contact lies at the junction of the tactile and deep ventral lateral cau-
dal thalamus. The ventral contact lies at the junction of the deep ventral caudal nucleus and
the ventral intermediate nucleus. Only the dorsal contact penetrates the ventral intermediate
nucleus proper. The dorsal-most contact lies in the ventral oral posterior nucleus of the thala-
mus. Neither moving the DBS lead up or down nor translating the lead anteriorly corrects the
situation.
The DBS lead was placed in a trajectory 3 mm anterior and 1 mm medial (the latter
to account for the shift of the homunculus medially as trajectories move anteriorly)
to a depth of 25 mm. The responses to macrostimulation through the DBS electrodes
appear in the next page.
INTRAOPERATIVE MONITORING – MACROELECTRODE
Electrode
selection
width phenomena
sustained
N O O P O 90 160 0 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1 none 0 1 2 3 0 1 2 3 0 1 2 3 4 0 1 2 3 4
2 Persistent paresthesias 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
O N O P O 90 160 1 Transient paresthesias 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2 none 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
3 none 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
4 Persistent paresthesias 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Move lead dorsal 2 mm
N O O P O 90 160 0 none 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1 none 0 1 2 3 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2 none 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
3 Transient paresthesias 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
4.5 Transient paresthesias 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4

Persistent parenthesis at 2 volts (delivered by a constant-voltage implanted pulse

generator simulator) resulted from stimulation in which the ventral-most electrode
(contact 0) served as the cathode and the dorsal-most electrode (contact 3) as the
anode. While the DBS lead remained in place, the next ventral electrode was enlisted
as the cathode. This had the effect of adjusting the cathode upward and increasing to
4 volts the threshold to persistent paresthesias—an effect that the neurophysiologist
interpreted as a too-ventral placement of the DBS lead. This effect also suggested that
in the sagittal plane there was described too shallow a DBS trajectory.
Elevating the DBS lead 2 mm had the effect of increasing the threshold on persis-
tent paresthesias to 4.5 volts. Since most patients experience satisfactory tremor control
with stimulation of less than 4 volts, this lead location ought to provide adequate relief
(as suggested by significant tremor reduction) while avoiding paresthesias. Upward
adjustment of the DBS lead upward permitted postoperative use of the ventral-most
contact, as well as greater flexibility in terms of possible electrode configurations.
Allowing the DBS lead to remain in its original location would render the most ventral
contact unusable. The neurophysiologist ruled out anteriorly repositioning the DBS
lead because she wished to preserve a sufficient number of contacts within the interme-
diate nucleus in order to accommodate the tangential trajectory.
Of note in this case are a couple of issues. First, the paresthesias affected the face.
Yet the microelectrode recordings demonstrated that the trajectory did not involve
the head representation. It is not unusual for the paresthesias to be referred to areas of
the body different from those suggested by the somatotopy of the DBS lead, because
stimulation may affect local neurons and produce paresthesias in the region of the
body that corresponds to the somatotopy in which the lead is placed (see chapter 1).
The stimulation may also affect axons passing in the vicinity of the stimulating elec-
trodes en route to the somatotopic representation elsewhere. The distribution of
paresthesias produced by stimulation therefore cannot be used to infer the somato-
topic representation, which contains the stimulating electrodes. Another interesting
feature, the inconsistency of the transient paresthesias with increasing voltage, as yet
defies explanation. Figure 13.6 shows the postoperative MRI scan of a DBS lead occu-
pying a reasonable position.
13. Cases / / 247
A B
FIGURE 13.6 Postoperative MRI scan. The DBS lead is indicated by the white arrow in the
coronal plane (A) and horizontal plane (B). The lead appears in the ventral lateral tier of the
thalamus.
CASE 3
A patient with Essential tremor. The microelectrode and macrostimulation reports

appear on successive pages. Reported in millimeters, the depths are relative to the
microdrive. The depth of the target, as determined by image-guided surgical naviga-
tion, is 25 mm relative to the microdrive. Image-guided navigation also determined
the initial trajectory. Provided also are the results of the macrostimulation through the
DBS lead.
First Trajectory
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Low density
Irregular
Transient
7.9 Moderate frequency Not tested Not tested
Moderate density
Irregular
Transient
Low density
Irregular
Transient
13. Cases / / 249
First Trajectory
Low density
Irregular
Transient
Moderate density
Irregular
Transient
Moderate density
Irregular
Transient
Low density
Transient
Injury
Moderate density
Transient
Bursting
11.1 Moderate frequency Possible response to wrist rotation; Not tested
Moderate density no response to muscle palpation
Irregular
Moderate density
Irregular
Moderate density
Irregular
Moderate density
Irregular
Transient
13.2 Moderate frequency No response Not tested
Moderate density
Irregular
Moderate density
Irregular
(Continued)
First Trajectory
Moderate density
Irregular
14.3 Moderate frequency Response to active hand opening Not tested
Moderate density and closing but not passive
irregular movement
Low density
Irregular
Low density and closing but not passive
Irregular movement
Low density
Irregular
Low density
Irregular
Low density and closing but not passive
Irregular movement
17.1 Moderate frequency Response to active shoulder Not tested
Moderate density movement but not passive
Irregular movement
17.5 Moderate frequency Joint rotation about elbow; Not tested
Low density response to muscle palpation
Irregular
Low density
Irregular
Lost
18.5 Low frequency Joint rotations about Not tested
Low density the metacarpal phalangeal joints
Irregular but not muscle palpation
Lost
19 Low frequency No response Not tested
Low density
Irregular
Low density
Irregular
Lost
13. Cases / / 251
First Trajectory
19.9 Low frequency Response to elbow rotation but not Not tested
Low density muscle palpation
Irregular
Low density
Irregular
Low density
Irregular
Low density
Irregular
22.0 Low frequency Light touch to fingers Not tested
Low density
Irregular
22.4 Low frequency Light touch to fingers Not tested
Low density
Irregular
22.7 Moderate frequency Light touch to upper lip Not tested
Low density
Irregular
23.1 Moderate frequency Light touch to upper lip Not tested
Low density
Irregular
23.6 Low frequency Light touch to back of upper arm Not tested
Low density
Irregular
Moderate density
Irregular
Transient
24.8 Moderate frequency Light touch to upper arm Not tested
Low density
Irregular
Low density
Irregular
Transient
Interpretation of the first trajectory: Depths 3.2 mm to 11.1 mm were characteristic

of the dorsal tier of the thalamus in all respects. The neuronal activity occupying the
recording sites were low frequency, low density, and transient. In addition, the dis-
tances between the neurons encountered at these depths—termed “low density in
the trajectory”—were fairly large compared with deeper segments of the trajectory.
The neuronal activity detected at 20 sites produced an average of approximately 1.7
neuronal sites per millimeter of trajectory. The microelectrode’s relative low imped-
ance at 0.55 megaohms rendered problematic the task of determining the absolute
density within the trajectory. In order to overcome this problem the neurophysi-
ologist estimated the densities in the trajectory by referring them to the densities
encountered along the entire trajectory. Such estimation had the effect of normaliz-
ing otherwise different microelectrode characteristics, and it was assumed that the
microelectrode accurately recorded extracellular action potentials as they occurred
at each site.
Increased frequencies and densities of neuronal activities at 13.6 mm led the neu-
rophysiologist to suspect that the microelectrode had entered the ventral tier—a dis-
covery that subsequent recording sites confirmed. Higher than expected for the ventral
caudal or ventral intermediate nuclei of the thalamus, the frequencies and the densities
were more typical of ventral oral posterior nucleus of the thalamus. There was moreover
a paucity of neurons responsive to sensorimotor driving in response to passive move-
ments. However, at depths of 14.3 mm, 15.5 mm, 16.7 mm, and 17.1 mm, neurons were
found that changed their activity with active movements generated by the patient but
not with passive movements initiated by the examiner. These findings also are consis-
tent with those involving the ventral oral posterior nucleus of the thalamus.
Encountered at a depth of 17.5 mm were neurons that responded to joint rotations
and muscle palpation. This encounter suggested that the microelectrode trajectory
had traversed the ventral oral posterior and entered the ventral intermediate nucleus
of the thalamus. At depths of 18.5 mm and 19.9 mm were encountered neurons that
responded to joint rotations but not muscle palpation. This encounter suggested that
the microelectrode trajectory had traversed the ventral intermediate nucleus of the
thalamus and entered the deep sensation region in the anterior portion of the ventral
caudal thalamic nucleus. At depths 22 mm to 24.8 mm were encountered six sites
whose neurons responded to light touch. This encounter suggested that the microelec-
trode trajectory traversed the deep sensation part and entered the tactile ventral caudal
nucleus of the thalamus. The fact that the length of the tactile ventral caudal thalamus
portion of the trajectory measured only 3.4 mm, however, suggested that the micro-
electrode entered the anterior border of the tactile ventral caudal thalamus relatively
13. Cases / / 253
low in the nucleus. Receptive fields in the face corresponding to two of these sites sug-
gested that the trajectory was excessively medial.
The microelectrode recordings bore out another important consideration, namely,
that the microelectrode in the single trajectory traversed tangentially through the ven-
tral oral posterior, ventral intermediate, deep ventral caudal, and tactile ventral caudal
nuclei of the thalamus. This suggested that the electrode pursued too shallow a trajec-
tory through the vertical axis of the thalamus (Figure 13.7). Because the planning MRI
showed that the entry point to the brain was immediately anterior to the motor cortex,
a more posterior entry point, which would have accommodated a more vertical trajec-
tory relative to the thalamus, would have proven impossible.
DBS lead
Vc-deep Vim
Vc-tactile
Vop
Sagittal plane

to the AC-PC
Posterior Anterior
line
Ventral
FIGURE 13.7 Schematic representation of the probable trajectory pursued during the initial
microelectrode recordings. The microelectrode traversed the ventral oral posterior (Vop),
ventral intermediate (Vim), deep ventral caudal (Vc-deep), and tactile ventral caudal nuclei of
the thalamus (Vc-tactile).
Pursuit of the tangential trajectory carried a risk of placing many of the DBS lead’s
electrode contacts on one side or the other of the ventral intermediate nucleus of the
thalamus. In order to eliminate this risk, the neurophysiologist placed the DBS lead
3 mm anterior to the initial microelectrode trajectory. This she did despite her previ-
ously encountering the tactile portion of the ventral caudal nucleus of the thalamus.
She thus did not know for certain whether, despite her precautions, she might place the
most ventral contact too close to the tactile region of the ventral caudal nucleus of the
thalamus. A more anterior placement of the DBS lead might have also resulted in fewer
electrical contacts entering the ventral intermediate nucleus of the thalamus. A more
anterior placement of the DBS lead might have also resulted in fewer electrical con-
tacts entering the ventral intermediate nucleus of the thalamus. The neurophysiologist
positioned the second trajectory 2 mm laterally in order to avoid the head homuncular
region, the penetration of which may have affected the patient’s language, speech, and
swallowing. The microelectrode traveled relatively quickly until it reached a depth of
15 mm, at which point it paused solely for the purpose of ascertaining that neurons situ-
ated the bottom of the ventral thalamic tier were amenable to recording. The results of
the microelectrode recordings during the second trajectory appear in Table 13.6.
13. Cases / / 255
Second Trajectory
2.0 Moderate frequency To ascertain ability to record Not tested
Low density neuronal activity
Irregular
2.6 Moderate frequency To ascertain ability to record Not tested
Low density neuronal activity
Irregular
Low density
Irregular
15.7 Moderate frequency Active hand opening and closing; no Not tested
Low density response to passive movements
Irregular
Transient
Low density
Irregular
Transient
17.9 Moderate frequency Elbow rotation and muscle Not tested
Moderate density palpation
Irregular
Low density
Irregular
Transient
Lost
18.5 Moderate frequency Wrist rotation and muscle Not tested
Low density palpation; no relation to tremor
Irregular
19.3 Low frequency Wrist rotation and muscle Not tested
Irregular
Transient
Lost
19.7 Low frequency Shoulder rotation and muscle Not tested
Irregular
Transient
Lost
(Continued)
Second Trajectory
Low density
Irregular
Transient
Injury
21.7 Low frequency Shoulder rotation and muscle Not tested
Moderate density palpation; no relation to tremor
Irregular
Transient
Lost
Low density
Irregular
Transient
Lost
23.7 Low frequency Hip rotation Not tested
Low density
Irregular
Transient
Lost
Moderate density
Irregular
Transient
Moderate density
Irregular
26.0 Decrease in background Not tested No response to 90 microamps
Interpretation of the second trajectory: As it did in the initial trajectory, the micro-
electrode traversed the ventral oral posterior, ventral intermediate, and anterior ventral
caudal thalamic nuclei, but stopped short of the tactile ventral caudal nucleus. Among
the homuncular representations encountered was that of the wrist. Given the fact the
patient’s tremor was predominantly distal, this encounter was important. No neurons
were encountered, however, in the head representation. Microstimulation at or below
90 microamps produced no paresthesias. These observations suggested the second
microelectrode trajectory had bracketed the anterior border of the tactile ventral cau-
dal nucleus of the thalamus. The DBS lead pursued a trajectory 2 mm anterior to the
second microelectrode trajectory to a depth of 26 mm. The responses to macrostimula-
tion through the DBS electrodes appear in the following page.
INTRAOPERATIVE MONITORING –MACROELECTRODE
Electrode
selection
width phenomena
sustained
N O O P O 90 160 0 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1 none 0 1 2 3 0 1 2 3 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4

Macrostimulation through the DBS lead to mimic the therapeutic effect met with an
excellent response from the patient. Figure 13.8 shows the postoperative MRI scan
with the DBS lead occupying a reasonable position.
A B
FIGURE 13.8 Postoperative MRI scan. The DBS lead is indicated by the dark hole in the coronal
plane (A) and horizontal plane (B). The lead appears in the ventral lateral tier of the thalamus.
CASE 4
The patient, who has Parkinson’s disease, is undergoing DBS surgery for the subtha-
lamic nucleus. The microelectrode and macrostimulation reports appear on successive
pages. Measured in millimeters, the depths are relative to the microdrive. Based on
image-guided surgical navigation, the depth of the target is 25 mm on the microdrive.
The first trajectory is based on image-guided navigation. In addition, the results of the
macrostimulation through the DBS lead are provided (Table 13.7).
First Trajectory
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Transient
Injury
Low density
Transient
Low density
Transient
Injury
Low density
Transient
13. Cases / / 263
First Trajectory
18.9 Moderate frequency Weakly related to ankle Not tested
Moderate density dorsiflexion
Irregular
Injury
19.5 High frequency Weakly related to flexion and Not tested
Moderate density extension at the knee
Irregular
20.1 High frequency Weakly related to ankle Not tested
Moderate density dorsiflexion
Irregular
20.4 High frequency Elbow and shoulder rotation Not tested
Moderate density
Irregular
Injury
Low density
Irregular
Moderate density
Irregular
Moderate density
Irregular
23.7 Moderate frequency Finger flexion Not tested
Mow density
Irregular
24.6 Moderate frequency Hand flexion and extension Reduction in muscle tone at 90
Moderate density microamps
Irregular
26.1 Decrease in background Not tested
Low density
Transient
Low density
Regularly irregular
Interpretation of the first trajectory: Recording site neuronal activity at depths between

6.9 mm to 15.2 mm was low to moderate frequency, low density, and transient. These
characteristics suggested that the electrode had encountered the anterior thalamus
above the subthalamic nucleus. Background activity decreased at a depth of 16.3 mm.
Only upon reaching a depth of 18.9 mm did neuronal activities resume, which sug-
gested that the electrode had entered the zona incerta between the thalamus and the
subthalamic nucleus.
Multiple sites residing at depths from 18.9 mm to 29.8 mm exhibited neuronal
activities whose frequencies ranged from low to high, with moderate frequencies and
densities predominating. Many demonstrated sensorimotor driving. As the recordings
demonstrated, the microelectrode had entered the sensorimotor region. Testing every
recording site for sensorimotor driving was therefore deemed unnecessary. At a depth
of 24.6 mm microstimulation produced a reduction in muscle tone.
At no recording site did there occur high-frequency, low-density, and regular neu-
ronal activity indicative of the presence of the substantia nigra pars reticulata, which,
depending on the laterality, may lie posterior to the subthalamic nucleus. Sampling
issues may prevent a trajectory from traversing the sustantia nigra pars reticulata, mak-
ing recording impossible. Sensorimotor responses corresponded with an extremely
long segment. If the trajectory does not lie vertical to the coronal plain, it may traverse
the long axis of the subthalamic nucleus. The patient’s spatial orientation may also
admit of a normal variant.
The trajectory met the following requirements: sensorimotor recordings of greater
than 5 mm; an absence of adverse response to microstimulation; and a site where the
last sensorimotor responsive recordings lying at a depth of 24.6 mm. The following
pages presents the results of macrostimulation through the DBS lead.
Electrode
selection
0 1 2 3 Case Pulse Rate Volts Rest tremor Effect on finger Effect on hand Tone
width tapping opening and closing
N O O P O 90 160 0 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1 Tonic contraction of face 0 1 2 3 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
P O O N O 1 Tonic contraction of face 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1 Transient sensation of 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
facial pulling
1.5 Transient sensation of 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
facial pulling
2.0 Tonic contraction of face 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Moved DBS dorsal 2 mm 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
N O O P O 90 160 1 Tonic contraction of face 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
O N O P O 90 160 1 Tonic contraction of face 0 1 2 3 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
O O N P O 90 160 1 Sensation of facial pulling 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Move lead 2 mm posterior 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4

and 2 mm ventral
N O O P O 90 160 1 Tonic contraction of face 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
O O N P O 90 160 1.5 Face feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2 Tonic contraction of face 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Moved lead 2 mm dorsal 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
N O O P O 90 160 0 Sensation of facial pulling 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Range of electrode impedances 2521 to 2799

Date: _______________ Page__ of __ ; Penetration number ___
Electrode
selection
0 1 2 3 Case Pulse Rate Volts Rest tremor Effect on finger Effect on hand Tone
N O O P O 90 160 1.5 Tonic contraction of face 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Moved lead 2 mm medial 0 1 2 3 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4

and 2 mm ventral
N O O P O 0.5 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1.0 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1.5 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2.0 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2.5 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
O O N P O 0.5 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1.0 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1.5 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2.0 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2.5 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
3.0 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
3.5 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
4.0 Feels funny 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
4.5 Feels funny, speech 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4

slurred
Insertion of the DBS leads resulted in a significant micro-subthalamotomy effect and

produced fairly minimal symptoms. Though the micro-subthalamotomy effect is sug-
gestive of appropriate placement, one must refrain from placing full confidence in it.
In this case, one must therefore base his assessments on any adverse effects associated
with macrostimulation.
Initial 1-v macrostimulation caused tonic contraction of the face, a reaction which
may owe to the DBS lead’s too ventral, anterior, or lateral placement. The active con-
tacts were reconfigured in such a way that the dorsal-most contact became the cathode
(negative contact). The rationale for reconfiguration rested on suspicion of a too ven-
tral or lateral original configuration (Figure 13.9). A caveat: A shallow trajectory in the
coronal plane—a trajectory that moves from lateral to medial, increasing in depth as it
does so—may place the DBS lead parallel to the posterior limb of the internal capsule.
This placement may not show differences in thresholds to tonic contraction between the
ventral-most contact, which is the cathode (negative contact), and the dorsal-most con-
tact, which is the cathode (negative contact). In these positions the dorsal-most contacts
are further away from the corticospinal tract in the posterior limb of the internal capsule
(Figure 13.9). In order to identify an excessively ventral, lateral, or anterior placement,
the lead was elevated 2 mm to detect any change in the thresholds to tonic contraction.
A B
Ventral intermediate thalamus

sule
Electrical Electrical
field
le
al cap
field
apsu
Thalamus
Intern
nal c
DBS lead DBS lead

Inter
Subthalaminc nucleus
Subthalaminc nucleus
Dorsal
Dorsal
Medial Lateral
Posterior Anterior
Ventral
Sagittal plane 17.0 Ventral
Coronal plane Fp 4.0
FIGURE 13.9 Schematic representations of the regional anatomy of the subthalamic nucleus

in the sagittal (A) plane 17 mm lateral to the AC-PC line and coronal (B) plane 4 mm posterior
to the midpoint of the AC-PC line. In A, the DBS lead is excessively anterior, and both the
ventral-most and dorsal-most contacts are nearly the same distance from the posterior limb of
the internal capsule. When used as cathodes, the ventral-most or dorsal-most contacts share
the same threshold to tonic contraction (negative contacts). The DBS lead in B is excessively
ventral, and the ventral-most contact lies closer to the posterior limb of the internal capsule
than does the dorsal-most contact. In cases where one contact or the other serves as the cath-
ode, the ventral-most contact’s threshold to tonic contraction is lower than the dorsal-most
contact’s. The same would hold true for an excessively lateral placement of the DBS lead.
The DBS lead was elevated dorsally by 2 mm. The second trajectory reached the same
tonic contraction threshold as that of the original. From this followed the conclusion
that the DBS lead placement was excessively anterior. The DBS lead was repositioned
2 mm posteriorly and 2 mm ventrally. Low thresholds to tonic contraction were again
evident despite any changes in electrode configuration. After the DBS was raised 2 mm
dorsally these low thresholds persisted. Because a posterior and dorsal repositioning
failed to reduce the threshold to tonic contraction, the DBS lead was moved 2 mm
medially to the original depth.
Though stimulation in the third DBS lead trajectory (2 mm medial) did not appear
to produce tonic contraction, it did produce an effect that the patient could only charac-
terize as “feeling funny,” which he experienced at low thresholds, and which occurred
despite changes in electrode configuration. The threshold was increased to 4.5 v in
order to detect any other effect that might provoke the same complaint, but none was
found. Readers may wish to gather their own impressions as to the DBS lead’s location
before proceeding.
The postoperative MRI scans appear in Figure 13.10. The DBS lead appears to lie above
the subthalamic nucleus, extremely close to the posterior limb of the internal capsule,
and appears to traverse the medial globus pallidus interna. The MRI scans reformat-
ted in the coronal and sagittal planes that contain the DBS lead (Figure 13.11), a new
estimate situated the microelectrode trajectory 2 mm anterior and 2 mm lateral to the
original trajectory (Figure 13.12).
A B C D
FIGURE 13.10 A series of axial MRI scans from ventral (A) to dorsal (D). The arrows in sections
B–D indicate the DBS lead. The DBS lead in A is extremely medial, nearly penetrating the red
nucleus. As the scans progress more dorsally, the DBS lead appears to lie in the posterior limb
of the internal capsule (B and C) and close to the medial globus pallidus interna.
13. Cases / / 273
A C
B
e
d
FIGURE 13.11 Axial MRI (A) and reconstructed sagittal (B) and coronal (C) planes containing
the DBS lead. The tip of the DBS lead in A appears to lie at the posterior edge of the posterior
limb of the internal capsule. In the coronal plane (C), there appear the DBS lead and remnants
of two posterior DBS lead trajectories (d and e). The lead appears to lie between the posterior
limb of the internal capsule with medial globus pallidus interna. The upper contacts appear to
lie within the medial globus pallidus interna.
A B
FIGURE 13.12 Postoperative MRI in which the microelectrode trajectory was reconstructed

from the position of the DBS lead. Thus, the coronal (A) and sagittal (B) planes containing the
microelectrode recordings based on the distances from the DBS lead. The trajectory appears
to traverse the medial globus pallidus interna.
Postoperative DBS programming continued to indicate low threshold to tonic contrac-

tions when ventral-most contacts were in use—an outcome consistent with the prox-
imity to the posterior limb of the internal capsule. Interestingly, a configuration of the
dorsal-most contact as cathode (negative contact) and the ventral-most contact anode
(positive contact) resulted in significant improvement in the patients tremor, rigidity,
and bradykinesia. The location of the DBS lead, particularly the dorsal-most contact as
cathode, suggests that the benefit may owe to activation of the axons exiting medially
through the globus pallidus interna, including the lenticular fasciculus and the ansa
lenticularis (Figure 13.13).
In retrospect, a number of usual findings might have alerted one to the fact that
the microelectrode and DBS leads had not reached the intended target. Unusually but
not impossibly long for the subthalamic nucleus, the length of sensorimotor recordings
suggested the electrode had encountered the globus pallidus interna. An encounter
with the substantia nigra pars reticulata increases one’s confidence that the trajectory
had passed through the subthalamic nucleus. A second, more posterior microelectrode
trajectory in this case would have helped one to gather recordings that suggest the elec-
trode had reached the substantia nigra pars reticulata, but the risks of additional micro-
electrode penetrations and prolonged surgery make this option unappealing.
Responses to macrostimulation through the DBS lead also provide some insight.
Had the lead entered the subthalamic nucleus, repositioning it posteriorly ought to have
increased the distance to the posterior limb of the internal capsule and, consequently,
the threshold to tonic contraction. Had the DBS lead entered medial globus pallidus
interna, repositioning it posteriorly would place it closer to the posterior limb of the
internal capsule, and the threshold to tonic contraction would thus be lowered (unless
IC
Th
A
Pt GPe
GPi B
FIGURE 13.13 Photomicrograph of a coronal section showing the putamen, globus pallidus

externa (GPe), the globus pallidus interna (GPi), the posterior limb of the internal capsule (IC),
the thalamus (Th), the lenticular fasciculus (A), and the ansa lenticularis (B). These last are the
pallidofugal fibers exiting the medical globus pallidus interna. Source: Modified from Mosby
(2002).
13. Cases / / 275
A B
FIGURE 13.14 Postoperative MRI scans (A axial; B coronal) depicting a tension pneumocepha-

lus (arrows) that may have caused the brain to shift posteriorly, from which resulted an exces-
sively anterior placement of microelectrode and DBS leads.
the threshold has already reached a minimum, as it has in the present case). Extensive
past experience suggests that the patient’s “feeling funny” is highly unusual. Such a
report may thus relate to stimulation of the medial globus pallidus interna, which may
involve the limbic or cognitive systems. Report of the same sensation was made by
another patient in whom the DBS lead had entered the medial globus pallidus interna.
The extremely anterior placement of the microelectrode and DBS lead may have
been due to a quantity of air in the skull, which caused tension pneumocephalus that
displaced the brain posteriorly (Figure 13.14). This displacement likely occurred prior
to the microelectrode recordings. A relatively shallow trajectory in the sagittal plane
meant that the microelectrode and the DBS lead might move from anterior to poste-
rior with increasing depth (Figure 13.15). This increases the risk of the microelectrode
trajectory’s traversing the globus pallidus interna. As demonstrated in this case, differ-
entiating the globus pallidus interna from the subthalamic nucleus by use of microelec-
trode recordings is problematic.
B
DBS lead
IC
A
Th
Pt
GPi
ZI
STN
Dorsal
Posterior Anterior
Ventral
Sagittal plane 16.0
FIGURE 13.15 Schematic representation of the potential consequence of an excessively shal-
low trajectory in the sagittal plane (A) when compared with a more typical trajectory (B).
The sagittal plane is 16 mm lateral to the AC-PC line. The typical trajectory (B) takes the
microelectrode through the anterior thalamus (Th), through the zona incerta (ZI), and into
the subthalamic nucleus (STN). An excessively anterior skull entry point leads to an exces-
sively shallow trajectory. The trajectory takes the microelectrode through the putamen (Pt),
whose differentiation from the anterior thalamus (Th) by use of microelectrode recordings
is difficult, and through the medial globus pallidus interna, whose differentiation from the
subthalamic nucleus by use of microelectrode recordings is likewise difficult. A great deal of
anatomical variability is possible, making the shallowness of trajectory appear extreme. As
the present case demonstrates, some patients’ rotated anatomy causes the excessively shal-
low trajectory.
CASE 5
This patient with Parkinson’s disease is undergoing DBS surgery to target the globus
pallidus interna. The microelectrode and macrostimulation reports appear on suc-
cessive pages. The depths, reported in millimeters, are relative to the microdrive. The
depth of the target, as determined by the microdrive image-guided surgical navigation,
is 25 mm. Image-guided navigation also determined the first trajectory. The results of
the macrostimulation through the DBS lead are provided (Table 13.8).
First Trajectory
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Transient
Injury
Low density
Injury
Low density
Transient
Low density
Transient
Low density
Transient
13. Cases / / 279
First Trajectory
Low density
Transient
Low density
Transient
Low density
Transient
Low density
Transient
Injury
Low density
Transient
12.8 High frequency pausing Not tested Not tested
High density
Irregular
Low density
Regular
Lower pitch
13.5 High frequency Knee flexion and extension Not tested
Moderate density
Irregular
13.9 High frequency Knee flexion and extension Not tested
Moderate density
Irregular
14.8 Injury Not tested Not tested
15.2 High frequency Not related to passive or active Not tested
Moderate density movement
Irregular
16.0 Moderate frequency Not related to passive or active Not tested
Irregular
16.8 Moderate frequency Ankle flexion and extension Not tested
Moderate density
Irregular
(Continued)
First Trajectory
17.1 Moderate frequency Knee flexion and extension Not tested
Moderate density
Irregular
17.7 High frequency No related to passive or active Not tested
Irregular
Low density
Irregular
19.0 Moderate frequency Shoulder movement Not tested
Moderate density
Irregular
19.3 High frequency Elbow flexion and extension Not tested
Moderate density
Irregular
19.6 High frequency Hip flexion and extension Not tested
Moderate density
Irregular
20.2 Moderate frequency Knee flexion and extension Not tested
Moderate density Shoulder movement
Irregular
20.5 High frequency Knee and ankle flexion and Not tested
Moderate density extension
Irregular
21.1 High frequency Knee and hip flexion and Not tested
Irregular Shoulder movement
Low density
Irregular
22.6 Reduced background Not tested Not tested
24.6 Reduced background Not tested No response to photic
stimulation
25.2 Reduced background Not tested No response to
microstimulation up to
90 microamps
Interpretation of the first trajectory: At depths between 3.9 mm and 11.3 mm, background
neuronal activity suggested low frequencies and low densities at the recording site. Neurons
at the site were consistent with the putamen. At a depth of 12.8 mm, activity suggested the
presence of high-frequency pausing neurons consistent with the globus pallidus externa.
A recording site at a depth of 13.1 mm contained moderate-frequency, low-density neuro-
nal activity. This activity was regular, and its action potentials were of a lower pitch than
that of action potentials at other sites. These findings are consistent with the presence of
border cells between the globus pallidus externa and globus pallidus interna.
Recording sites lying at depths of 13.5 mm to 21.7 mm contained neuronal activity
of irregular moderate-to-high frequencies and moderate densities. Neurons generating
this activity also indicated sensorimotor driving, which predominantly occurred in the
lower extremities. These findings were consistent with the lower extremity homuncu-
lar representation of the sensorimotor globus pallidus interna. Placement of the DBS
lead in this position risked failing to improve the patient’s upper extremity symptoms.
Because the lower extremity homunculus occupies an anterior, medial, and dorsal posi-
tion in the globus pallidus interna, a second trajectory was placed 2 mm posterior and
2 mm lateral to the initial trajectory. The results of microelectrode recordings in this
trajectory are shown in Table 13.9.
Second Trajectory
Depth Activity Description Sensorimotor Driving Stimulation Response
Low density
Irregular
Transient
Low density
Irregular
Transient
Moderate density
Irregular
Moderate density
Irregular
Injury
17.5 Moderate frequency Knee extension Not tested
Moderate density
Irregular
13. Cases / / 283
Second Trajectory
low density
Regular
Lower pitch
Low density
Regular
Lower pitch
19.0 Moderate frequency Wrist flexion and extension Not tested
Moderate density
Irregular
19.3 Moderate frequency Ankle dorsiflexion Not tested
Moderate density
Irregular
19.6 Moderate frequency Ankle dorsiflexion Not tested
Moderate density
Irregular
Moderate density
Irregular
Moderate density
Irregular
21.6 Lost Not tested Not tested
21.9 Moderate frequency Elbow flexion and extension Not tested
Moderate density
Irregular
Moderate density
Irregular
Moderate density
Irregular
Lost
26.3 None Not tested No response to
90 microamps
Interpretation of the second microelectrode trajectory: The relative position of the glo-

bus pallidus interna having been determined in the first trajectory, the microelectrode
was quickly repositioned to a depth of 15 mm. At 3.1 mm and 4.2 mm the microelec-
trode was halted in order to demonstrate the ability to record neuronal activity. These
pauses were necessary to assure the quality of the microelectrode recordings.
At depths from 15.8 mm to 17.5 mm neuronal activity was characterized by
high-frequency pause activity, moderate frequency, and moderate density at the record-
ing sites. These findings are consistent with globus pallidus externa. At a depth of
17.8 mm, neuronal activity was regular, of moderate frequency, low density within the
recording site, and of a lower pitch than encountered elsewhere. These findings are con-
sistent with border cells between the globus pallidus externa and globus pallidus interna.
At depths of 18.4 mm to 22.5 mm, neuronal activity ranged from moderate to irreg-
ular high frequency, and its density within the recording site was moderate. Neuronal
activity also indicated sensorimotor driving, which had extended to the upper extrem-
ity homuncular representation. At a depth of 26.3 mm, microstimulation below 90 μa
did not appear to produce any adverse effects.
Because the representation was discovered in the proximal upper extremity, some
concern was felt that DBS lead placement in this trajectory would fail sufficiently to
help hand function. Also, the fact that the anterior border of the posterior limb of the
internal capsule went unidentified made necessary a third trajectory positioned 3 mm
posterior and 2 mm lateral to the second trajectory. The results of the microelectrode
recordings in the third trajectory are shown in Table 13.10.
TABLE 13.10 Documentation of the microelectrode recordings from the third trajectory.
Third Trajectory
Moderate density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
13. Cases / / 285
Third Trajectory
Moderate density
Irregular
Low density
Regular
Lower pitch
Moderate density
Irregular
19.0 Lost Not tested Not tested
Low density
Irregular
20.1 High frequency Ankle flexion and extension Not tested
Moderate density
Irregular
20.7 High frequency Wrist flexion and extension Not tested
Moderate density
Irregular
21.5 Moderate density Finger flexion Not tested
Irregular
Moderate density
Irregular
22.7 High frequency No active movement of Not tested
Moderate density jaw or tongue
Irregular
low density
Regular
Lower pitch
26.1 Reduced background Not tested No response to photic
stimulation or microstimulation
up to 90 pps
Interpretation of the third microelectrode trajectory: The relative depth of the globus

pallidus interna having been established by previous trajectories, the microelectrode
was quickly advanced to a depth of approximately 15 mm. Care was taken during the
microelectrode’s advancement to ensure that neuronal activities were continually
encountered. This measure was necessary in order to eliminate the possibility, however
unlikely, of the microelectrode’s entering the posterior limb of the internal capsule, the
only indication of which would have been a loss of neuronal activity. Twice in its rapid
advance—at 4.0 mm and 4.7 mm—the microelectrode was halted in order to deter-
mine the quality of the microelectrode recordings.
Encountered at a depth of 14.8 mm was neuronal activity characterized by irreg-
ular, high-frequency pausing, which was consistent with the globus pallidus externa.
Encountered at a depth of 17.0 mm was neuronal activity of regular moderate fre-
quency. It was low density within the site, and its pitch was lower than that of other
recorded activities. All these findings were consistent with border cells lying between
the globus pallidus externa and globus pallidus interna.
Neuronal activities recorded at depths between 18.5 mm and 24.1 mm were of
high-to-moderate irregular frequencies. Their densities within the recording site were
moderate. These findings were consistent with the globus pallidus interna. In addition,
sensorimotor driving suggested that homuncular representations extended to the dis-
tal upper extremity. To save time, tests of sensorimotor driving at deeper sites were
limited to the jaw and tongue. This was done to determine whether the head homun-
cular representation lay within the trajectory such that DBS stimulation would lead to
speech or swallowing difficulties. These findings suggest that the trajectory had settled
in the appropriate medial-lateral position.
At a depth of 24.1 mm, low-density and low-pitch activities of regular and moder-
ate frequency within the site were encountered. This finding was consistent with the
presence of border cells at the bottom of the globus pallidus interna. Microstimulation
up to 90 pps produced no responses, and photic stimulation generated no changes in
background neuronal activity. These findings if present would suggest the presence of
the optic tract nearby.
The posterior limb of the internal capsule was not encountered, which is necessary
to avoid an excessively posterior DBS placement with the risk of tonic contractions
complicating DBS therapy. Similarly, identifying the anterior border of the posterior
limb of the internal capsule helps avoid DBS lead placement excessively anterior to the
sensorimotor regions of the globus pallidus interna. Merely demonstrating that a tra-
jectory reaches the sensorimotor region and placing the DBS lead in that trajectory
fails to exclude the possibility of an excessively posterior DBS lead placement. Such
13. Cases / / 287
possibility, however, would have been detected during macrostimulation through the
DBS lead.
The third trajectory was 4 mm posterior to the first trajectory, which demonstrated
sensorimotor driving. This third trajectory’s use suggested that it was likely that an
excessively anterior DBS lead placement was avoided, consequently, the DBS lead was
placed in this trajectory of the third microelectrode recordings. The results of macro-
stimulation through the DBS lead are shown in the following page.
Electrode
selection
0 1 2 3 Case Pulse Rate Volts Rest tremor Postural tremor Action tremor Cup task
width
N O O P O 90 160 0 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1 No side effects 0 1 2 3 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1.5 No side effects 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2 No side effects 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2.5 No side effects 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2.0 No side effects 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
3 No side effects 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
3.5 No side effects 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
4 No side effects 0 1 2 3 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4

Interpretation of macrostimulation through the DBS lead: A remarkable micropalli-

dotomy effect is evident. The DBS lead placement appears sound, but whether the
micropallidotomy effect owed to any one penetration of the brain is difficult to deter-
mine. It is also problematic to assume that the micropallidotomy effect resulted from
the actual DBS lead placement trajectory. With stimulation at 2 v, however, the patient’s
hand opening and closing improved. No further side effects of macrostimulation were
encountered.
Figure 13.16 shows images taken from a postoperative MRI. The scan shows the
DBS lead in the posterior lateral globus pallidus interna. One notes the presence of a
small amount of air in the skull.
FIGURE 13.16 Axial MRI scan showing the DBS lead in the posterior lateral region of the globus
pallidus interna.
CASE 6
This patient with Parkinson’s disease is undergoing DBS surgery to target the globus
pallidus interna. The microelectrode and macrostimulation reports appear on suc-
cessive pages. The depths, reported in millimeters, are relative to the microdrive. The
depth of the target, as determined by the microdrive image-guided surgical navigation,
is 25 mm. Image-guided navigation also determined the first trajectory. The results of
the macrostimulation through the DBS lead are provided (Table 13.11).
First Trajectory
Microstimulation
Low density
Irregular
Transient
9.1 High frequency pause Not tested Not tested
Moderate density
Irregular
Lost
Moderate density
Irregular
12.0 High frequency Not tested Not tested
Regular
Low density
12.7 Moderate frequency Elbow and wrist flexion and Not tested
Irregular
Moderate density
Irregular
15.3 Moderate frequency Finger flexion and extension Not tested
Moderate density
Irregular
13. Cases / / 293
First Trajectory
Microstimulation
15.8 Moderate frequency Finger flexion and extension Not tested
Moderate density
Irregular
Moderate density
Irregular
18.3 Moderate frequency Shoulder rotation Not tested
Moderate density
Irregular
Moderate density
Irregular
20.2 Moderate frequency Not related to jaw or tongue Not tested
Moderate density movements
Irregular
Irregular
21.1 High frequency Not related to jaw or tongue Not tested
High density movements
Irregular
Irregular Tremor related
Irregular
Moderate density
Irregular
Lost
Interpretation of the first trajectory: Few neurons were encountered at recording sites

deeper than 9 mm. A recording site lying at a depth of 8.8 mm contained neurons
whose activity was consistent with activity characteristic of neurons in the putamen.
Recording sites at depths between 9.1 mm to 11.1 mm contained high-frequency-pause
neurons, a finding consistent with the activity characteristic of the neurons in the glo-
bus pallidus externa. Neurons recorded at a depth of 12.0 mm were low density, high
frequency, and regular—all characteristics consistent with those of border cells. At the
recording sites neuronal activity was consistent with activity characteristic of neurons
of the globus pallidus interna.
The neurons recorded, which were in the upper extremity of the homuncular rep-
resentation, established this trajectory as lying in the optimal medial-lateral sagittal
plane. The next task became that of determining the location of the anterior border
of the internal capsule’s posterior limb. A second trajectory was therefore made 3 mm
posteriorly. The results of the microelectrode recordings are shown in Table 13.12.
Second Trajectory
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Low density
Irregular
13. Cases / / 295
Second Trajectory
Low density
Irregular
Low density
Irregular
Moderate density
Irregular
15.0 Moderate frequency Not responsive Not tested
Moderate density
Irregular
15.3 Moderate frequency Not responsive Not tested
Moderate density
Irregular
Moderate density
Irregular
Moderate density
Irregular
Moderate density
Irregular
Irregular
Irregular
(Continued)
Second Trajectory
Irregular
Irregular
Irregular
Low density movements
27.0 None Not tested No response to
90 pps
Reduction in muscle
tone in the contralateral
elbow and wrist
Interpretation of the second trajectory: Because the top of the globus pallidus externa
was identified at 10 mm in the trajectory, the microelectrode was rapidly advanced to
that depth. Neurons were identified at depths of 4.2 mm and 6.9 mm to demonstrate a
functioning microelectrode. The recording site at a depth of 11.3 mm was thought to
lie in the putamen. Because recording sites from 12.0 mm to 15.3 mm were thought
to lie in the globus pallidus externa, they were not tested for sensorimotor driving.
Recording sites from 16.7 mm to 23.4 mm were thought to lie in the globus pallidus
interna. At a depth of 16.7 mm, sensorimotor driving indicated units related to the
upper extremities. Because the homuncular representation was determined to lie in the
upper extremities, the task became that of ensuring that trajectory avoided the homun-
culus’s head region in order to reduce the risk of affecting speech. The subsequent sen-
sorimotor testing was consequently confined to two tasks: jaw opening and closing and
tongue protrusion and retraction.
The lowest point in the trajectory associated with neuronal recordings was the same
as the lowest point of the first trajectory. It was therefore concluded that the second
trajectory had not entered the posterior limb of the internal capsule. There could have
been made a third trajectory more posterior than the second, attempting to encounter
the posterior limb of the internal capsule to avoid an excessively anterior placement of
most of the DBS lead contacts not in the sensorimotor region. Yet because the first tra-
jectory demonstrated sensorimotor driving 3 mm anterior, such risk would most likely
have been avoided. Risk lay, rather, with an excessively posterior placement of this third
trajectory. Macrostimulation through the DBS lead, however, would have allowed this
risk to be detected. For these reasons the DBS lead was placed in the third trajectory
3 mm anteriorly, assuming the worst case scenario of the posterior limb of the internal
capsule being immediately posterior to the second trajectory, and implanted at a depth
of 24 mm. The results of macrostimulation appear in the following pages.
Microstimulation was performed solely at the bottom of the trajectory. This owed
to the fact that a tungsten microelectrode was used, and it likely would not have with-
stood the stimulation. Stimulation therefore occurred solely at such times as it was
anticipated that microelectrode recordings in this trajectory were no longer needed.
Intraperative DBS testing results are shown on the following page.
Jaw tightening suggested involvement of the corticobulbar fibers during macrostimu-

lation DBS lead. Also, propagation to the corticospinal fibers was suggested by the
worsening of the finger-tapping and hand-opening and -closing tasks as stimulation
was increased from 2 volts to 2.5 volts. Because there was concern of an excessively
deep or posterior DBS lead placement, the electrical field was moved dorsally by
selecting dorsal contact (Contact 2) as the cathode (negative contact) and allowing
the dorsal most contact (Contact 3) to remain the anode (positive contact). This met
with no significant change in the threshold to the side effects, suggesting that the DBS
lead was placed excessively posterior and excessively close to the posterior limb of the
internal capsule.
The DBS lead was repositioned 2 mm anteriorly, because the primary concern
became that of possible side effects, which the patient nonetheless continued to expe-
rience. This suggested that during the time that the ventral-most contact (Contact
0) served as the cathode (negative contact) and the dorsal-most (Contact 3) the anode
(positive contact), stimulation current spread to the posterior limb of the internal cap-
sule. The electrical field was thus increased by changing the cathode to dorsal contact
(Contact 2) and allowing the dorsal-most contact (Contact 3) to remain the anode.
Any threshold to side effects, had the latter been present, would have been greater than
4 volts. The DBS lead was therefore allowed to remain in this trajectory but was moved
2 mm dorsally. Figure 13.17 shows several consecutive axial MRI scans with the loca-
tion of the DBS lead.
FIGURE 13.17 Sequence of axial MRI scans depicting the location of the DBS lead. The
sequences range from most ventral (left) to most dorsal (right). The white arrow appearing in
the fourth scan indicates the paramagnetic defect caused by the DBS lead, which was placed
in the posterior lateral ventral globus pallidus interna.
CASE 7
The patient, who has Parkinson’s disease, is undergoing DBS surgery to target the sub-
thalamic nucleus. The microelectrode and macrostimulation reports appear on suc-
cessive pages (Table 13.13). The depths, reported in millimeters, are relative to the
microdrive. The depth of the target, as determined by the microdrive image-guided
surgical navigation, is 25 mm. Image-guided navigation also determined the first tra-
jectory. The results of the macrostimulation through the DBS lead are provided.
First Trajectory
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Transient
Low density
Irregular
Transient
13. Cases / / 303
First Trajectory
14.7 Moderate frequency Tremor related Not tested
Low density
Irregular
Transient
15.8 Moderate frequency Finger flexion and Not tested
Irregular
16.5 Moderate frequency Finger flexion and Not tested
Irregular
19.5 Moderate frequency Wrist flexion and Not tested
Irregular
19.8 High frequency pause Elbow flexion and Not tested
Irregular
20.4 High frequency Wrist flexion and Not tested
Irregular
20.9 High frequency Tremor related Not tested
High density
Irregular
Moderate density
Irregular
Moderate density
Irregular
Moderate density
Irregular
22.5 High frequency pause Tremor related Not tested
Moderate density
Irregular
High density
Irregular
(Continued)
First Trajectory
Moderate density
Irregular
Moderate density
Irregular
Low density
Regular
24.3 No neurons Not tested No adverse effects, slight
tremor reduction
Interpretation of the first trajectory: At depths of 7.0 mm to 14.7 mm neuronal activ-

ity recorded was generally low to moderate, and its density at and between recording
sites was moderate. At certain times traversal of 1 mm or greater was required before a
recording site was encountered. These recordings were consistent with the thalamus.
Recorded at a depth of 13.8 mm was neuronal activity of moderate frequency and
low density. Such activity appeared consistent with high-frequency-pause neurons
often encountered in the globus pallidus interna. Some concern arose that the globus
pallidus may have been traversed, as happens in situations in which the subthalamic
nucleus is the target. In such situations air may enter the skull and thus cause tension
pneumocephalus. Yet there were encountered no additional recording sites suggestive
of the presence of high-frequency-pause neurons.
At sites ranging in depth from 14.7 mm to 24.0 mm, recorded neuronal activity was
moderate to high frequency, and it occurred in moderate-to-high density often discern-
ible at intervals of a few tenths of a millimeter. Activity at many recording sites sug-
gested sensorimotor driving. Neuronal activity at a number of sites were observed, both
visually and aurally, to be periodic, oscillating at the same frequency as the patient’s
tremor on the contralateral side. Activity at these recording sites was consistent with
activity characteristic of sensorimotor driving.
Note that the precise homuncular representation at the site was not discriminated.
Unlike the globus pallidus interna and the ventral intermediate thalamus, the homun-
cular representation at this site does not affect the targeting. At recording sites whose
periodic activity clearly occurred at the same frequency as did the patient’s tremor,
it was not necessary formally to identify the homuncular representation; the tremor
effectively accomplished this. Also, traversal of 5 mm of the sensorimotor subthalamic
nucleus satisfied a major criterion, thus eliminating the need for sensorimotor testing
and leaving only the need of discovering that structure’s bottom.
The relatively small volume, compared with the globus pallidus interna, of the sen-
sorimotor subthalamic nucleus or ventrolateral thalamus eliminates the necessity of
identifying the homuncular representation. The size of these latter structures is such
that it is not possible to affect solely that part of the contralateral body which corre-
sponds to the homuncular representation. The representation, in other words, occupies
an area roughly equivalent to the typical volume of tissue activation with DBS. It is
therefore not important to identify the specific homuncular representation that corre-
sponds to the part of the body responsible for the greatest loss of function as is the case
for the globus pallidus interna or ventral intermediate nucleus of the thalamus.
Traversal of the subthalamic nucleus’s length began at 14.7 mm and ended at
24.0 mm. Though this trajectory met the criterion of spanning at least 5 mm of
13. Cases / / 307
sensorimotor subthalamic nucleus, its unusually great length likely relates to the angle
in the coronal plane, which would suggest that it projected to the long axis of the sub-
thalamic nucleus (Figure 13.18, coronal view).
Of lower density also was neuronal activity recorded at sites along that segment of
the trajectory which traversed the middle of the subthalamic nucleus (see chapter 10).
This lower density, which is common, may lead one to mistake the center of the subtha-
lamic nucleus for its bottom. Yet had one in fact reached the bottom rather than the cen-
ter, the trajectory would have failed to meet the criterion of sufficient length, namely, at
least 5 mm of sensorimotor subthalamic nucleus.
Neuronal activity recorded at the end of the trajectory, which lay at a depth of 26.2
mm, was high frequency and of low density. As such, it was consistent with activity
characteristic of the substantia nigra pars reticulata. This evidence strongly suggests
that, rather than reaching the medial globus pallidus, the trajectory did in fact reach
the intended target, the subthalamic nucleus. For this trajectory, the distal edge of the
DBS lead’s ventral-most contact reached a depth of 24.3 mm. A reduction in back-
ground activity at this depth suggested that the trajectory exited the bottom of the
subthalamic nucleus. The results of macrostimulation through the DBS lead appear in
the following page.
INTRA-OPERATIVE MONITORING – MACROELECTRODE
Electrode
selection
0 1 2 3 case Pulse Rate Volts Effects on tone Effects on finger Effects on hand Effects on tremor
N O O P O 90 160 0 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1 No side effects 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
2 No side effects 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
3 No side effects 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
4 No side effects 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
5 No side effects 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4

Remarkable improvement without side effects followed macrostimulation through the

DBS lead. Though the need to do so was unlikely to arise, the voltage was increased to
5 volts. Such high voltage was shown to be safe and thus offered greater flexibility to
the programmer responsible for postoperative DBS maintenance. Figure 13.18 consists
of images gathered from the postoperative MRI scan. They show the DBS lead to be
positioned well.
1 2
B D
A
C
E
FIGURE 13.18 Coronal-view (1) and axial-view (2) postoperative MRI scans. The coronal view
shows the four contacts of the DBS lead (A). The ventral-most and ventral contacts lie in the
lateral STN (D), whose decreased density owes to paramagnetic effects of increased iron
concentration in the STN (D and B) and the red nucleus (C and E). In B and C the triangle and
circle indicate the location of the STN and red nucleus respectively. The axial view shows the
DBS lead’s tip, which is indicated by the white arrow. In the coronal view (1), the four contacts
(A) form an angle that is not vertical. This owes to the need of avoiding the lateral ventricle.
The angle carried the microelectrode and DBS leads tangentially through the STN in a lateral
(dorsal) to medial (ventral) direction, which perhaps explains the greater length of STN in the
microelectrode recordings.
CASE 8
The patient, a 60-year-old right-handed male, has a 15-year history of Essential tremor.
The tremor, which is bilateral, involves the upper extremities distally and has pro-
gressed to the point of causing significant disability. Medications either brought the
patient no effective relief or produced intolerable adverse effects. The microelectrode
and intraoperative macrostimulation report appears below.
Not before arriving 13 mm above the target did the electrode record any extracel-
lular action potentials. Of a low density at the recording site, neuronal activity was
low frequency and produced occasional bursting activity and injury currents. As the
electrode progressed to 5 mm above the target, neuronal activity continued to be low
frequency, low density within the recording site, and low density in the trajectory. At
2.7 mm above the target were found extracellular action potentials that responded to
voluntary mouth-opening and tongue-protrusion tasks. Microstimulation produced
no adverse effects. The forgoing represents the entirety of the documentation.
Is this an optimal location to place the DBS lead, or is another microelectrode
recording trajectory required? If a second trajectory is needed, then in which direction
ought it be moved?
The trajectory was excessively medial for at least two reasons: (1) rather than the
head, the distal upper extremity experienced the greatest limiting symptoms; and
(2) placing the DBS lead in the head representation increases the risk of language,
speech, and swallowing problems. Unfortunately, the documentation did not pre-
cisely describe the extent of the sensory testing to determine whether the response was
attributable to tactile, deep sensation, or muscle spindle receptor stimulation. Thus, it
is unclear why the intraoperative neurophysiologist chose to move 1 mm posteriorly.
A second trajectory was placed 2 mm lateral and 1 mm posterior to the first trajec-
tory. Of low density at the recording site and along the trajectory, neuronal activity was
encountered once the electrode progressed to 13 mm above the target. Beyond this
distance high-frequency extracellular action potentials were encountered. These were
thought to be consistent with activity characteristic of tremor. At 2.3 mm above the tar-
get, neuronal activities were encountered that were thought to respond to joint rotation
about the elbow. The operative record, however, contained no mention as to whether
there was a response to muscle palpation, but this does not mean that it was not done.
Microstimulation at the target did not produce any responses.
The DBS lead was placed 1 mm above the image-guided target in the second trajec-
tory. Stimulation with the ventral-most contact as the cathode (negative contact) and
the dorsal-most contact as anode (positive contact) reduced tremor at 2 volts and up to
4 volts produced no side effects.
The patient’s postoperative course was complicated. Programming or readjustment
of the stimulation parameters and electrode configurations would produce satisfac-
tory benefit lasting a few weeks to a few days after each attempt at DBS programming.
Stimulation through the ventral-most contact produced a feeling of dysphoria and anx-
iety. Monopolar DBS with the ventral and dorsal contacts as cathodes the patient found
tolerable. The postoperative MRI scan appears in Figure 13.19.
13. Cases / / 315
A D E
H
G
B F
FIGURE 13.19 Axial views of the postoperative MRI (A) with an expanded view of the region
around the third ventricle (D). As can be seen in B and H, the DBS lead is extremely anterior
and medial. In the expanded view, the AC-PC line runs from E to F. (Note that F is not the PC,
which fell outside the plane of the MRI scan, but the stalk of the pineal, which lies immedi-
ately above and posterior to the PC.) The DBS lead is approximately 1.5 mm posterior to the
midpoint (G) of the AC-PC line. (Typically, the usual target, which is approximately 8 mm pos-
terior to the midpoint of the AC-PC line, confirms that the DBS lead is excessively anterior.)
Also, whereas the typical target is 12 to 15 mm lateral, the distance from the midline (AC-PC
line) to the DBS lead laterally was approximately 10 mm. In this case, the relatively wide third
ventricle displaces the optimal target even more laterally and shifts the DBS lead to an exces-
sively medial position. A large amount of intracranial air is apparent, producing a tension
pneumocephalus (C) as evidenced by effacement of the gyri.
The DBS lead is extremely anterior, lying approximately 1 mm posterior to the mid-
point of the AC-PC line. The DBS lead typically lies 4 mm anterior to the posterior
commissure. Also, the DBS lead is extremely medial, lying approximately 10 mm from
the AC-PC line. In fact, the extremely wide third ventricle is making the DBS lead
probably even more medial in the thalamus. Indeed, the ventral-most contact may lie
in the lateral hypothalamus, which accounts for the dysphoria with DBS.
This case has a number of important instructive points. First, the misplaced leads
indicated the importance of minimizing intracranial air, which in this case caused the
brain to shift posteriorly and laterally. This shift then caused the DBS lead to be placed
too anteriorly and medially. The temporary benefit from the DBS also is consistent with
a misplaced DBS lead.
The intraoperative neurophysiological monitoring was suboptimal as well. First, the
documents of the microelectrode recordings did not provide information regarding the
trajectory in the sagittal plane—its anterior to posterior position, specifically, though
this does not mean is was not done. This possible failure owed primarily to inability to
find the tactile ventral caudal nucleus of the thalamus, the location of which would have
been revealed by neuronal activity in response to light touch. Though neurons respon-
sive to joint rotation and therefore tremor were found, to a great extent such neuronal
activity can be found in the sagittal plane, which extends from the anterior region of
the ventral caudal thalamus, through the ventral intermediate of the thalamus to the
ventral oral posterior thalamus. Further, the operative record made no mention of neu-
rons responsive to muscle palpation, which is indicative of muscle spindle–responsive
neurons related to the ventral intermediate thalamus. Identifying the anterior border of
the ventral caudal thalamus’s tactile region is thus important, because this region offers
best evidence that the trajectories are not excessively anterior.
A site in which the extracellular action potentials appeared to respond to
mouth-opening and tongue-protrusion tasks suggests that the trajectory was medial
in the head representation. The head representation is generally avoided in order to
minimize the risk of dysarthria and dysphagia with DBS. The second trajectory was
correspondingly placed 2 mm laterally and 1 mm posteriorly. The homunculus shifts
laterally as the trajectory moves posteriorly. This suggests that the 2-mm lateral move
probably was less than 2 mm relative to the homunculus. However, recording sites
in the second trajectory demonstrated responses to joint rotations about the elbow.
Though the second trajectory, which subsequently was used for the DBS lead, may not
have been excessively medial for the nuclei situated well anterior to the ventral interme-
diate nucleus of the thalamus, it still would have been extremely medial with respect to
the ventral intermediate nucleus.
/ / / 14 / / / FUTURE INTRAOPERATIVE
NEUROPHYSIOLOGICAL
MONITORING
INTRODUCTION: THE FUTURE
The indications for DBS in neurological and psychiatric disorders are rapidly expand-
ing. DBS itself should be considered part of a larger therapeutic domain based on
exploiting electrophysiology. Examples include neuroprosthetics such as artificial
retinas and machine-brain interfaces. It is highly likely that some form of intraopera-
tive neurophysiological monitoring will be used during surgeries to implement these
therapies even if it is only testing the device and only indirectly related to localizing the
targets for these therapies.
The potential for electrophysiologically based therapies is limited only by the
imagination because the brain basically is an electronic device. The brain processes
and transmits information electronically. Neurotransmitters, which are the basis for
many pharmacological treatments, are the messengers between neurons and not the
message. The message is the pattern of electrical discharges, neuronal action potentials.
There is growing realization that disability, at least in the central nervous system, pre-
dominantly is due to misinformation or loss of information. Deep brain stimulation
and other related therapies should be able intervene electronically to correct the effects
of the misinformation.
It also is possible that other therapies not based on electrophysiology but based on
stereotactic and functional neurosurgery may evolve that could require intraoperative
neurophysiological monitoring, such as cell transplantation or gene therapies. Again,
the same principles would apply to the rational use of intraoperative neurophysiologi-
cal monitoring in those cases. To be sure, cell transplantation and gene therapies using
317
stereotactic neurosurgery have been applied and, to the knowledge of this author,
without intraoperative neurophysiological monitoring. The results thus far have been
poor for fetal cell transplants and unimpressive for gene therapies, although the latter
is very early in its development and may achieve therapeutic efficacy that might rival
DBS. Yet, it is reasonable to ask whether intraoperative neurophysiological monitoring,
based on the principles described above, would have led to more effective results due
to better targeting. It would be a shame for these therapies, particularly gene therapy,
to be abandoned because failure to address the principles described above resulted in
less-than-optimal targeting.
Predicting the future of intraoperative neurophysiological monitoring is very
problematic. Unfortunately, one cannot look to the past to inform future expecta-
tions. The approaches used by some physicians and surgeons are those they learned
through apprenticeships, and the habits instilled often are confused for knowledge.
Consequently, many current approaches are unprincipled, and therefore there is a
dearth of principles by which to shape the future. This is not to say that surgeons, physi-
cians, intraoperative neurophysiologists, and healthcare professionals are unprincipled
but only to say that epistemic analyses are not commonplace in medical education or
practice (the interested reader may see ReasonBasedMedicineAndScience.com).
Whether there are any mechanisms in place to ensure that each and every practitio-
ner utilizes the best approaches, whether established by prospective randomized trials
(unlikely to happen in this author’s opinion) or by careful critique based on principle,
is unclear. Currently, the lack of any meaningful external supervision or consequence
of what approach any practitioner chooses is not encouraging. Ultimately, it is a matter
of an individual conscious commitment to excellence and willingness to be self-critical
and constantly reevaluate assumptions and presuppositions. Such consciousness
requires a degree of metacognitive skills and humility, things not often taught in med-
ical school or in postgraduate education. And, if present in physicians after medical
school, these attributes often do not survive postgraduate education.
THE PRINCIPLES
Define Success
The first principle is to carefully define success. Clearly, DBS therapy requires a team
effort including the proper selection of patient, typically the purview of the neurologist;
proper implantation of the DBS system, typically the combined effort of the surgeon
and the intraoperative neurophysiologist; and effective postoperative management,
14. Future Intraoperative Neurophysiological Monitoring / / 319
again typically the purview of the neurologist. Certainly, there are neurosurgeons
who are excellent in patient selection and postoperative managements, but the typi-
cal training of a neurosurgeon does not provide for gaining sufficient expertise in the
pharmacological management that is critical to both patient selection and postopera-
tive management. It is a historical fact that DBS has been portrayed as a neurosurgical
therapy, but it is not. The therapy is not established with the implantation of the DBS
system but rather with the effective postoperative initiation and programming of the
DBS system, typically in the hands of nonsurgeons.
The mindset that DBS is a neurosurgical therapy belies the necessity of carefully
defining success. The success of DBS is not the placement of the DBS lead in the appro-
priate target but rather a patient whose disability if sufficiently improved. Certainly,
accurate placement of the DBS lead is a necessary prerequisite to success as just defined,
but it is a means and not an end.
The practical result of recognizing the accurate surgical placement of the DBS lead
as a means to an end requires recognition of the needs of the physician and healthcare
professional who will effect the ends that define success. Thus, the goal of surgery is
not to place the DBS lead in the appropriate anatomical target but in the physiological
target in a manner that facilitates the postoperative programming. Thus, the endpoint
of DBS lead implantation is to provide a sufficiently wide therapeutic window between
the stimulation intensity that improves symptoms and the stimulation intensity that
causes adverse effects (see c hapter 9). Establishing a sufficiently wide therapeutic win-
dow requires accommodating the “worst case” scenarios that may confront the physi-
cian and healthcare professionals responsible for the postoperative care.
Recognition of the true definition of success also means conducting the DBS lead
implantation surgery in such a manner as to instill confidence in the physician and
healthcare professionals responsible for the postoperative management. Confidence is
needed because postoperative management is complex and labor intensive. The effort
to meet the complexity and intensity has to be balanced against the expectation of suc-
cess. Confidence in the DBS lead implantation likely will lead to greater perseverance
on the part of those providing the postoperative care in problematic patients. In this
author’s opinion, microelectrode recordings provide the needed level of confidence
more so than other techniques (see c hapter 1).
The importance of success depends on the degree to which the patient’s disability
is improved. Improvements less than what was reasonable to expect, in some respects,
is worse than if no improvement resulted. The decision to undergo DBS lead revi-
sion surgery depends on the risk-to-benefit ratio. The benefit is directly related to the
degree of disability. In the case of continued severe disability that would attend no
improvement following surgery, the potential benefits are correspondingly high and
the risk-to-benefit ratio is stilled in favor of DBS lead revision surgery. In the case of
partial improvement, the disability is less and consequently, the potential benefit is less.
The effect on the risk-to-benefit ratio may be such that DBS lead revision surgery will
be deferred and the patient continue to suffer some disability, a situation termed the
“tyranny of partial benefit.”
Define the Target
The second principle is to describe the target as precisely as possible. The first presump-
tion requires jettisoning the notion that the target is anatomical. Rather, it is physiolog-
ical, as to improve the symptoms and disabilities associated with disease is to restore
more normal physiology. This is not to say that the anatomical structure and the physi-
ological target cannot be synonymous. However, experience to date suggests that they
are not. For example, currently it is insufficient to state that the target is the subtha-
lamic nucleus; rather it is the sensorimotor regions (and avoidance of the limbic and
cognitive regions). For cervical dystonia, the target is the head representation in the
sensorimotor homunculus of the globus pallidus interna and not just the sensorimotor
region of the globus pallidus interna.
Assess Target Relative to Proposed Intervention
The third principle is to address how the nature of the target relates to the interven-
tion proposed, as this will establish the specifications of how the intervention will be
employed. As experience has demonstrated, the typical volume of tissue activation with
DBS is approximately the size of the sensorimotor region of the subthalamic nucleus.
Thus, finer resolution than just the sensorimotor region is not relevant. However, in the
case of the globus pallidus interna and the ventral intermediate nucleus of the thala-
mus, the sensorimotor region is larger than the volume of tissue activation. Thus, the
placement of the DBS lead implies a choice of where within the sensorimotor region
would be optimal. In the case of the ventral intermediate nucleus of the thalamus and
the globus pallidus interna, the precise and appropriate homuncular representation is
targeted.
The issue of the spatial extent of the structures intended to be stimulated and those
to be avoided depends on how the spatial extent of the stimulation can be controlled. In
currently commercially available DBS leads, there is some control in the axis contain-
ing the contacts by controlling which are active cathodes. This control allows for some
A B
Lead Lead
Structure A Structure A
Electrical field
Electrical field
Structure B Structure B
FIGURE 14.1 Schematic representation of the effects of lack of control of the stimulus electri-
cal field in the plan orthogonal to the axis containing the electrical contacts in the lead. The
electrical field is depicted as a disc emanating from a contact on the lead. In A, the electrical
field (represented by lightning bolts) affects structure A and B, producing adverse effects.
Moving the electrical field up the long axis avoids stimulating structure B, but structure A con-
tinues to be stimulated, producing adverse effects.
control of the spatial extent, which consequently affects the specificity and sensitivity
issues related to localization for DBS lead implantation (Montgomery 2010). However,
there is very little, if any, control of the spatial extent in the plane orthogonal to the long
axis of the lead (Figure 14.1).
In the future, it may be possible to have segmented contacts where the cathode
has a restricted effect in the plane orthogonal to the axis of the contacts on the lead
(Figure 14.2). Thus, having some control of the spatial extent in all axes could lessen the
required specificity and sensitivity in DBS lead implantation and consequently, change
the effectiveness of different methods for target localization. For example, a DBS lead
placed too anterior in the subthalamic nucleus and resulting in electrical current spread
to the posterior limb of the internal capsule could be compensated by using the cath-
odal contact that projects posterior.
Assess Target Relative to Nontarget Areas
The fourth principle is to address the nature of the target relative to the nontarget areas
in the vicinity. If the target is immediately adjacent a structure whose stimulation pro-
duces significant adverse effects, the tolerance allowed for targeting will be very differ-
ent from the tolerance should the immediately adjacent structures not cause significant
adverse effects when stimulated. The globus pallidus interna and the subthalamic
nucleus are cases in point, with respect to cognitive and psychological adverse effects.
A B
Structure A Structure A
Lead Lead
Electric field Electric field
Structure B Structure B
FIGURE 14.2 Schematic representation of the effects of lack of control of the stimulus electrical
field in the plan orthogonal to the axis containing the electrical contacts in the lead. The elec-
trical field is depicted as a disc emanating from a contact on the lead. In A, the electrical field
(represented by lightning bolts) affects structure A, producing adverse effects, while also stim-
ulating structure B to produce benefit. Moving the electrical field up or down would not resolve
the adverse effects. A segmented contact that faces only one direction in the plane orthogonal
to the long axis of the lead would be able to “project” the electrical field away from structure
A, thereby avoiding side effects, and at the same time stimulate structure B to provide benefit.
These are more common with subthalamic nucleus DBS compared to DBS of the glo-
bus pallidus interna. The reason may be that the limbic and cognitive regions within
the globus pallidus interna are further distant than is the case within the subthalamic
nucleus.
In some situations, the fourth principle necessitates a change in the target from
the physiological system that mediates therapeutic efficacy to the systems that medi-
ate adverse effects. For example, the primary objective of globus pallidus interna
DBS lead implantation is to avoid the posterior limb of the internal capsule. Thus,
the target becomes the anterior border of the posterior limb of the internal capsule.
Similarly, the anterior border of the tactile ventral caudal thalamus becomes the
primary target of implantation of the DBS lead in the ventral intermediate thalamic
nucleus.
Assess Method of Visualizing the Target
The fifth principle follows from the need to operationalize the primary target based on
the methods necessary to identify the primary target. As there are many ways in which
a target may be “seen” such as by measuring proton densities in MRI, radiodensities
in CT, electrical or optical impedance, or sensorimotor driving of neuronal extracel-
lular action potentials, the question is “What is the nature of the target in terms of the
methods used for ‘seeing’?” For example, the proton density of the sensorimotor region,
the target, is the same as the limbic and cognitive regions, the definite nontarget (in
other words, structures to be avoided). Thus, MRI scanning cannot “see” the target
within the subthalamic nucleus as different from the nontargets. However, recording
of extracellular action potentials while conducting sensorimotor stimulation can “see”
the target.
An extension of the corollary described above is if the direct target cannot be “seen”
it is possible that another landmark that can be “seen” may be used as a surrogate for the
actual target. Indeed, this is the reason for the use of stereotactic coordinates relative
to the midpoint of the line connecting the anterior and posterior commissures, AC and
PC, respectively. Alternatively, special MRI sequences allow more direct visualization
of the subthalamic nucleus, often based on the paramagnetic effect of its high iron (Fe)
concentration. The ventral and lateral borders of the visualized subthalamic nucleus
can act as a surrogate for the sensorimotor region. Similarly, the posterior limb of the
internal capsule abutting the globus pallidus interna could be used as the surrogate
marker of the sensorimotor region of the globus pallidus interna, which is just anterior
to the posterior limb of the internal capsule. Perhaps, high resolution functional MRI
(fMRI) may be able to directly visualize the sensorimotor regions. However, currently
there is evidence that at least with respect to the midpoint of the line connecting the
anterior and posterior commissure, the accuracy is insufficient to obviate the need for
microelectrode recordings of extracellular action potentials (see chapter 1).
Assess Method Validity
The sixth principle relates to the quality of the methods used to “see” the target, partic-
ularly with regard to accuracy and precision, which provide some measure of the valid-
ity of the methods within a patient and between patients. There are different types of
validity, but the most important one here is ecological validity. In other words, what is
the validity (in the case of DBS surgery, accuracy of finding the target) in the context in
which the DBS electrode is actually placed into the patient’s brain? Thus, the accuracy
of a stereotactic device in a phantom and how often that target can be reached (preci-
sion) is only partially relevant (the accuracy in the phantom only being the most precise
possible, not the actual accuracy). The accuracy of placing a trajectory in the target
surrogate based on the midpoint of the line connecting the anterior and posterior com-
missure on the preoperative MRI scan is only partially relevant; it represents only the
ceiling effect beyond which no further increase in accuracy is possible. Reasons are
discussed in chapter 1 as to why the preoperative MRI or CT scan is only partially rel-
evant, such as brain shift due to failure to control for air let into the intracranial space.
Some errors may be device errors, that is, due to the intrinsic nature and prop-
erties of the device. However, there are many more sources of error, and focusing
exclusively on device error is misleading—a point not often appreciated. For exam-
ple, a device may reliably get an electrode within 1 mm of the intended spatial tar-
get (accuracy) such as the midpoint of the line between the anterior and posterior
commissure, and do so repeatedly (precision), at least in a phantom. However, if the
actual target, the physiological sensorimotor region of the subthalamic nucleus, for
example, is not precise but highly variable across different patients, then the error
will increase due to both the variability between patients and the error inherent in
the device. To that error must be added error introduced by the surgical techniques
such as obtaining an MRI scan with the patient supine and then operating with the
patient more in a semi-sitting position and thereby allowing brain shift. The surgical
source of error can increase greatly if air is not prevented from entering the skull.
Thus, a device with high accuracy and precision, as it relates to instrumental error,
may not result in optimal placement of the electrodes. Indeed, other methods are
needed.
Certainly, intraoperative MRI and/or CT scans could compensate for the variabil-
ity induced by the surgery, such as intracranial air and brain shift, but then the accuracy
of DBS lead implantation is equal to (or less than) that the accuracy of targeting based
on the midpoint of the line connecting the anterior and posterior commissures. If the
methods of intraoperative MRI and/or CT scan prevent the concurrent use of micro-
electrode recordings of extracellular action potentials, then it is reasonable that intra-
operative MRI and/or CT scans will be less accurate than microelectrode recordings
of extracellular action potentials.
The virtue of the microelectrode recordings is that each subsequent trajectory
is relative to the prior trajectories and not to the preoperatively visualized surrogate
for the target, thereby mitigating, at least to some degree, the effects of brain shift.
Consequently, the advantage of intraoperative MRI and/or CT scans relative to micro-
electrode recordings seems doubtful. One could argue that at least intraoperative MRI
and/or CT scans could identify the degree of brain shift and perhaps make intraop-
erative microelectrode recordings more efficient. However, if there is significant brain
shift, particularly as a result of intracranial air, then one would not want to place the
DBS lead anyways, even if the microelectrode recordings were able to identify the tar-
get. Large amounts of brain shift are likely to cause the DBS lead to migrate even if
intraoperative microelectrode recordings allowed the DBS lead to be optimally placed
initially. Intraoperative MRI and/or CT scans combined with intraoperative micro-
electrode recordings of extracellular action potentials may offer the best alternative;
however, such methods would have to be judged in terms of risks and costs compared
to outcomes.
INTRAOPERATIVE MONITORING AS A DIAGNOSTIC TEST
Ultimately, any intraoperative monitoring is basically a diagnostic test. In the case of

DBS, the diagnostic question is whether a particular trajectory is diagnostic or predic-
tive of a good clinical response postoperatively (see the first principle described above).
Consequently, intraoperative monitoring should be viewed as a diagnostic test with all
the inherent limitations of any diagnostic test. As can be appreciated, diagnostic tests
are problematic. Every clinician is confronted with considering whether an abnormal
diagnostic test is a true positive (the entity actually is abnormal) or a false positive (the
entity actually is normal). Similarly, every clinician needs to consider whether a nor-
mal diagnostic test is a true negative (the entity actually is normal) or a false negative
(the entity actually is abnormal). In intraoperative neurophysiological monitoring, the
physiologist is confronted with the question of whether the findings are a true posi-
tive (the trajectory actually is a good trajectory that has a high probability of providing
therapeutic DBS) or a false positive (while it looks like a good trajectory, DBS in this
trajectory is unlikely to produce a benefit). Alternatively, the physiologist is confronted
with a decision of a negative result, meaning that the results from the monitoring are
not indicative of a good trajectory, and whether the negative result is a true negative or
a false negative.
Consider the situation where the microelectrode recordings in a specific trajectory
failed to encounter any neuronal extracellular action potentials. The physiologist has to
determine whether or not this represents an intracerebral hematoma. However, there is
nothing unique about the lack of recordings of neuronal extracellular action potentials
that is pathognomonic for intracerebral hematoma. It also could indicate being present
in the posterior limb of the internal capsule or an electrode failure. Thus, the physiolo-
gist has to decide whether his or her conclusion of an intracerebral hematoma is a true
positive or false positive. A bit of detective work is required, which is made easier by an
understanding of the nature of the clues and a critical mind. For example, the neuro-
surgeon could have given a medication, unbeknownst to the physiologist, that would
suppress neuronal activities. The physiologist might falsely conclude, after determining
that the electrode properties were nominal, that there was an intracerebral hematoma.
The surgery would be aborted, and in retrospect it would be determined that the medi-
cation had silenced the neuronal activities. The patient would have to undergo a second
operation.
Next consider a person who develops a test for determining whether or not the
stimulating electrode is in the posterior limb of the internal capsule by studying the
cortical electroencephalographic potentials evoked by stimulation. A critical investiga-
tor would ask at least two questions. First, what is the probability that an appropriate
evoked potential truly is associated with the electrode being in the posterior limb of
the internal capsule (probability of a true positive)? Second, what is the probability
that a lack of the appropriate evoked potential truly is associated with the electrode
not being in the posterior limb of the internal capsule (probability of a true negative)?
These questions address the issues of specificity, meaning the probability that an appro-
priate evoked potential actually is associated with the electrode being in the posterior
limb of the internal capsule. Sensitivity relates to the probability that an appropriate
evoked potential may not be associated with the electrode being in the posterior limb of
the internal capsule. For example, the electrode may not be in the posterior limb of the
internal capsule but may stimulate axons projecting from the cortex to the subthalamic
nucleus which could result in a false positive evoked potential. It is important to note
that while specificity and sensitivity are linked they are not the reciprocal of each other.
It is not sufficient to only consider the specificity and sensitivity. Rather one has to
consider the specificity and sensitivity in the context of what is called the prior prob-
ability. For example, consider a predictive test for persons at risk for Parkinson’s disease
that is 95% specific and 95% sensitive. At first, this would appear to be a very good test.
However, if the incidence of persons at risk for Parkinson’s disease is only 5%, then there
will be an equal number of true and false positives. Thus the positive predictive value (the
ratio of true to false positives) is only 50%. The incidence of persons at risk is referred to
as the prior probability and is often used in Bayesian analyses. Similar considerations
hold for the negative predictive value, which is the ratio of true to false negatives.
These considerations apply to intraoperative neurophysiological monitoring. In the
case of the evoked potential described above, the prior probability would be the prob-
ability of placing the electrode initially into the posterior limb of the internal capsule
independent of the use of the evoked potential. For example, if using MRI for the initial
targeting results in a very low probability of the electrode being placed on the posterior
limb of the internal capsule, then the prior probability correspondingly would be low
and the probability of a false positive would be high; perhaps sufficiently high as to not
recommend the use of the evoked potential.
Ultimately, the purpose of intraoperative neurophysiological monitoring is to iden-
tify the trajectory that has the highest probability of producing a therapeutic benefit,
and most often this depends on knowing where to place a subsequent trajectory when
the prior trajectory does not meet the criteria for optimal location. Thus, the utility of
identifying the posterior limb of the internal capsule only makes sense if it can help
direct the direction of a more optimal trajectory, as identifying the posterior limb of
the internal capsule already demonstrates that the current trajectory is not optimal. In
the case of the subthalamic nucleus, the posterior limb of the internal capsule borders
the subthalamic nucleus on three sides (lateral, anterior, and ventral). Thus, to be use-
ful, the evoked potential would have to be able to differentiate the spatial relation of
the identified posterior limb of the internal capsule relative to the subthalamic nucleus.
Finally, any intraoperative neurophysiological monitoring test must be considered
in the context of the entire surgical procedure. Clearly there are the costs associated
with the time and materials, but there is also the cost of the intraoperative neurophysi-
ologist, which is directly proportional to the knowledge and skills required. A more
complex test may have better localizing value but may not be feasible or the incremental
benefit only of marginal value.
It is important to note that the discussion above does not mean that stimulation
evoked potentials in the electroencephalogram could not be useful in the future.
Rather, the discussion described above serves to demonstrate the necessary issues that
any such diagnostic intraoperative neurophysiological monitoring technique would
have to face. The discussion, importantly, sets the criteria that any intraoperative neu-
rophysiological monitoring must meet before acceptance.
POSSIBLE APPLICATIONS OF THE PRINCIPLES
The range of targets is expanding, and the methods for localization for DBS lead implan-
tation still follow from the principles described above. Consider DBS of the anterior
limb of the internal capsule for obsessive-compulsive disorder, currently approved by
the USFDA under a humanitarian device exemption (HDE). Some centers have used
intraoperative neurophysiological monitoring of neuronal extracellular action poten-
tials for localizing the anterior limb of the internal capsule. The rationale is that no
neuronal extracellular action potentials could be encountered. Doing so would sug-
gest that the electrode is not in the anterior limb of the internal capsule. Further, as
the microelectrode moves ventrally the anterior limb of the internal capsule becomes
fenestrated with increasing “islands” of gray matter that reveal themselves by the pres-
ence of neuronal extracellular action potentials. This information can be used for estab-
lishing the depth for the subsequently DBS lead.
Similarly, microelectrode recordings could be performed in DBS of the subgenu
cingulum for the treatment of depression, which is undergoing clinical trials. In this
case, the effort is to demonstrate neuronal extracellular action potentials in order to be
sure that the trajectory is not too medial in the subcortical gray matter or too lateral in
the interhemispheric fissure.
This author has limited experience with DBS of the pedunculopontine nucleus
but has used microelectrode recordings and microstimulation. The rationale is that
pedunculopontine nucleus is adjacent to the ascending medial lemniscus, which, if
inadvertently stimulated, could produce intolerable paresthesias. In addition, the
pedunculopontine nucleus is adjacent to the brachium conjunctivum, which conducts
outputs from the deep cerebellar nuclei to the brainstem and thalamus. The concern is
that inadvertent stimulation of the brachium conjunctivum could cause gait ataxia, as
has been observed with subthalamic nucleus DBS.
Specific neuronal discharge properties also provide reassurance as to the optimal
location. Two general discharge types are found in the pedunculopontine nucleus. The
first is a relatively high discharge rate with extracellular action potentials of relatively
long durations and a lower discharge rate with shorter extracellular action potentials.
Most importantly, the neurons demonstrate sensorimotor driving.
The range of technologies for intraoperative neurophysiological monitoring is
expanding and thus, the importance of intraoperative neurophysiological monitoring
is likely to expand. Many will focus on electrophysiological techniques, particularly
local field potentials, which typically are easier to record. For example, it may be pos-
sible to record local field potentials for the DBS lead, and once the optimal location
is identified the DBS lead can be left in place (Chen, Pogosyan, Zrinzo et al. 2006).
These studies have demonstrated increased power in the high beta frequencies (15–30
Hz) within the subthalamic nucleus and perhaps a difference between the dorsal and
ventral regions (Trottenberg, Kupsch, Schneider et al. 2007). However, the positive
and negative predictive value for localizing the physiologically defined target, that is, at
least 5 mm of sensorimotor driving, is unknown.
Local field potentials represent the summed activity of postsynaptic action poten-
tials in the dendrites and neuronal cell bodies over a volume that is large relative to the
volume of tissue recorded from microelectrodes or semi-microelectrodes. Thus, loss
of resolution becomes an issue. For example, computational modeling suggests that
the spatial extent of the local field potential may extend over several millimeters, rais-
ing questions as to spatial resolution and specificity (Lempka and McIntyre 2013). To
some extent the low resolution can be improved by special techniques such as triangu-
lating the source of the electrical potentials between multiple recording contacts (as
is done with bipolar recordings in standard electroencephalography). However, the
efficacy of such approaches is unknown.
An exciting development is the use of optogenetics which involves transfecting neu-
rons with genetic material that produces light sensitive ion gated or G-protein-coupled
ionic conductance channels to initiate action potentials (see chapter 4). An example
would be the gene for channelrhodopsin derived from green algae. This gene, with vari-
ous promoters to control expression, can be delivered into neurons using viral vectors.
Once the channelrhodopsin or others are expressed on the neuronal membrane, shin-
ing a light (such as through a fiberoptic electrode) can cause changes in the neuronal
membrane, which may either depolarize, leading to an increased probability of an action
potential, or hyperpolarize, reducing the probability of an action potential. One of the
potential advantages is that the selectivity of action can be controlled by the promoters,
which can be specific to a type of neuron (Gradinaru, Mogri, Thompson et al. 2009). The
nature of the selectivity of the promoters could dramatically change the physical spatial
resolution of the device to activate the channelrhodopsin. Clearly, there are many tech-
nological obstacles to overcome, but the potential is exciting and hopeful.
Ultimately, optogenetic techniques likely will depend on the precise patterns of
excitations evoked in the stimulated structures. Thus, optogenetic techniques will have
much in common with other electrophysiologically based techniques such as DBS.
SUMMARY
The potential for electrophysiological based therapies is limited only by the imagina-
tion. Consequently, there will be an expanding need to understand the fundamental
principles of electrophysiology both for control (such as recording) and effecting thera-
pies (such as stimulation). While the specific techniques may vary fundamentally, all
these methods will entail operating at the spatially and temporal resolutions not gen-
erally available with pharmacological methods. This will be true no matter what the
methods are for recording or eliciting electrical activities in neurons.
REFERENCES
Chen C, Pogosyan A, Zrinzo L, et al.: Intra-operative recordings of local field potentials can help localize
the subthalamic nucleus in Parkinson’s disease surgery. Experimental Neurology 198: 214–221, 2006.
Gradinaru V, Mogri M, Thompson KR, et al.: Optical deconstruction of parkinsonian neural circuitry.
Science 324(5925): 354–359, 2009.
Lempka S and McIntyre C: Theoretical analysis of the local field potential in DBS applications. PLoS One
8: e59839, 2013.
Montgomery EBJ: Deep Brain Stimulation Programming: Principles and Practice. Oxford, Oxford
Trottenberg T, Kupsch A, Schneider G-H, et al.: Frequency-dependent distribution of local field potential
activity within the subthalamic nucleus in Parkinson’s disease. Experimental Neurology 205: 287–
291, 2007.
APPENDIX A
SUBTHALAMIC NUCLEUS DEEP BRAIN

STIMULATION ALGORITHM
DESCRIPTION AND DISCLAIMER
The algorithm described here utilizes a heuristic device in which the microelectrode
trajectories are first divided into and upper and lower trajectory (Figure A.1). The lower
trajectory is divided into an initial, middle, and bottom region. The algorithm uses the
characteristics of the neuronal activities in each of these segments to relate the micro-
electrode trajectory to the regional anatomy. These segments are of relative lengths,
which may vary in different microelectrode trajectories. The neuronal characteristics of
each segment provide a guide to the relative location in the patient’s anatomy surround-
ing the trajectory. Often, the interpretation is that the site suggests trajectories that are
too medial, too lateral, too posterior, too anterior, too shallow in the sagittal plane, or
too shallow in the coronal plane. The intraoperative neurophysiologist should look to
interpretations that are most consistent across the four segments.
The algorithm is a distillation of the author’s experience and has not been subjected
to rigorous prospective validation because of the complexity and expense such valida-
tion would entail. Note, this algorithm is educational in nature and is not intended to
direct the care of any individual patient. The physician, surgeon, and healthcare profes-
sional must exercise their judgment in using the algorithm and assume full and exclu-
sive responsibility for the care of the patient.
The criteria of a physiologically optimal target are: (1) 5 mm or more of
sensorimotor-driven neuronal activities, (2) no adverse effects with macrostimula-
tion, and (3) some improvement in symptoms (see chapter 9). For interpretation of the
results from macrostimulation through the DBS lead see chapter 12.
331
332 / / A ppendix A
Dorsal Coronal plane
Medial Lateral
Ventral
Upper trajectory
Low frequency, transient, low density
Th
Initial lower trajectory
Low frequency, transient, very low density
ZI
STN Middle lower trajectory
High frequency, sustained, high density,
irregular, sensori-motor driven
SNr
Bottom lower trajectory

High frequency, sustained, low density,
regular
Fp 4.0
FIGURE A.1 Schematic representation of the heuristic employed in the algorithms to relate
findings during microelectrode recoding trajectories. The anatomy represents structures in
the coronal plane approximately 4 mm anterior to the midpoint of the line connecting the
anterior and posterior commissures (AC-PC line). As can be seen in this trajectory, the upper
trajectory occupies the thalamus (Th), the initial lower trajectory contains the zona incerta (ZI),
the middle lower traject ory represents the subthalamic nucleus, and the bottom trajectory is
in the substantia nigra pars reticulata (SNr) for this specific trajectory. This trajectory is typi-
cal of those that contain the physiologically defined optimal target (see chapter 10). Source:
Modified from Schatlenbrand and Wahren (1997).
Subthalamic Nucleus DBS Algorithm
1. Upper trajectory

1.1 No neuronal activity
1.1.1 Consider electrode failure
1.1.1.1 Check impedance
1.1.1.1.1 If too high or too low may suggest electrode failure
1.1.1.2 Tap on wire leads from electrode to see if typical artifact is produced
1.1.1.2.1 If silent, inputs may be in electrical contact with ground, there may
be a broken electrical connection or a failure in the amplifier system
1.1.2 May be in the posterior limb of the internal capsule from a trajectory that is
too shallow in the coronal or sagittal plane
1.1.2.1 Microstimulation or macrostimulation may produce tonic muscle
contraction
1.1.3 Consider an intracerebral hematoma
Appendix A / / 333
1.1.3.1 Inspect electrode and guide cannulas for blood as these are removed
1.1.3.2 The brain may wipe the cannula and electrodes of blood on the surface
1.1.3.2.1 Place the stylet into the cannula to see if blood can be expelled
1.1.3.2.2 Extend the tip of the microelectrode or semi-microelectrode to see if
blood can be expelled
1.2 Irregular, low frequency, transient, low density within recording site, low den-
sity in trajectory, and no sensorimotor driving
1.2.1 Consistent with anterior thalamus
1.2.1.1 Expected at this height in the trajectory
1.3 Irregular, low to moderate frequency, transient, low density within recording
site, low density in trajectory, and with sensorimotor driving but more with active than
passive joint rotations
1.3.1 Consistent with ventral oral posterior thalamus
1.3.1.1 Suggests that trajectory may be too posterior depending on findings in
the lower trajectory
2. Lower trajectory

2.1. Initial lower trajectory
2.1.1. Irregular, low frequency, very transient, possibly bursting, very low density
within recording site, very low density in trajectory, and no sensorimotor
driving
2.1.1.1. Consistent with zona incerta and expected at this height in the trajectory
2.1.1.1.1. Width of zona incerta important
2.1.1.1.1.1. Small width suggests posterior and/or medial zona incerta and
the trajectory may be too posterior and/or medial depending
on the remainder of the trajectory
2.1.1.1.1.2. Wide zona incerta suggests anterior and/or lateral zona
incerta and the trajectory may be too anterior and/or lateral
depending on the remainder of the trajectory
2.1.2. No neuronal activity
2.1.2.1. Consider electrode failure
2.1.2.1.1. Check impedance
2.1.2.1.1.1. If too high or too low may suggest electrode failure
2.1.2.1.2. Tap on wire leads from electrode to see if typical artifact is produced
2.1.2.1.2.1. If silent, inputs may be in electrical contact with ground,
there may be a broken electrical connection or a failure in the
amplifier system
334 / / A ppendix A
2.1.2.2. May be in the posterior limb of the internal capsule from a trajectory
that is too shallow in the coronal or sagittal plane
2.1.2.2.1. Microstimulation or macrostimulation may produce tonic muscle
contraction
2.1.2.3. Consider an intracerebral hematoma
2.1.2.3.1. Inspect electrode and guide cannulas for blood as these are removed
2.1.2.3.2. The brain may wipe the cannula and electrodes of blood on the surface
2.1.2.3.2.1. Place the stylet into the cannula to see if blood can be expelled
2.1.2.3.2.2. Extend the tip of the microelectrode or semi-microelectrode
and examine for blood
2.2. Middle lower trajectory
2.2.1.1. Irregular, high frequency, sustained, moderate to high density within
site, high density in trajectory
2.2.1.1.1. Encountered relatively high in trajectory and <5 mm of sensorim-
otor driving, may be too lateral depending on the length of this
type of neuronal activities
2.2.1.1.2. Encountered relatively low in trajectory and more moderate dis-
charge frequency and >5 mm of sensorimotor driving, may be too
medial, anterior, or posterior
2.2.1.1.2.1. If no or poor sensorimotor driving, too medial
2.2.1.1.2.2. If paresthesias with stimulation, too posterior
2.2.1.1.2.3. If tonic muscle contraction, too anterior
2.2.1.1.3. Encountered approximately ≥5 mm above the image-guided bot-
tom of the subthalamic nucleus, trajectory with ≥5 mm of robust
sensorimotor driving, no side effects to stimulation, and improve-
ment in symptoms with stimulation (optional), may be appropri-
ate trajectory
2.2.1.2. No neuronal activity
2.2.1.2.1. May be too lateral and/or anterior in the posterior limb of the
internal capsule
2.2.1.2.1.1. Risk if trajectory is too shallow in the coronal plane with a
very lateral entry point, for example when trying to avoid the
lateral ventricle
2.2.1.2.1.2. Stimulation may produce tonic contraction
2.2.1.2.2. Check for electrode failure
2.2.1.2.2.1. Check impedances
2.2.1.2.3. Consider intracerebral hematoma
Appendix A / / 335
2.2.1.2.3.1. Examine cannula and electrode for evidence of blood

2.2.1.2.3.2. Withdrawing cannula through brain may wipe blood off
2.2.1.2.3.3. Insert stylet to see if it pushes blood out
2.2.1.2.3.4. Extend electrode tip to see if it pushes out blood
2.2.1.2.3.5. Check top of cannula to see if blood is exiting
2.3. Bottom lower trajectory

2.3.1. Regular, high frequency, sustained, low density in site, low density in tra-
jectory, no sensorimotor driving (may respond to eye movements)
2.3.1.1. Consistent with substantia nigra pars reticulata
2.3.1.1.1. Important to demonstrate to assure trajectory actually in the sub-
thalamic nucleus and not in medial globus pallidus interna where
the neuronal activities just above are nearly identical.
2.3.1.1.1.1. Can happen if there is significant brain shift secondary to intra-
cranial air
2.3.2. Moderately regular, moderate frequency, sustained, low density in site, low
density in trajectory, and sound of the action potentials of lower frequency
due to longer extracellular action potential
2.3.2.1. Consistent with border cells at the bottom of the medial globus pal-
lidus interna
2.3.2.1.1. Trajectory too posterior and suggests significant brain shift likely
related to intracranial air
2.3.3. No or rare neuronal activity
2.3.3.1. May be too lateral and/or anterior in the posterior limb of the internal
capsule
2.3.3.1.1. Risk if trajectory is too shallow in the coronal plane with a very
lateral entry point, for example when trying to avoid the lateral
ventricle
2.3.3.1.2. Stimulation may produce tonic contraction, interference with
speech, or conjugate deviation of the eyes (see chapter 12)
2.3.3.2. Check for electrode failure
2.3.3.2.1. Check impedances
2.3.3.3. Consider intracerebral hematoma
2.3.3.3.1. Examine cannula and electrode for evidence of blood
2.3.3.3.2. Withdrawing cannula through brain may wipe blood off
2.3.3.3.3. Insert stylet to see if it pushes blood out
2.3.3.3.4. Extend electrode tip to see if it pushes out blood
2.3.3.3.5. Check top of cannula to see if blood is exiting
336 / / A ppendix A
3. DBS macrostimulation at ventral-most (deepest) aspect of the neuronal activities

identified as the sensorimotor subthalamic nucleus
3.1. Paresthesias
3.1.1. Too posterior
3.2. Tonic contraction
3.2.1. Check threshold to tonic contraction at ventral-most contact as cathode
(negative) and with the dorsal-most contact as cathode
3.2.1.1. If thresholds the same
3.2.1.1.1. Suggests trajectory is parallel to the posterior limb of the internal
capsule
3.2.1.1.2. Typically suggests the trajectory is too anterior or too lateral if the
trajectory is shallow in the coronal plane
3.2.1.1.3. If DBS lead too shallow in the coronal plane, such as to avoid the
lateral ventricle, DBS lead may be parallel to the posterior limb of
the internal capsule
3.2.1.2. If threshold at the ventral-most contact lower than upper contact
3.2.1.2.1. DBS lead too ventral or too lateral (unless the lead is very shallow
in the coronal plane, see 3.2.1.1.2.
3.3. Skewed deviation of the eyes producing disconjugate gaze and diplopia
3.3.1. Suggests stimulation too medial affecting the fascicles of the oculomotor nerve
3.4. Speech and language abnormalities

3.4.1. If tonic facial contraction is seen with higher stimulation intensity
3.4.1.1. May be related to spread of stimulation current to the corticospinal tract
3.4.2. If not associated with facial contraction with higher stimulation intensity
3.4.2.1. May be intrinsic to stimulation of the subthalamic nucleus
3.4.2.2. Repositioning of the DBS lead unlikely to result in improvement
3.5. Unusual or strange feeling

3.5.1. May be associated with spread of stimulation current to the hypothalamus
3.5.1.1. Stimulation may be too medial and/or too dorsal
3.5.2. May be seen with stimulation of the medial globus pallidus interna if there
is significant brain shift posteriorly
REFERENCE
Schaltenbrand G, Wahren W: Atlas for Stereotaxy of the Human Brain. Stuttgart, Thieme, 1977.
APPENDIX B
VENTRAL INTERMEDIATE THALAMIC

DEEP BRAIN STIMULATION
ALGORITHM
DESCRIPTION AND CAVEAT
The algorithm below is to aid in the optimal placement of DBS leads in the ventral inter-
mediate thalamus using microelectrode recordings. Semi-microelectrode recordings
may not have sufficient spatial resolution to precisely identify the homuncular repre-
sentation; however, the algorithm can be adapted for semi-microelectrode recordings.
Also, there is a guide to aid in the interpretation of DBS macrostimulation. The algo-
rithm is a distillation of the author’s experience and has not been subjected to rigorous
prospective validation because of the complexity and expense such validation would
entail. Note, the algorithm is educational in nature and is not intended to direct the
care of any individual patient. The physician, surgeon, and healthcare professional must
exercise their judgment in the utilization of the algorithm and assume full and exclusive
responsibility for the care of the patient.
The approach used in the algorithm is to divide the microelectrode trajectory into
an upper and lower section (Figure B.1). The upper section represents the dorsal tier
of the thalamus. Typically, the neuronal activity in the upper section does not convey
much information other than the fact that if neurons can be clearly recorded the micro-
electrode recording system would appear to be working. The lower section relates to
the ventral tier nuclei that contain various nuclei whose neuronal behavioral responses
have localization value.
The physiologically defined optimal target is the most anterior border of the tactile
ventral caudal thalamus. Most often, the most anterior border can only be detected by
bracketing (Figure B.1). In this case, the posterior trajectory detects neuronal activities
337
338 / / A ppendix B
Upper trajectory Upper trajectory

Low frequency, low density within site, Low frequency, low density within site,
low density in trajectory, transient, low density in trajectory, transient, irregular
irregular
Initial lower trajectory Initial lower trajectory

Low to moderate frequency, low density Low to moderate frequency, low density
within site, low density in trajectory, Vc within site, Low density in trajectory,
Vim persistent, irregular
persistent, irregular
Middle lower trajectory Middle lower trajectory

Low to moderate frequency, low density Low to moderate frequency, Low density
within site, low density in trajectory, within site, low density in trajectory,
persistent, irregular persistent, irregular
Bottom lower trajectory Bottom lower trajectory

Low to moderate frequency, low density Low to moderate frequency, Low density
within site, low density in trajectory, Dorsal within site, Low density in trajectory,
persistent, irregular persistent, irregular
Posterior Anterior 16.0
Ventral
FIGURE B.1 Schematic of a saggital section through the thalamus containing the ventral cau-
dal nucleus (Vc) and the ventral intermediate nucleus (Vim) showing two microelectrode tra-
jectories. As can be seen, the trajectories ares organized into an upper and lower half and
the lower half is divided into the initial, middle and bottom one-third. The purpose of the two
trajectories is to bracket the anterior border of the tactile Vc nucleus. The trajectories show the
repsentative structures often associated with each segment. Modified from Schaltenbrand G
and Wahren W: Atlas for Stereotaxy of the Human Brain. Stuttgart, Thieme, 1977.
related to tactile stimulation while another trajectory just anterior does not. Thus, one
can conclude that the anterior border of tactile ventral caudal thalamus lies between
the two trajectories. Occasionally, a trajectory will demonstrate only a very small
extent of neuronal activity driven by tactile stimulation (Figure B.2). In that case, the
intraoperative neurophysiologist may infer that the trajectory is very near the anterior
border of the tactile ventral caudal thalamus and may infer that the optimal trajectory
for the DBS lead is 2–3 mm anteriorly, provided it is in the sagittal plane appropriate to
the homuncular representation.
The ventral nuclei are organized as stacked “slabs” in the horizontal plane where the
short axis of the slabs are oriented in an anterior-posterior direction. The most caudal
nucleus of interest is the tactile ventral caudal thalamus, whose neurons respond to light
touch. Just anterior is the anterior region of the ventral caudal thalamus, whose neurons
respond to joint rotation but not tactile stimulation nor muscle palpation. Next in the
anterior direction is the ventral intermediate thalamus, whose neurons respond to joint
rotation and muscle palpation. Finally, the most anterior nucleus of concern, which
would not be unusual to encounter, is the ventral oral posterior thalamus. Neurons here
respond to joint rotation but are more active when the patient actively produces the
joint rotation rather than with passive rotation exerted by the examiner.
Appendix B / / 339
Upper trajectory
Low frequency, low density within site,
low density in trajectory, transient,
irregular
Initial lower trajectory

Low to moderate frequency, low density
within site, Low density in trajectory, Vc
persistent, irregular Vim
Middle lower trajectory

within site, Low density in trajectory,
persistent, irregular

within site, Low density in trajectory,
persistent, irregular, Tactile responsive neurons only Dorsal
at the very bottom of the trajectory
Posterior Anterior 16.0
Ventral
FIGURE B.2 Schematic of a saggital section through the thalamus containing the ventral cau-
dal nucleus (Vc) and the ventral intermediate nucleus (Vim) showing a microelectrode trajec-
toys. As can be seen, the trajectories ares organized into an upper and lower half and the lower
half is divided into the initial, middle and bottom one-third. In this case, the tactile region of
the Vc thalamus is found at the very bottom of the trajectory suggesting that the anterior bor-
der is very close. In this case, it may not be necessary to have an additional microelectrode
trajectory to bracket the anterior border of the tactile Vc nucleus. The trajectories show the
repsentative structures often associated with each segment. Modified from Schaltenbrand G
Often the angle of the trajectory in the sagittal plane is such that the trajectory may
pass tangentially through more than one ventral tier thalamic nucleus. Consequently,
the lower trajectory is divided into the initial, middle, and bottom of the lower trajectory.
The number of different ventral tier nuclei traversed in a single trajectory is an indication
of the angle of the trajectory in the sagittal plane. Which particular ventral tier nuclei are
encountered provides information as to the anterior-posterior position of the trajectory
The “slabs” are wider in the medial-lateral direction to accommodate the homun-
cular representation. The representation is organized like half of an onion (Figure B.3)
where the core represents the head region, the next layer represents the upper extremity
and the outer layer represents the lower extremity. Which homuncular representations
are encountered in the trajectory provides information as to the angle of the trajectory
in the coronal plane and the relative medial-lateral position.
The approach also is intended to find the anterior border of the tactile region of the
ventral caudal thalamus toward the bottom of the trajectory. The reason is that if the
DBS lead is too close to the tactile region of the ventral caudal thalamus, intolerable
340 / / A ppendix B
Dorsal
Medial Lateral
Upper trajectory Ventral
Initial lower trajectory Lower extremity
Upper extremity
Head
FIGURE B.3 Schematic representation of the homuncular organization of the ventral interme-
diate nucleus of the thalamus (Vim). As can be seen, a microelectrode trajectory shows each
of the segments of the trajectory and the structures that may be associated with each seg-
ment. The key is that one has to traverse the entire Vim to ensure that the proper homuncluar
representations are contained within the trajectory and that other representations, typically
the head, are not in the trajectory.
paresthesias to DBS stimulation may limit the therapeutic effectiveness of the DBS.
The typical trajectory tends to move posterior as that trajectory advances. Thus, the
ventral-most (deepest) part of the DBS lead will be the most posterior of the DBS loca-
tion and hence, closest to the tactile region of the ventral caudal thalamus. Therefore,
it is important that the DBS lead be at least 2–3 mm anterior to the anterior border of
the tactile region of the ventral caudal thalamus at the deepest point in the trajectory.
Some items in the algorithm have “This scenario is highly unlikely” based on the
anatomy. This suggests that there may have been an error in interpreting the character-
istics of the neuronal activities and the recordings or the stimulation results should be
reconsidered.
Ventral Intermediate Thalamic DBS Algorithm
1.0 Upper trajectory

1.1.1.2.1 If silent, inputs may be in electrical contact with ground, there may
be a broken electrical connection, or a failure in the amplifier system
Appendix B / / 341
far too lateral or anterior (the posterior limb of the internal capsule moves
medially as the trajectory moves anteriorly)
1.1.2.1 Microstimulation or macrostimulation may produce tonic muscle con-
traction, alternations in speech, or conjugate eye deviation
1.1.3.2 The brain may wipe the cannula and electrodes of blood on the
surface
1.1.3.2.2 E xtend the tip of the microelectrode or semi-microelectrode to
see if blood can be expelled
1.2 Irregular, very low frequency, transient, low density within recording site, low den-
sity within trajectory
1.2.1 Consistent with dorsal tier thalamus
1.2.2 Expected at this height in the trajectory
1.2.3. Not informative of medial-lateral or anterior-posterior location
2.0 Lower trajectory

2.1 Initial lower trajectory
2.1.1 No neuronal activity
2.1.1.1 Consider electrode failure
2.1.1.1.1 Check impedance
2.1.1.1.2 Tap on wire leads from electrode
2.1.1.2 May be in the posterior limb of the internal capsule from a trajectory
that is far to lateral or anterior (the posterior limb of the internal cap-
sule moves medially as the trajectory moves anteriorly)
2.1.1.2.1 Microstimulation or macrostimulation may produce tonic muscle
contraction
2.1.1.3 Consider an intracerebral hematoma
2.1.1.3.1 Inspect electrode and guide cannulas for blood as these are removed
2.1.1.3.2 The brain may wipe the cannula and electrodes of blood on the surface
2.1.1.3.2.1 Place the stylet into the cannula to see if blood can be expelled
2.1.1.3.2.2 Extend the tip of the microelectrode or semi-microelectrode
to see if blood can be expelled
2.1.2 Irregular, low to moderate frequency, transient, low density within record-
ing site, low density within trajectory
342 / / A ppendix B
2.1.2.1 Consistent with ventral tier nuclei, expected at this height in the
trajectory
2.1.2.1.1 Neurons related to tactile stimulation, not joint rotation or muscle
palpation
2.1.2.1.1.1 Consistent with posterior or tactile ventral caudal thalamus
2.1.2.1.1.1.1 Suggests trajectory very posterior as tactile responsive
neurons encountered high in the trajectory
2.1.2.1.1.1.2 May want to move electrode 4 mm anteriorly
2.1.2.1.1.2 Homuncular representation
2.1.2.1.1.2.1 Head
2.1.2.1.1.2.1.1 Too medial unless head symptoms specifically tar-
geted; however, if next trajectory moved anteriorly,
the homunculus will shift medially approximately
1 mm medially for every 4-mm anterior movement
2.1.2.1.1.2.2 Arm
2.1.2.1.1.2.2.1 Good medial-lateral position; however, if next tra-
jectory moved anteriorly, the homunculus will shift
medially approximately 1 mm medially for every
4-mm anterior movement
2.1.2.1.1.2.3 Leg
2.1.2.1.1.2.3.1 Too lateral unless leg symptoms specifically tar-
2.1.2.1.2 Neurons related to passive joint rotation but not tactile stimula-
tion or muscle palpation
2.1.2.1.2.1 Consistent with the anterior region of the ventral caudal thalamus
2.1.2.1.2.1.1 Suggests trajectory posterior
2.1.2.1.2.1.1.1 If intent is to detect the anterior border of the tactile
region of the ventral caudal thalamus, the degree of
posterior movement depends on what is found in the
lower part of the lower trajectory
2.1.2.1.2.2.1 Head
Appendix B / / 343

2.1.2.1.2.2.2 Arm
4-mm anterior movement
2.1.2.1.2.2.3 Leg
2.1.2.1.2.2.3.1 Too lateral unless leg symptoms specifically
targeted; however, if next trajectory moved
anteriorly, the homunculus will shift medially
approximately 1 mm medially for every 4-mm
anterior movement
2.1.2.1.3 Neurons related to passive joint rotation and muscle palpation but
not tactile stimulation
2.1.2.1.3.1 Consistent with the ventral intermediate thalamic nucleus
2.1.2.1.3.1.1 U ltimate target for the DBS lead; however, placement
in this trajectory is predicated on having identified the
anterior border of the tactile ventral caudal thalamus
toward the bottom of the trajectory at least 2 mm poste-
riorly and this trajectory being in the proper homuncular
representation
2.1.2.1.3.2.1 Head
1 mm medially for every 4-mm anterior movement or
1 mm laterally for every 4-mm posterior movement
2.1.2.1.3.2.2 Arm
4 mm anterior movement or 1 mm laterally for every
4 mm posterior movement
2.1.2.1.3.3 Leg
344 / / A ppendix B
2.1.2.1.3.3.1 Too lateral unless leg symptoms specifically targeted;

however, if next trajectory moved anteriorly, the homun-
culus will shift medially approximately 1 mm medially
for every 4-mm anterior movement or 1 mm laterally for
every 4-mm posterior movement
2.1.2.1.4 Neurons related somewhat to passive joint rotation but greater
with active rotation
2.1.2.1.4.1 Consistent with the ventral oral posterior thalamic nucleus
2.1.2.1.4.1.1 Suggests the trajectory is too anterior or that the trajectory
is very shallow in the sagittal plane; however, the decision
depends on what is found in the lower regions of the lower
trajectory and whether or not the anterior border to the
tactile region of the ventral caudal thalamus was found
2.1.2.1.4.1.1.1 If the anterior border of the tactile ventral caudal
thalamus is known to be 2–3 mm posterior at
the bottom of a previous trajectory, then this
trajectory may be appropriate for the DBS lead,
as moving the trajectory posteriorly would risk
stimulation current spread to the tactile ven-
tral caudal thalamus, producing intolerable
paresthesias
2.1.2.1.4.1.2 Homuncular representation
2.1.2.1.4.1.2.1 Head
2.1.2.1.4.1.2.1.1 Too medial unless head symptoms specifically
targeted; however, if next trajectory moved
anteriorly, the homunculus will shift medially
approximately 1 mm medially for every 4-mm
anterior movement or 1 mm laterally for every
4-mm posterior movement
2.1.2.1.4.1.2.2 Arm
2.1.2.1.4.1.2.2.1 Good medial-lateral position; however, if next
trajectory moved anteriorly, the homunculus
will shift medially approximately 1 mm medi-
ally for every 4-mm anterior movement or 1 mm
laterally for every 4-mm posterior movement
2.1.2.1.4.1.2.3 Leg
Appendix B / / 345
2.1.2.1.4.1.2.3.1
Too lateral unless leg symptoms specifically tar-
1 mm medially for every 4-mm anterior move-
ment or 1 mm laterally for every 4-mm posterior
movement
2.2 Middle lower trajectory

2.2.1 No neuronal activity
2.2.1.1 Consider electrode failure
2.2.1.1.1 Check impedance
2.2.1.1.2 Tap on wire leads from electrode
2.2.1.1.3 May be in the posterior limb of the internal capsule from a trajec-
tory that is far to lateral or anterior (the posterior limb of the inter-
nal capsule moves medially as the trajectory moves anteriorly)
2.2.1.1.3.1 M icrostimulation or macrostimulation may produce tonic
muscle contraction, speech abnormalities, or conjugate eye
deviation
2.2.1.1.4 Consider an intracerebral hematoma
2.2.1.1.4.1 Inspect electrode and guide cannulas for blood as these are
removed
2.2.1.1.4.2 The brain may wipe the cannula and electrodes of blood on
the surface
2.2.1.1.4.2.1 Place the stylet into the cannula to see if blood can be
expelled
2.2.1.1.4.2.2 Extend the tip of the microelectrode or semi-microelec-
trode to see if blood can be expelled
2.2.2 Neurons related to tactile stimulation
2.2.2.1 Consistent with the tactile region of the ventral caudal thalamus
2.2.2.1.1 If initial lower trajectory had neurons responsive to tactile
stimulation
2.2.2.1.1.1 Suggests that trajectory is very too posterior
2.2.2.1.2 If initial lower trajectory had neurons not responsive to tactile
stimulation but responsive to passive joint rotation but not muscle
palpation
2.2.2.1.2.1 Suggests a transition from the anterior ventral caudal thala-
mus to the tactile region of the ventral caudal thalamus, thus
identifying the anterior border of the ventral caudal thalamus
346 / / A ppendix B
2.2.2.1.2.1.1 The anterior border of the tactile ventral caudal thalamus

is detected but fairly high in the trajectory
2.2.2.1.2.1.1.1 Suggests that the anterior border extends more ante-
riorly deeper in the trajectory and that the trajectory
is just a bit too posterior
2.2.2.1.2.1.2 If initial lower trajectory had neurons not responsive to
tactile stimulation but is responsive to passive joint rota-
tion and muscle palpation
2.2.2.1.2.1.2.1 Suggests a transition from the ventral intermediate
thalamus to the tactile region of the ventral caudal
thalamus, thus identifying the anterior border of the
ventral caudal thalamus
2.2.2.1.2.1.2.1.1 The anterior border of the tactile ventral cau-
dal thalamus is detected but fairly high in the
trajectory
2.2.2.1.2.1.2.1.2 Suggests that the anterior border extends more
anteriorly deeper in the trajectory and that the tra-
jectory is just a bit too posterior
2.2.2.1.2.1.2.1.3 While the ventral intermediate thalamus is the
ultimate target, placing the DBS lead in this tra-
jectory would place the DBS lead too close to the
tactile ventral caudal thalamus and risk intolerable
paresthesias with DBS
2.2.2.1.2.1.2.1.4 The trajectory may be too shallow in the coronal plane,
which risks having few stimulation contacts actually in
the ventral intermediate nucleus of the thalamus
2.2.2.1.2.1.2.1.5 If possible, may want to move the point of entry
posteriorly and the next trajectory more anteriorly
2.2.2.1.2.1.3 If initial lower trajectory had neurons not responsive to
tactile stimulation and modestly responsive to passive
joint rotation but active with active joint rotation
2.2.2.1.2.1.3.1 Suggests a transition from the ventral oral posterior
thalamus to the tactile region of the ventral caudal
thalamus, thus identifying the anterior border of the
ventral caudal thalamus
2.2.2.1.2.1.3.2 The anterior border of the tactile ventral caudal thal-
amus is detected but fairly high in the trajectory
Appendix B / / 347

anteriorly deeper in the trajectory and that the
trajectory is just a bit too posterior
2.2.2.1.2.1.3.3 The trajectory may be too shallow in the sagittal
plane, which risks having few stimulation contacts
actually in the ventral intermediate nucleus of the
thalamus
2.2.2.1.2.1.3.3.1 If possible may want to move the point of
entry posteriorly and the next trajectory more
anteriorly
2.2.2.2 Neurons related to passive joint rotation but not tactile stimulation or
muscle palpation
2.2.2.2.1 Consistent with the anterior region of the ventral caudal thalamus
2.2.2.2.1.1 If initial lower trajectory had neurons responsive to tactile
stimulation
2.2.2.2.1.1.1 This scenario would be highly unlikely
2.2.2.2.1.2 If initial lower trajectory had neurons not responsive to tac-
tile stimulation but responsive to passive joint rotation but
not muscle palpation
2.2.2.2.1.2.1 Suggests a trajectory within anterior ventral caudal
thalamus
2.2.2.2.1.2.1.1 How far too anterior relative to the tactile region
of the ventral caudal thalamus depends on what is
found in the bottom of the trajectory
2.2.2.2.1.2.1.1.1 Assumes that the anterior border of the ventral
caudal thalamus has not been identified and
that this is the intent
tile stimulation but responsive to passive joint rotation and
muscle palpation
2.2.2.2.1.3.1 Suggests a trajectory within the ventral intermediate
thalamus and that the electrode has moved from the ven-
tral intermediate thalamus to the anterior ventral caudal
thalamus
2.2.2.2.1.3.1.1 Even though this location may be in the ventral inter-
mediate thalamus, which is the ultimate DBS tar-
get, this may not be a good place for the DBS lead,
348 / / A ppendix B
depending on whether the distance to the tactile

region of the ventral caudal thalamus can be detected.
2.2.2.2.1.3.1.1.1 Also suggests that the trajectory is too shal-
low in the sagittal plane and that with most of
the trajectory will be posterior to the ventral
intermediate thalamus and thus not likely to be
effective and may produce paresthesias, as the
trajectory will move closer to the tactile ventral
caudal thalamus at the bottom of the trajectory
2.2.2.2.1.3.1.1.2 If possible may want to move the point of
anteriorly
2.2.2.2.1.3.1.1.3 How far too anterior relative to the tactile region
of the ventral caudal thalamus depends on what
is found in the bottom of the trajectory
2.2.2.2.1.3.1.2 Assumes that the anterior border of the ventral cau-
dal thalamus has not been identified and that this is
the intent
tile stimulation but is responsive to passive joint rotation and
muscle palpation
2.2.2.2.1.4.1 Suggests a transition from the ventral intermediate thala-
mus to the tactile region of the ventral caudal thalamus,
thus identifying the anterior border of the ventral caudal
thalamus
2.2.2.2.1.4.1.1 While the ventral intermediate thalamus is the ulti-
mate target, placing the DBS lead in this trajectory
would place the DBS lead too close to the tactile ven-
tral caudal thalamus and risk intolerable paresthe-
sias with DBS
2.2.2.2.1.4.1.2 The anterior border of the tactile ventral caudal thal-
amus is detected but fairly high in the trajectory
anteriorly deeper in the trajectory and that the
trajectory is just a bit too posterior
2.2.2.2.1.4.1.2.2 The trajectory may be too shallow in the sagit-
tal plane, which risks having few stimulation
Appendix B / / 349
contacts actually in the ventral intermediate

nucleus of the thalamus
2.2.2.2.1.4.1.2.3 If possible, may want to move the point of
anteriorly
tile stimulation or muscle palpation and modestly responsive
to passive joint rotation but more responsive to active joint
rotation
2.2.2.2.1.5.1 Suggests a transition from the ventral oral posterior
thalamus to the anterior region of the ventral caudal
thalamus
2.2.2.2.1.5.1.1 The trajectory may be too shallow in the sagittal
plane, which risks having few stimulation contacts
actually in the ventral intermediate nucleus of the
thalamus
2.2.2.2.1.5.1.1.1 If possible, may want to move the point of
anteriorly
2.2.3 Neurons related to passive joint rotation better than active rotation and
muscle palpation
2.2.3.1 Consistent with the ventral intermediate thalamic nucleus
2.2.3.1.1 Ultimate target for the DBS lead; however, placement in this tra-
jectory is predicated on having identified the anterior border of the
tactile ventral caudal thalamus toward the bottom of the trajec-
tory and being in the proper homuncular representation
2.2.3.1.2 I f initial lower trajectory had neurons responsive to tactile
stimulation
2.2.3.1.2.1 This scenario would be highly unlikely
palpation
stimulation but responsive to passive joint rotation and muscle
palpation
350 / / A ppendix B
2.2.3.1.4.1 Suggests a trajectory within the ventral intermediate thala-

mus and that the electrode has continued the ventral inter-
mediate thalamus
2.2.3.1.4.1.1 Even though this location may be in the ventral interme-
diate thalamus, which is the ultimate DBS target, this
may not be a good place for the DBS lead, depending on
whether the distance to the tactile region of the ventral
caudal thalamus has been detected. Also, must be in
proper homuncular representation
2.2.3.1.4.1.1.1 Depends on the findings at the bottom of the
trajectory
stimulation or muscle palpation but is modestly responsive to pas-
sive joint rotation and more responsive to active joint rotation
2.2.3.1.5.1 Suggests a transition from the ventral oral posterior thalamus
to the ventral intermediate thalamus
2.2.3.1.5.1.1 While the ventral intermediate thalamus is the ultimate
target, placing the DBS lead in this trajectory would place
the DBS lead too close to the tactile ventral caudal thala-
mus and risk intolerable paresthesias with DBS
2.2.3.1.5.1.1.1 Depends on the findings at the bottom of the trajectory
2.2.3.1.5.1.2 The trajectory may be too shallow in the sagittal plane,
which risks having few stimulation contacts actually in
the ventral intermediate nucleus of the thalamus
2.2.3.1.5.1.2.1 If possible, may want to move the point of entry pos-
teriorly and the next trajectory more anteriorly
2.2.4 Neurons not responsive to tactile stimulation or muscle palpation but mod-
estly responsive to passive joint rotation and more responsive to active joint
rotation
2.2.4.1 Consistent with the ventral oral posterior thalamus
2.2.4.1.1 Too anterior and risks most DBS electrodes not being in the ven-
tral intermediate thalamus
2.2.4.1.2 If initial lower trajectory had neurons responsive to tactile stimulation
palpation
Appendix B / / 351

stimulation but responsive to passive joint rotation and muscle
palpation
2.2.4.1.5 If initial lower trajectory had neurons not responsive to tac-
tile stimulation or muscle palpation but modestly responsive
to passive joint rotation and more responsive to active joint
rotation
2.2.4.1.5.1 Suggests a the trajectory remains in the ventral oral posterior
thalamus and is too anterior
2.2.4.2
Homuncular representation
2.2.4.2.1 Head
2.2.4.2.1.1 Too medial unless head symptoms specifically targeted; how-
ever, if next trajectory moved anteriorly, the homunculus will
shift medially approximately 1 mm medially for every 4-mm
anterior movement or 1 mm laterally for every 4-mm move-
ment posteriorly
2.2.4.2.1.1.1 If homuncular representation in the initial lower trajec-
tory is the arm, the trajectory may be too shallow in the
coronal plane and the trajectory will continue medially as
the electrode is advanced
2.2.4.2.2 Arm
2.2.4.2.2.1 Good medial-lateral position; however, if next trajectory
moved anteriorly, the homunculus will shift approximately
1 mm medially for every 4-mm anterior movement or 1 mm
laterally for every 4-mm movement posteriorly
tory is the leg, the trajectory may be too shallow in the
coronal plane and the trajectory will continue medially
as the electrode is advanced
2.2.4.2.3 Leg
2.2.4.2.3.1 Too lateral unless leg symptoms specifically targeted; how-
ment posteriorly
352 / / A ppendix B
2.2.5 Bottom lower trajectory

2.2.5.1 No neuronal activity
2.2.5.1.1 Consider electrode failure
2.2.5.1.1.1 Check impedance
2.2.5.1.1.2 Tap on wire leads from electrode
2.2.5.1.2 May be in the posterior limb of the internal capsule from a trajec-
tory that is far to lateral or anterior (the posterior limb of the inter-
nal capsule moves medially as the trajectory moves anteriorly)
2.2.5.1.2.1 M icrostimulation or macrostimulation may produce tonic
muscle contraction, speech abnormalities, or conjugate eye
deviation
2.2.5.1.3 Consider an intracerebral hematoma
2.2.5.1.3.1 Inspect electrode and guide cannulas for blood as these are
removed
2.2.5.1.3.2 The brain may wipe the cannula and electrodes of blood on
the surface
2.2.5.1.3.2.1 Place the stylet into the cannula to see if blood can be
expelled
2.2.5.1.3.2.2 Extend the tip of the microelectrode or semi-microelec-
trode to see if blood can be expelled
2.2.5.2 Neurons related to tactile stimulation
2.2.5.2.1 Consistent with the tactile region of the ventral caudal thalamus
2.2.5.2.1.1 If initial and middle lower trajectory had neurons responsive
to tactile stimulation
2.2.5.2.1.1.1 Suggests that trajectory is far too posterior
2.2.5.2.1.2 If initial not responsive to tactile stimulation while the mid-
dle lower trajectory is responsive to tactile stimulation
2.2.5.2.1.2.1.1 Suggests that the anterior border of the tactile region
of the ventral caudal thalamus has been detected
fairly low in the trajectory and the trajectory is not
too posterior
2.2.5.2.1.2.1.1.1 May want another trajectory 2 mm anteriorly
and if no neurons responsive to tactile stimula-
tion are encountered then the anterior border of
the tactile ventral caudal thalamus is up to 2 mm
posterior and the DBS lead can be placed in a
trajectory 2–3 mm anterior to the last trajectory
Appendix B / / 353
that did not demonstrate neurons responsive

to tactile stimulation if the proper homuncular
representation has been found
2.2.5.2.1.3 If neither the initial or middle trajectory had neurons respon-
sive to tactile stimulation
2.2.5.2.1.3.1 Suggests the anterior border of the tactile region has
been found and is low in the trajectory
2.2.5.2.1.3.1.1 May want another trajectory 2 mm anteriorly and
if no neurons responsive to tactile stimulation are
encountered then the anterior border of the tactile
ventral caudal thalamus is up to 2 mm posterior and
the DBS lead can be placed in a trajectory 2–3 mm
anterior to the most posterior trajectory that did
not demonstrate neurons responsive to tactile
stimulation.
2.2.5.2.1.3.1.2 Alternatively, the DBS lead may be placed 3–4 mm
anteriorly, provided the proper homuncular rep-
resentation has been targeted and DBS test stimu-
lation does not produce persistent paresthesias at
reasonable voltages or currents.
2.2.5.3 Neurons related to passive joint rotation better than active joint rota-
tion but not tactile stimulation or muscle palpation
2.2.5.3.1 Consistent with the anterior region of the ventral caudal thalamus
2.2.5.3.1.1 If initial lower or middle trajectories had neurons responsive
2.2.5.3.1.2 If initial lower and middle trajectories had neurons not
responsive to tactile stimulation but responsive to passive
joint rotation better than passive rotation but not muscle
palpation
2.2.5.3.1.2.1 Suggests a trajectory within anterior ventral caudal
thalamus
2.2.5.3.1.2.1.1 How far too anterior relative to the tactile region of
the ventral caudal thalamus
2.2.5.3.1.2.1.1.1 Assumes that the anterior border of the ventral
caudal thalamus has not been identified and
that this is the intent
354 / / A ppendix B
2.2.5.3.1.2.1.1.2 May want to move posteriorly 2 mm and repeat

microelectrode recordings
tile stimulation but responsive to passive joint rotation but
not muscle palpation and the middle trajectory had neurons
responsive to muscle palpation
2.2.5.3.1.3.1 This is an unlikely scenario
2.2.5.3.1.4 If either the initial lower or middle trajectories had neurons
not responsive to tactile stimulation or muscle palpation but
more responsive to active rather than passive joint rotation
2.2.5.3.1.4.1 Suggests a trajectory moving from the ventral oral poste-
rior thalamus to the anterior region of the ventral caudal
thalamus
2.2.5.3.1.4.1.1 Suggests that the trajectory is too shallow in the sag-
ittal plane
2.2.5.3.1.4.1.1.1 May want to move the entry point posteriorly if
possible
2.2.5.4 Neurons related to muscle palpation
2.2.5.4.1 Consistent with the ventral intermediate thalamus
2.2.5.4.1.1 If initial lower or middle trajectories had neurons responsive
2.2.5.4.1. If initial lower or middle trajectories had neurons were not respon-
sive to tactile stimulation or muscle palpation but were responsive
to passive joint rotation better than active joint rotation
2.2.5.4.1.3 If initial lower and middle trajectories had neurons respon-
sive to muscle palpation
2.2.5.4.1.3.1 Suggests a trajectory within ventral intermediate thalamus
2.2.5.4.1.3.1.1 Th is would be a good trajectory for the DBS lead,
assuming that the anterior border of the tactile
region of the ventral caudal thalamus was demon-
strated to be at least 2–3 mm posterior and assum-
ing the correct medial-lateral position relative to the
homunculus
2.2.5.4.1.4 If initial lower trajectory had neurons responsive to muscle
palpation and middle trajectory did not
Appendix B / / 355

2.2.5.4.1.5 If initial lower trajectory did not have neurons responsive
to muscle palpation and middle trajectory did have neurons
2.2.5.4.1.5.1 Suggests the trajectory passes from the ventral oral pos-
terior thalamus into the ventral intermediate thalamus
and that the trajectory is a bit shallow in the sagittal plane
2.2.5.4.1.5.1.1 One could move the trajectory posteriorly, however
that might move the deepest portion closer to the
tactile region of the ventral caudal thalamus
2.2.5.4.1.5.1.1.1 This may be detectable by macrostimulation
through the DBS lead
2.2.5.4.1.6 If initial and middle lower trajectories did not have neurons
2.2.5.4.1.6.1 Suggests the trajectory passes from the ventral oral poste-
rior thalamus into the ventral intermediate thalamus low
in the trajectory and that the trajectory is a bit shallow in
the sagittal plane and there may not be sufficient stimu-
lation contacts within the ventral intermediate thalamus
for effective DBS
2.2.5.4.1.6.1.2 One could move the trajectory posteriorly, however
that might move the deepest portion closer to the
tactile region of the ventral caudal thalamus
2.2.5.5 Neurons not responsive to tactile stimulation or muscle palpation but
more responsive to active rather than passive joint rotation
2.2.5.5.1 Consistent with ventral oral posterior thalamus
2.2.5.5.1.1 If initial or middle lower trajectory had neurons responsive to
tactile stimulation, muscle palpation, and passive more than
active joint rotation
2.2.5.5.1.1.1 Th is would be a highly unlikely scenario
2.2.5.5.1.2 These findings at the bottom of the trajectory would suggest
that the DBS lead is far too anterior
2.2.5.6 Homuncular representation
356 / / A ppendix B
2.2.5.6.1
Head
2.2.5.6.1.1 Too medial unless head symptoms specifically targeted; how-
ment posteriorly
2.2.5.6.1.1.1 I f homuncular representation in the initial lower trajec-
tory is the arm, the trajectory may be too shallow in the
coronal plane and the trajectory will continue medially as
the electrode is advanced
2.2.5.6.2
Arm
2.2.5.6.2.1 Good medial-lateral position; however, if next trajectory
moved anteriorly, the homunculus will shift medially approx-
imately 1 mm medially for every 4-mm anterior movement or
1 mm laterally for every 4-mm movement posteriorly
tory is the leg, the trajectory may be too shallow in the
coronal plane and the trajectory will continue medially
as the electrode is advanced
2.2.5.6.3
Leg
2.2.5.6.3.1 Too lateral unless leg symptoms specifically targeted; how-
ment posteriorly
3.0 DBS macrostimulation at the target

3.1 Configuration and stimulation parameters
3.1.1 Assuming wide bipolar configuration where ventral-most (deepest) contact
is the cathode (negative) contact and the dorsal-most (highest) contact is
the anode as initial configuration
3.1.1.1 Most intense but localized electrical field
3.1.2 Pulse width of 120 microseconds
3.1.2.1 Sufficient to activate smaller axons
3.1.3 Frequency of 160 pulses per second (pps)
3.2 No response up to 10 voltages (constant-voltage stimulation) or 5 mA (constant-current

stimulation)
Appendix B / / 357
3.2.1 Test electrode impedances and current flow

3.2.1.1 Refer to manufacturer’s guidelines for excessive impedance or possible
open circuit
3.2.2 Check stimulator battery levels
3.2.3 Check connections
3.2.4 Change cathodes
3.2.5 Replace DBS lead
3.3 No microthalamotomy effect
3.3.1 Therapeutic effect without adverse effects with stimulation voltage/current
of 4 v/3 mA
3.3.1.1 May be optimal location for DBS lead
3.3.2 No therapeutic effect without adverse effects with stimulation voltage/cur-
rent of 4 v/3 mA
3.3.2.1 Check system, see 3.2
3.3.2.2 May be too anterior or medial
3.3.3 Persistent paresthesias at voltages/current less than 2 v/2 mA more than the
threshold for benefit with the ventral-most (deepest) contact the cathode and
the dorsal-most (highest) contact the anode (positive contact) and persistent
paresthesias at voltages/current less than 2 v/2 mA more than the threshold
for benefit with the ventral-most (deepest) contact the anode (positive con-
tact) and the dorsal-most (highest) contact cathode (negative contact)
3.3.3.1 Suggests that the DBS contact is running parallel to the anterior border
of the ventral caudal thalamus and is too posterior
3.3.3.1.1 May want to move the DBS lead anteriorly depending on the
threshold to persistent paresthesias
3.3.4 Persistent paresthesias at voltages/current less than 2 v/2 mA more than the
threshold for benefit with the ventral-most (deepest) contact the cathode and
the dorsal-most (highest) contact the anode (positive contact) and persistent
paresthesias at voltages/current more than 2 v/2 mA more than the threshold
for benefit with the ventral-most (deepest) contact the anode (positive con-
tact) and the dorsal-most (highest) contact cathode (negative contact)
3.3.4.1 Suggests that the dorsal-most contact is further from the anterior
border of the tactile ventral caudal thalamus than is the ventral-most
contact
3.3.4.1.1 May consider moving the DBS lead out (dorsal) to move the
ventral-most contact anteriorly as the DBS lead moves dorsally
358 / / A ppendix B
3.3.5 Tonic muscle contraction at voltages/current less than 2 v/2 mA more than

the threshold for benefit with the ventral-most (deepest) contact the cath-
ode and the dorsal-most (highest) contact the anode (positive contact) and
tonic muscle contraction at voltages/current more than 2 v/2 mA more
than the threshold for benefit with the ventral-most (deepest) contact the
anode (positive contact) and the dorsal-most (highest) contact cathode
(negative contact)
3.3.5.1 Suggests that the DBS lead is too ventral (deep) and is affecting the
corticospinal and corticobulbar pathways in the posterior limb of the
internal capsule
3.3.5.2 Alternatively, the DBS lead may be too lateral
3.3.6 Tonic muscle contraction at voltages/current less than 2 v/2 mA more than
the threshold for benefit with the ventral-most (deepest) contact the cath-
ode and the dorsal-most (highest) contact the anode (positive contact) and
tonic muscle contraction at voltages/current less than 2 v/2 mA more than
the threshold for benefit with the ventral-most (deepest) contact the anode
(positive contact) and the dorsal-most (highest) contact cathode (negative
contact)
3.3.6.1 Suggests that the DBS lead is running parallel to the posterior limb of
the internal capsule, which would be the case if the trajectory were too
shallow in the coronal plane
3.3.6.1.1 May want move the entry point medially or, alternatively, move
the DBS lead medially
3.3.7 Dysarthria or aphasia at voltages/current less than 2 v/2 mA more than the
threshold for benefit with the ventral-most (deepest) contact as the cathode
(negative contact)
3.3.7.1 The DBS lead may be too medial affecting the head representation,
spread to the corticobulbar fibers, or a nonspecific effect of thalamic
stimulation
3.3.7.1.1 Increase the voltage/current and if tonic contraction appears, the
DBS lead may be too ventral (deep) or lateral (see above)
APPENDIX C
GLOBUS PALLIDUS INTERNA DEEP

BRAIN STIMULATION ALGORITHM
DESCRIPTION AND CAVEAT
The algorithm below is to aid in the optimal placement of DBS leads in the sensorim-
otor region of the posterior and lateral globus pallidus interna using microelectrode
recordings. The algorithm can be adapted for semi-microelectrode recordings. Also,
there is a guide to aid in the interpretation of DBS macrostimulation. The algorithm is
a distillation of the author’s experience and has not been subjected to rigorous prospec-
tive validation because of the complexity and expense such validation would entail.
Note, this algorithm is educational in nature and is not intended to direct the care of
any individual patient. The physician, surgeon, and healthcare professional must exer-
cise their judgment in the utilization of the algorithm and assume full and exclusive
responsibility for the care of the patient.
The approach used in the algorithm is to divide the microelectrode trajectory into
an upper and lower section. The upper section represents the putamen. The lower tra-
jectory is divided into three levels, the initial level, the middle level, and the bottom
level (Figure C.1). Components of the lower trajectories include the putamen, globus
pallidus externa, globus pallidus interna, border cells that surround the globus pallidus
interna, and the posterior limb of the internal capsule. The algorithm described below
incorporates information from the four levels to help establish where the trajectory is
relative to the desired trajectory and, hence, where to move the electrode should the
present trajectory not meet criteria for DBS lead implantation.
The approach first intends to find the anterior border of the posterior limb of the
internal capsule, as the sensorimotor region of the globus pallidus abuts the posterior
limb. Thus, just targeting the sensorimotor region of the globus pallidus interna with-
out identifying the posterior limb of the internal capsule risks placing the DBS contacts
359
360 / / A ppendix C
Dorsal
Medial Lateral
Vental
Internal
capsule
Upper trajectory Low frequency, transient, irregular, low density
Pt High frequency, sustained, irregular, high density

High frequency-pause or low frequency burst
Initial lower trajectory In Parkinson’s disease in upper region, moderate
GPe frequency, transient, regular, moderate density in
GPi lower region
High frequency, sustained, irregular, high density

Bottom lower trajectory Border cell area
High frequency, sustained, regular, low density in

Optic tract upper region, related to light flashes or phosphenes
with stimulation in the lower region
FIGURE C.1 Schematic of a coronal section through the putamen, globus palldius externa, bor-
der cells and globus palldius interna showing a microelectrode trajectory. As can be seen, the
trajectory is organized into an upper and lower half and the lower half is divided into the initial,
middle and bottom one-third. The trajectories show the repsentative structures often associ-
ated with each segment. Modified from Schaltenbrand G and Wahren W: Atlas for Stereotaxy
of the Human Brain. Stuttgart, Thieme, 1977.
too close to the corticobulbar and corticospinal tracts contained within the posterior
limb of the internal capsule. The result may be a low stimulation threshold for produc-
ing tonic muscle contractions, which may limit the effectiveness of DBS. Further, in
order to place as many of the DBS lead contacts as possible in the sensorimotor region
it is important that the trajectory pass through at an angle where the span of contacts is
in the sensorimotor region of the globus pallidus interna. It is at this ventral limit that
the DBS contacts will be closest to the posterior limb of the internal capsule.
Most often, the most anterior border of the posterior limb of the internal capsule
can only be detected by bracketing (Figure C.2). In this case, the posterior trajectory
does not detect neuronal activities as it traverses the bottom of the lower trajectory,
while another trajectory just anterior does detect neuronal activities throughout the
course of the trajectory. Thus, one can conclude that the anterior border of the posterior
limb of the internal capsule lies between the two trajectories. Occasionally, a trajectory
will demonstrate only very small extent of the posterior limb of the internal capsule
(Figure C.3). In that case, the intraoperative neurophysiologist may infer that the tra-
jectory is very near the anterior border of the posterior limb of the internal capsule and
may infer that the optimal trajectory for the DBS lead is 2–3 mm, in the case of patients
with Parkinson’s disease or 4–5 mm for patients with dystonia, anteriorly, provided it is
in the sagittal plane appropriate to the homuncular representation.
Appendix C / / 361
Dorsal
Posterior Anterior
Vental
Internal
capsule
Upper trajectory
Initial lower trajectory Pt

GPe
GPi
Border cell area

Optic tract
FIGURE C.2 Schematic of a saggital section through the putamen, globus palldius externa, bor-
der cells and globus palldius interna showing a microelectrode trajectory. As can be seen, the
trajectory is organized into an upper and lower half and the lower half is divided into the initial,
middle and bottom one-third. The trajectories show the repsentative structures often associ-
ated with each segment. This figure shows how the anterior border of the posterior limb of the
internal capsule can be bracketed between the trajectories. Modified from Schaltenbrand G
Also intended is to find at least 5 mm or more of sensorimotor globus pallidus

interna in the trajectory, as less than that suggests that the trajectory is skimming over
the top or through the bottom of the globus pallidus interna. This typically occurs
when the trajectory is too shallow in the sagittal plane. The potential consequences are
that the ventral-most contacts are too close to the posterior limb of the internal capsule
to be useful and the dorsal-most contacts are anterior to the sensorimotor region and
likewise less effective.
An algorithm to guide the interpretation of test DBS macrostimulation also is pro-
vided. Again, intraoperative neurophysiologists must use their own judgment when
interpreting stimulation results for their individual patient. There are two aspects to the
test DBS macrostimulation. First, the responses to stimulation can help understand the
regional anatomy around the DBS lead electrodes. The nature of the responses helps
identify the structure in the regional anatomy, and the threshold to the response helps
interpret relative distances between the DBS lead cathode (negative) contact and the
structure. The second purpose is to establish the therapeutic window, which is the differ-
ence between the threshold that produces benefit and the threshold that produces adverse
effects (see chapter 10). In cases of a marked micro-pallidotomy effect, the threshold for
benefit cannot be determined. Consequently, it is reasonable to assume that the thresh-
old is somewhat higher than the voltages associated with benefit that are reported in the
literature. In these cases, this author assumes a threshold to benefit of 4 volts, if using a
constant-voltage stimulator, or 3 mA if using a constant-current stimulator.
Dorsal
Posterior Anterior
Vental
Internal
capsule
Upper trajectory
Initial lower trajectory Pt

GPe
GPi
Border cell area

Optic tract
FIGURE C.3 Schematic of a saggital section through the putamen, globus palldius externa,
border cells and globus palldius interna showing a microelectrode trajectory. As can be seen,
the trajectory is organized into an upper and lower half and the lower half is divided into the
initial, middle and bottom one-third. The trajectories show the repsentative structures often
associated with each segment. This figure shows how the anterior border of the posterior
limb of the internal capsule can found low in the trajectory suggesting that the anterior bor-
der is relatively close and bracketing by the use of another trajectory may not be necessary.
Modified from Schaltenbrand G and Wahren W: Atlas for Stereotaxy of the Human Brain.
Stuttgart, Thieme, 1977.
Some items in the algorithm have “This scenario is highly unlikely” based on the
anatomy. This suggests that there may have been an error in interpreting the character-
istics of the neuronal activities and the recordings or the stimulation results should be
reconsidered.
Globus Pallidus Interna DBS Algorithm
1. Upper trajectory

1.1.1.2.1 If silent, inputs may be in electrical contact with ground, there
may be a broken electrical connection, or a failure in the amplifier
system
1.1.1.3 Replace microelectrode
far too medial or posterior (the posterior limb of the internal capsule moves
anteriorly as the trajectory moves medially)
1.1.2.1 Microstimulation or macrostimulation may produce tonic muscle con-
traction, speech abnormalities, or conjugate eye deviation
Appendix C / / 363

1.1.3.2 The brain may wipe the cannula and electrodes of blood on the
surface
1.1.3.2.2 E xtend the tip of the microelectrode or semi-microelectrode to
see if blood can be expelled
1.2 Irregular, low to moderate frequency, transient, low density within recording site,
low density in trajectory
1.2.1 Consistent with putamen
1.2.1.1 Expected at this height in the trajectory
1.3 Irregular, high frequency, sustained, high density within recording site, may have
high-frequency-pause neurons and low frequency bursting neurons, high density
in trajectory and no sensorimotor driving
1.3.1 Consistent with globus pallidus externa
1.3.1.1 If found in the upper trajectory suggests trajectory may be medial
2. Lower trajectory
2.1. Initial lower trajectory
2.1.1. No neuronal activity
2.1.1.1.2. Replace electrode
2.1.1.3. In the absence of neuronal activities in the upper trajectory
2.1.1.3.1. May be too medial and/or posterior in the posterior limb of the
internal capsule
lateral ventricle
2.1.1.3.1.2. Stimulation may, but not necessarily, produce tonic contrac-
tion, speech abnormalities, or conjugate eye deviation
364 / / A ppendix C
2.1.2. Irregular, low frequency, transient, possibly bursting, low density within

recording site, low density in trajectory
2.1.2.1. Consistent with putamen
2.1.2.1.1.1. At this height in the trajectory, may suggest trajectory is
medial
2.1.3. Irregular, high frequency, sustained, high density within recording site,
may have high frequency pause neurons and low frequency bursting neu-
rons, high density in trajectory, and no sensorimotor driving
2.1.3.1. Consistent with globus pallidus externa
2.1.3.1.1. Expected at this depth of the trajectory
2.1.4. Regular, low to moderate frequency, persistent, low density within record-
ing site, low density in trajectory, action potential of longer duration pro-
ducing a lower pitch in the audio signal
2.1.4.1. Consistent with border cell area
2.1.4.1.1. At this height may suggest a medial trajectory
2.1.5. Irregular, high frequency, sustained, high density within recording site,
high density in trajectory, and sensorimotor driving
2.1.5.1. Consistent with globus pallidus interna
2.1.5.1.1. At this height suggests a medial trajectory, but appropriate sagit-
tal plane is determined by the homuncular representation if found
2.1.5.1.2. Homuncular representation, important for determining the cor-
rect sagittal plane
2.1.5.1.2.1. Head
2.1.5.1.2.1.1. Consistent with a lateral trajectory, may be optimal in
sagittal plane for treatment of cervical dystonia or hyper-
kinesia provided that the trajectory is 2–3 mm posterior
in patients with hyperkinesia and 4–5 mm in patients
with dystonia anterior to the anterior border of the pos-
terior limb of the internal capsule. Would not be appro-
priate for patients with Parkinson’s disease
2.1.5.1.2.2. Arm
2.1.5.1.2.2.1. Consistent with middle portion of the sensorimotor
region, may be optimal in sagittal plane for treatment
of upper extremity focal dystonia, hyperkinesia or
Parkinson’s provided that the trajectory is 2–3 mm pos-
terior in patients with Parkinson’s disease and 4–5 mm
in patients with dystonia
Appendix C / / 365
2.1.5.1.2.3.
Leg
region, may be optimal in sagittal plane for treatment of
lower extremity dystonia or hyperkinesia provided tra-
jectory is 2-3 mm and 4–5 mm in patients with dysto-
nia anterior to the anterior border of the posterior limb
of the internal capsule. Would not be appropriate for
patients with Parkinson’s disease
2.2. Middle lower trajectory

2.2.1.1. No or rare neuronal activity
2.2.1.1.1. May be too medial and/or posterior and in the posterior limb
of the internal capsule when encountered at this height in the
trajectory
lateral ventricle
2.2.1.1.1.2. Stimulation may produce tonic contraction
2.2.1.1.2. Check for electrode failure
2.2.1.1.2.1. Check impedances
2.2.1.1.3. Consider intracerebral hematoma
2.2.1.1.3.1. Examine cannula and electrode for evidence of blood
2.2.1.1.3.2. Withdrawing cannula through brain may wipe blood off
2.2.1.1.3.5. Check top of cannula to see if blood is exiting
2.2.2. I rregular, high frequency, sustained, high density within recording site,
2.2.2.1.1. At this height may be an appropriate trajectory in the coronal
plane
rect sagittal plane
2.2.2.1.2.1. Head
366 / / A ppendix C

2.2.2.1.2.2. Arm
terior in patients with hyperkinesia or Parkinson’s dis-
ease and 4–5 mm in patients with dystonia
2.2.2.1.2.3. Leg
lower extremity dystonia or hyperkinesia provided and
4–5 mm in patients with dystonia anterior to the anterior
border of the posterior limb of the internal capsule. Would
not be appropriate for patients with Parkinson’s disease
rons, high density in trajectory
2.2.3.1.1. At this depth of the trajectory suggests a lateral trajectory
2.2.4.1.1. At this height may be expected, if the segment just above (middle
lower trajectory) was typical of globus pallidus externa
2.2.4.1.2. If segment just above (middle lower trajectory) typical of globus
pallidus interna, suggests that trajectory is too posterior
2.3. Bottom lower trajectory

2.3.1. No or rare neuronal activity
2.3.1.1. If more than a couple of mm then may be too medial and/or posterior
and in the posterior limb of the internal capsule
2.3.1.1.1. Stimulation may produce tonic contraction
Appendix C / / 367
2.3.1.1.2. If more than less than couple of mm then may be in an optimal
trajectory in sagittal plane, as the posterior limb of the internal
capsule may be just posterior
2.3.1.1.2.1. If stimulation produces tonic contraction, trajectory is too
posterior
2.3.1.3.3. Insert stylet to see if it pushes blood out
2.3.1.3.4. Extend electrode tip to see if it pushes out blood
2.3.2.1.1. At this height may be an appropriate trajectory in the coronal
plane
rect sagittal plane
2.3.2.1.2.1. Head
2.3.2.1.2.2. Arm
2.3.2.1.2.2.1 Consistent with middle portion of the sensorimotor
terior in patients with hyperkinesia or Parkinson’s dis-
ease and 4–5 mm in patients with dystonia
2.3.2.1.2.3. Leg
368 / / A ppendix C

lower extremity dystonia or hyperkinesia provided and
4–5 mm in patients with dystonia anterior to the ante-
rior border of the posterior limb of the internal capsule.
Would not be appropriate for patients with Parkinson’s
disease
rons, high density in trajectory, and no sensorimotor driving
2.3.3.1.1. At this depth of the trajectory suggests a lateral trajectory
2.3.4.1.1. At this height may be expected, if the segment just above (middle
lower trajectory) was typical of globus pallidus externa
2.3.4.1.2. If segment just above (middle lower trajectory) typical of globus
pallidus interna, suggests that trajectory is too posterior
3. DBS macrostimulation at ventral (deepest) aspect of the neuronal activities identi-

fied as the sensorimotor subthalamic nucleus
3.1. Configuration and stimulation parameters
3.1.1. A ssuming wide bipolar configuration where ventral-most (deepest) con-
tact is the cathode (negative) contact and the dorsal-most (highest) contact
is the anode as initial configuration
3.1.1.1. Most intense but localized electrical field
3.1.2. Pulse width of 120 microseconds
3.1.2.1. Sufficient to activate smaller axons
3.1.3. Frequency of 160 pulses per second (pps)
3.2. No response up to 10 voltages (constant-voltage stimulation) or 5 mA (constant-current

stimulation)
3.2.1. Test electrode impedances and current flow
3.2.1.1. Refer to manufacturer’s guidelines for excessive impedance or pos-
sible open circuit
3.2.2. Check stimulator battery levels
Appendix C / / 369
3.2.3. Check connections

3.2.4. Change cathodes
3.2.5. Replace DBS lead
3.2.6. Contacts may be too anterior to the sensorimotor region
3.2.6.1. Note, symptoms of certain disorders, such as dystonia, may not respond
initially to DBS
3.3. Tonic contraction
3.3.1. Typically suggests trajectory is too posterior
3.4. Produces phosphenes or other visual distortions

3.4.1. Suggests trajectory too deep
3.5. Unusual or strange feeling

3.5.1. Suggests trajectory too medial
APPENDIX D
MICROELECTRODE RECORDING FORM

FOR SUBTHALAMIC NUCLEUS DEEP
BRAIN STIMULATION
370
Patient Name Date:
MRN: Target: Subthalamic nucleus (STN) Page __ of ___
Start time: Stop time: ◻ Left ◻ Right
Location* Electrode Activity description Prob Target S-m driving Stim response Grade Notes Stimulation
current
AP _____ ◻ Low frequency ◻ Regular ◻ Thalamus ◻ Yes ◻ No ◻ not tested impedance _______ go
Lat ______ ◻ Not tested down quickly to _____
Vert _____ Ring ◻ Moderate frequency ◻ Irregular ◻ Zona Incerta ◻ head ◻ move ____ant post ____ med
_____ ◻ prox UE lat ◻ XY ◻ frame
Anterior ↑ ◻ High frequency ◻ Bursting ◻ lat STN Bradykinesia ◻+◻-◻0
Arc ______ ◻ distal UE
Lateral → ◻ Low density ◻ transient ◻ med STN Rigidity ◻ + ◻ - ◻ 0
Depth ◻ LE
◻ Moderate density ◻ lost ◻ SNr Tremor ◻ + ◻ - ◻ 0
◻ High density ◻ injury Tonic contract ◻ + ◻ 0
Backgnd ◻ ↑ ◻ ↓ Eye deviation ◻+◻0
Vert _____ ◻ Moderate frequency ◻ Irregular ◻ Zona Incerta ◻ head ◻ move ____ant post ____ med
Ring _____ ◻ prox UE lat ◻ XY ◻ frame
Depth ◻ LE
Vert _____ ◻ Moderate frequency ◻ Irregular ◻ Zona Incerta ◻ head ◻ move ____ ant post ____ med
Depth ◻ LE
AP _____ ◻ Low frequency ◻ Regular ◻ Thalamus ◻ Yes ◻ No ◻ not tested impedance_______ go
Vert _____ ◻ Moderate frequency ◻ Irregular ◻ Zona Incerta ◻ head ◻ move ____ ant post ____ med
Depth ◻ LE
Vert _____ ◻ head
Ring _____ ◻ Moderate frequency ◻ Irregular ◻ Zona Incerta ◻ prox UE ◻ move ____ ant post ____ med
Arc ______ Anterior ↑ ◻ distal UE lat ◻ XY ◻ frame
Depth Lateral → ◻ High frequency ◻ Bursting ◻ lat STN ◻ LE Bradykinesia ◻+◻-◻0
◻ Low density ◻ transient ◻ med STN Rigidity ◻ + ◻ - ◻ 0
* frame coordinates relative to center hole
APPENDIX E
MICROELECTRODE RECORDING FORM

FOR GLOBUS PALLIDUS INTERNA
373
Patient Name Date:
MRN: Target: Globus pallidus interna (GPi) Page ____ of ____
current
AP ________ ◻ Low frequency ◻ Regular ◻ CD/PT ◻ Yes ◻ No ◻ not tested impedance ________ go
Lat ________ ◻ Not tested down quickly to ________
Vert ________ ◻ Moderate frequency ◻ Irregular ◻ GPe ◻ head ◻ move _______ ant post _______
Ring ________ ◻ prox UE med lat ◻ XY ◻ frame
Arc ________ Anterior ↑ ◻ High frequency ◻ Bursting ◻ GPi ◻ distal UE Bradykinesia ◻+◻-◻0
Depth Lateral → ◻ Low density ◻ transient ◻ Int cap ◻ LE Rigidity ◻ + ◻ 0
◻ Moderate density ◻ lost ◻ Optic tract Tremor ◻ + ◻ 0
◻ High density ◻ injury ◻ Border Tonic contract ◻ + ◻ 0
Backgnd ◻ ↑ ◻ ↓ ◻ HFP ◻ LFP Phosphenes ◻+◻0
Vert ________ ◻ Moderate frequency ◻ Irregular ◻ GPe ◻ head ◻ move ________ ant post
Ring ________ ◻ prox UE ________ med lat ◻ XY
Arc ________ ◻ distal UE ◻ frame
Depth Anterior ↑ ◻ High frequency ◻ Bursting ◻ GPi ◻ LE Bradykinesia ◻+◻-◻0
Lateral → ◻ Low density ◻ transient ◻ Int cap Rigidity ◻ + ◻ 0
Arc ________ Anterior ↑ ◻ distal UE ◻ frame
Depth Lateral → ◻ High frequency ◻ Bursting ◻ GPi ◻ LE Bradykinesia ◻+◻-◻0
◻ Low density ◻ transient ◻ Int cap Rigidity ◻ + ◻ 0
Arc ________ Anterior ↑ ◻ distal UE ◻ frame
AP _____ ◻ Low frequency ◻ Regular ◻ CD/PT ◻ Yes ◻ No ◻ not tested impedance ________ go
Lat ______ ◻ Not tested down quickly to ________
Vert _____ ◻ Moderate frequency ◻ Irregular ◻ GPe ◻ head ◻ move ________ ant post
Ring _____ ◻ prox UE ________ med lat ◻ XY
Arc ______ Anterior ↑ ◻ distal UE ◻ frame
APPENDIX F
MICROELECTRODE RECORDING
FORM FOR VENTRAL INTERMEDIATE
THALAMUS
376
Patient Name Date:
MRN: Target: Thalamic nucleus (Vim) Page __ of __
current
AP _____ ◻ Low frequency ◻ Regular ◻ dorsal thal ◻ Yes ◻ No ◻ not tested impedance _______ go down
Lat _____ ◻ Not tested quickly to _______
Vert _____ ◻ Moderate frequency ◻ Irregular ◻ Vc tactile ◻ head ◻ move _______ ant post _______
Ring _____ ◻ prox UE med lat ◻ XY ◻ frame
Arc ______ Anterior ↑ ◻ High frequency ◻ Bursting ◻ Vc deep ◻ distal UE Rest tremor ◻+◻-◻0
Depth Lateral → ◻ Low density ◻ transient ◻ Vim ◻ prox LE Postural tremor ◻ + ◻ - ◻ 0
◻ Moderate density ◻ lost ◻ Vop ◻ distal LE Action tremor ◻ + ◻ - ◻ 0
◻ High density ◻ injury ◻ joint Tonic contract ◻ + ◻ 0
Backgnd ◻ ↑ ◻ ↓ ◻ spindle Paresthesias ◻ + ◻ 0
◻ light touch Speech ◻ -- ◻ 0
Arc _____ Anterior ↑ ◻ High frequency ◻ Bursting ◻ Vc deep ◻ distal UE Rest tremor ◻+◻–◻0
Depth Lateral → ◻ Low density ◻ transient ◻ Vim ◻ prox LE Postural tremor ◻ + ◻ – ◻ 0
◻ Moderate density ◻ lost ◻ Vop ◻ distal LE Action tremor ◻ + ◻ – ◻ 0
◻ light touch Speech ◻–◻0
Arc _____ Anterior ↑ ◻ High frequency ◻ Bursting ◻ Vc deep ◻ distal UE Rest tremor ◻+◻-◻0
Depth Lateral → ◻ Low density ◻ transient ◻ Vim ◻ prox LE Postural tremor ◻ + ◻ - ◻ 0
◻ Moderate density ◻ lost ◻ Vop ◻ distal LE Action tremor ◻ + ◻ - ◻ 0
Arc _____ Anterior ↑ ◻ High frequency ◻ Bursting ◻ Vc deep ◻ distal UE Rest tremor ◻+◻–◻0
Depth Lateral → ◻ transient ◻ prox LE
◻ Low density ◻ Vim Postural tremor ◻ + ◻ – ◻ 0
◻ lost ◻ distal LE
◻ Moderate density ◻ Vop Action tremor ◻ + ◻ – ◻ 0
◻ injury ◻ joint Tonic contract ◻ + ◻ 0
◻ High density
* frame coordinates relative to center hole
APPENDIX G
INTRAOPERATIVE MACROSTIMULATION
FOR CLINICAL EFFECT IN
PARKINSON’S DISEASE
This form may be used to document the clinical and adverse effects of macrostimulation,
typically through the DBS lead. The reader is encouraged to photocopy this form for use
in the operating room. Electronic versions that the reader may modify are available for
downloading from www.grnneuromod.com. This form is not intended to direct the treat-
ment of any individual patient, and the user assumes full and sole responsibility for its use.
The nomenclature of the contact names in the DBS lead varies greatly. Typically, the
DBS lead contains four contacts in a row along the long axis if the DBS lead. This form
names the contacts ventral-most for the deepest contact, ventral for the next contact, dorsal
for the contact above the ventral contact, and dorsal-most for the highest contact. The user
may wish to enter the specific numbering of the contacts appropriate to the DBS lead used.
For the stimulation intensity, provision is made for both constant-voltage and
constant-current stimulation. Values are in volts for the former and in milliamps (mA)
for the latter. To avoid confusion, the reader may wish to strike through the parameter
that is not appropriate.
Visual analog rating scales are provided for the major motoric symptoms assessable in
the operating room. The scales are derived from the motor part of the Unified Parkinson
Disease Rating Scales (UPDRS) (Goetz CG, Tilley BC, Shaftman SR, et al.: Movement
Disorder Society-sponsored revision of the Unified Parkinson’s Disease Rating Scale
(MDS-UPDRS): scale presentation and clinimetric testing results. Movement Disorders
23(15): 2129–2170, 2008. In these scales, 0 is normal and 4 is the most severe (see
chapter 12). The exception is rating of muscle tone. Typically 0 would be normal, which
indicates the normal resistance to passive joint rotations. However, with DBS the muscle
tone actually may be less than normal. For this reason, normal muscle tone is assigned a
value of 1 and the reduction of muscle tone below normal is assigned a value of 0.
379
Intraoperative Macrostimulation for Parkinson’s disease
Electrode
selection
C Pulse Rate Volts Adverse effects, Effect on tone Effect on finger Effect on rapid Effect on tremor
a width or mA where and transient vs tapping hand opening and
Ventral-most
Dorsal-Most
s sustained closing
Ventral
e
Dorsal
mosst
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Notes:
APPENDIX H
INTRAOPERATIVE
MACROSTIMULATION FOR CLINICAL
EFFECT IN TREMOR DISORDERS
This form may be used to document the clinical and adverse effects of macrostimu-
lation, typically through the DBS lead. The reader is encouraged to photocopy this
form for use in the operating room. Electronic versions that the reader may modify
are available for downloading from www.grnneuromod.com. This form is not intended
to direct the treatment of any individual patient, and the user assumes full and sole
responsibility for its use.
names the contacts ventral-most for the deepest contact, ventral for the next contact,
dorsal for the contact above the ventral contact, and dorsal-most for the highest con-
tact. The user may wish to enter the specific numbering of the contacts appropriate to
the DBS lead used.
Visual analog rating scales are provided for the various forms of tremor assess-
able in the operating room. In these scales, 0 is normal and 4 is the most severe (see
chapter 12).
381
Intraoperative Macrostimulation for Tremor Disorders
Electrode selection
Case Pulse Rate Volts Adverse effects, Effect on resting Effect on postural Effect on tremor Effect on tremor
Dorsal-Most mosst
width or mA where and transient vs tremor tremor during the finger-to- during bringing a cup
Ventral-most
sustained nose task to the lips
Ventral
Dorsal
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Notes:
APPENDIX I
INTRAOPERATIVE
MACROSTIMULATION FOR
CLINICAL EFFECT ON DYSTONIA
the DBS lead used.
Visual analog rating scales are provided for four forms of dystonia assessable in the
operating room. These forms may differ in the region of the body affected or by the joints
rotated to produce the dystonia. In these scales, 0 is normal and 4 is the most severe (see
chapter 12). As the patient may have multiple and varied forms of dystonia simultane-
ously, the scales are assigned to four forms of dystonia presented by the patient. It is
important to note that many forms of dystonia will not be affected by macrostimulation
383
384 / / A ppendix I
in the operating room. Some forms of dystonia may take weeks to months to respond.
In this author’s experience, “phasic” dystonic symptoms, that is, those that are rapidly
fluctuating, such as dystonic tremor, respond most rapidly and therefore may be more
feasible to measure during intraoperative macrostimulation. Also, cervical dystonia
may be difficult to assess when rigid frame systems are used, as they do not allow move-
ment of the head. Many frameless systems allow some movement of the head, which
may allow assessment for response to intraoperative stimulation.
Intraoperative Macrostimulation for Dystonia
Electrode
selection
Case Pulse Rate Volts Adverse effects, Effect on dystonia 1 Effect on dystonia 2 Effect on dystonia 3 Effect on dystonia 4
Dorsal-Most mosst
width or mA where and transient vs
Ventral-most
sustained
Ventral
Dorsal
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Notes:
APPENDIX J
INTRAOPERATIVE
CLINICAL EFFECT ON TICS
the DBS lead used.
Visual analog rating scales are provided for four forms of tics assessable in the oper-
ating room. These forms may differ in the region of the body affected or by the joints
rotated to produce the tic. In these scales, 0 is normal and 4 is the most severe (see
chapter 12). As the patient may have multiple and varied forms of tics simultaneously,
the scales are assigned to four forms of tics presented by the patient.
386
Intraoperative Macrostimulation for Tics
Electrode
selection
Case Pulse Rate Volts Adverse effects, Effect on tic 1 Effect on tic 2 Effect on tic 3 Effect on tic 4
Dorsal-Most mosst
width or where and transient vs
Ventral-most
mA sustained
Ventral
Dorsal
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Notes:
APPENDIX K
INTRAOPERATIVE
CLINICAL EFFECT ON DYSKINESIA
The nomenclature of the contact names in the DBS lead varies greatly. Typically,
the DBS lead contains four contacts in a row along the long axis of the DBS lead. This
form names the contacts ventral-most for the deepest contact, ventral for the next con-
tact, dorsal for the contact above the ventral contact, and dorsal-most for the highest
contact. The user may wish to enter the specific numbering of the contacts appropriate
to the DBS lead used.
Visual analog rating scales are provided for four forms of dyskinesia assessable in
the operating room. These forms may differ in the region of the body affected or by the
joints rotated to produce the dyskinesia. In these scales, 0 is normal and 4 is the most
severe (see c hapter 12). As the patient may have multiple and varied forms of dyskine-
sia simultaneously, the scales are assigned to four forms of dyskinesia presented by the
patient.
388
Intraoperative Macrostimulation for Dyskinesia
Electrode
selection
Case Pulse Rate Volts Adverse effects, Effect on Effect on Effect on Effect on
Dorsal-Most mosst
width or where and dyskinesia 1 dyskinesia 2 dyskinesia 3 dyskinesia 4
Ventral-most
mA transient vs
Ventral
Dorsal
sustained
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Notes:
INDEX
abdominal reflex, superficial, 225f ankle, schematic representation of range of

AC-coupled amplifiers, 99 motion, 210f
acetylcholine, 5 anodes, 214
Achilles’ deep tendon reflex, schematic ansa lenticularis, photomicrograph of, 274f
representation of, 224f anterior commissures (AC), 32, 32f
AC-PC line, 32 anterior ventral caudal thalamus, 196–197
AC power lines, interference from, 107, 107f antidromic action potential, 76f, 214
action potentials, 71, 145f antidromic conduction, 77
antidromic, 76f aphasia, subcortical, 189
extracellular, failure to record, 221 arrays, horizontal, 139–140
extracellular, intraoperative microelectrode artifacts, 93–94, 106, 111–112
recordings of, 2 defined, 107
generation, 74f from mechanical sources, 110, 113f
graded vs., 67–75 from patient speech, 122
potential effects of filters on, 98f preventing, 112–115
recording, 75–77 reducing, 119–123
regenerative process, 72, 73f removing, 100
A/D conversion, digitization noise from, 94 shielding to control, 115–117, 116f
air athethosis, 210
entry to intracranial vault, and brain shift, 177 audio monitoring, 103
entry to intracranial vault, and misplaced leads, cleaning signal to, 120f
316 average stimulation, and optimal benefit, 225–226
MRI scan showing in skull, 290f, 315f axial view, postoperative MRI scan, 310f
alertness, inability to maintain, 222 axon hillock, 71
alligator clip, 113 axon of neuron, 63–64, 64f
attaching to conductor, 114f junction with cell body, 71
alternating current, 55–61
Alzheimer’s disease, 3 basal ganglia
amperes, 53 dopamine release in, 6f
amplifiers, 88–89 schematic of nuclei subset, 151f
microelectrode output impedance relative to basal ganglia-thalamic-cortical system, 151f
input impedance of amplifier system, 92 neurons in, 144
noise internal to, 93 stimulation of, 150
analog conductor lines, 93 battery
analog multiplexers, noise from, 94 construction of, 53f
anesthesia, 43–45 neuron as, 65, 65f
anions, 49 bed-of-nails effect, 140
392 / / I ndex
benzodiazepines, 222 complications, 227

biophysics, of electrical activation, 144, 146 Essential tremor, 228–237, 230f, 238–247,
biosafety, 146–148 248–260, 312–316
bipolar biphasic stimulation, 155 Parkinson’s disease, 262–276, 278–290,
bipolar microelectrodes 292–300, 302–310
macrostimulation through indifferent or cathode follower, 92
reference contact, 138–139 cathode ray tubes (CRTs), 109
schematic representation of, 21f, 145f cathodes, 214
bipolar recordings, 75, 76f tissue activation volume surrounding, 17
border cells cations, 49
for globus pallidus interna, 181 cerebellar climbing fibers, 150
microelectrode recordings containing, 135f, 182f cerebellar outflow tremor, 205–207
recording consistent with, 294 cerebellar relay nucleus, 188f
schematic of coronal section through, 358f cerebrospinal fluid, and brain shift, 42
schematic of saggital section through, 359f, 360f cervical dystonia, torticollis in, 208f
brachium conjunctivum, inadvertent charge balance biphasic waveforms, 147f
stimulation, 328 chassis ground, 117f, 118
bradykinesia chemical reaction, and electrostatic field, 50
impact of stimulation current propagating to children
corticospinal tract, 216 DBS for dystonia, 174
tests to assess, 203 ventral intermediate nucleus DBS in, 191
brain chloride (Cl–), 49
electrical stimulation, 145f chorea, 210
as electronic device, 48 choroid plexis, 31
injuries from electrical stimulation, 147 clinical assessments, 201–226
reducing function, 4 of corticospinal and corticobulbar stimulation,
return path for current, 155 211–213
brain shift, 16, 42–43, 163 of dystonia, 207–209
air entry to intracranial vault, 177 of essential tremor and cerebellar outflow tremor,
electrophysiological mapping to accommodate, 205–207
198 hints, 225–226
hematoma and, 222 of hyperkinetic syndromes, 210–211
MRI or CT scan to identify degree, 324 of Parkinson’s disease, 202–205
from opening skull, 14 of potential complications, 221–224
of speech and language, 213–214
cables of tic disorders, 211
length of, 113 common-mode rejection, 100–101
shielding, 116 effects of, 101f
capacitance noise removal, 100f
development of, 58f compact fluorescent lights, 115
at electrode-brain interface, 22 complacency, 24–25
capacitive reactance, 56–57, 58, 90, 154 concentration gradient, 52, 72
effect of frequency on, 60f concordant paresthesias
effect on changes in source waveform, 59f mechanism in response to stimulation, 11f
frequencies in source voltage and, 59 possible mechanisms of, 149f, 194f
schematic representation of effects, 154f conduction, orthodromic or antidromic, 77
capacitor, saturation of, 60 conductors, 56
cardioballistic effects, 111, 121 attaching alligator clip to, 114f
Cartesian system, 36–37 capacitive coupling between, 110, 110f
reference plane for, 32, 33f forces resisting flow of electrical charges, 53
case studies metals, 146
Index / / 393
confidence, 140–141, 225, 319 globus pallidus interna algorithm, 357–367

connectors, effects of straight and twisted importance of, 2–3
pair, 114f lead placement, 190
consciousness, and reticular activating nomenclature of contact names in lead, 377
system, 222 vs. pharmacological therapies, 3–6
constant current, 23 qualities of therapeutic effects, 6
vs. constant voltage, 153–154 relief from, 2
constant voltage subthalamic nucleus algorithm, 329–334, 330f
vs. constant current, 153–154 success, 1, 319
impedance effects on, 154 trajectory angle of lead, 12
coronal plane value of experience-based training, xiii
angle of entry, 177f ventral intermediate thalamic algorithm,
photomicrograph section, 274f 335–356, 336f
postoperative MRI scan, 310f deep sensation, 196–197
shallow trajectory in, 268 delusions, 222
corticobulbar tract dendrites, 64f
clinical assessments of stimulation, dendritic tree, 130
211–213 depolarization, 69
damage to, 222–223 depression, 3, 219
corticospinal tract, 215 from medial subthalamic nucleus stimulation,
clinical assessments of stimulation, 211–213 169
damage to, 222–223 device errors, 324
impact on independence of finger movements, dexmedetomidine, 45
203, 204f diagnostic test, intraoperative monitoring as,
involvement, and excessively posterior 325–327
stimulation, 220 digital conductor lines, 93
propagation of stimulation current to, 201 digitization noise, from A/D conversion, 94
stimulation, and dystonia assessment, 209 dimpling of skin surfaces, 212
coulombs, 48 diplopia, 218
crosstalk, 93 direct pathway, 151f
cup task, for tremor evaluation, 206 disconjugate gaze, 218–219
current, 107 from hemorrhage, 224
alternating or fluctuating, 55–61 discordant paresthesias
in brain, return path for, 155 mechanism in response to stimulation, 11f
constant, vs. constant voltage, 153–154 possible mechanisms of, 149f, 194f
current density, and safety limit for electrical documentation
stimulation, 147–148 missing information, 316
cutoff frequency, 95–96, 95f for Parkinson’s, microelectrode recordings,
262–263, 278–280, 282–285, 292–296,
daisy chain, 112f, 115 302–304
DBS. See Deep Brain Stimulation (DBS) for tremor, microelectrode recordings, 228, 231,
DC-coupled amplifiers, 99 233, 248–251, 255–256
DC recordings, 93 value of detail, 141
decision tree for possible trajectories, 170–171 See also forms
Deep Brain Stimulation (DBS) doors, shielding for, 117
adverse effects, 187 dopamine receptors, 4
clinical evaluation of target, 214–221 dopamine, release in basal ganglia, 6f
clinical success of, x dorsal thalamus, 195
decision to undergo lead revision surgery, double vision, 218
319–320 drift, by DC-coupled amplifiers, 99
future for, 317–318 dysarthria, 189
394 / / I ndex
dyskinesia fields affecting, 49

intraoperative macrostimulation form, 386–387 noise from movement, 108
scale for, 210 electrophysiological analyses, automated and
dysphagia, 189 sophisticated, 131
dystonia, 2, 173, 182 electrophysiologically based therapies, future
in children, DBS for, 174 for, 317
clinical evaluation, 207–209 electrostatic charge, 49
estimating degree of departure from neutral electrostatic fields, 106
position for wrist, 209f and chemical reaction, 50
globus pallidus interna vs. globus pallidus eloquent areas, 30
externa, 181 ependymal surface, traversal, and DBS lead
intraoperative macrostimulation form, 381–383 deflection, 163
stimulation of corticospinal tract and assessment epilepsy, 3
of, 209 equipment errors, 324
Essential tremor
earth ground, 117f, 118 case studies, 228–237, 230f, 248–260, 312–316
electrical activity case studies, MRI scan, 236f, 237f, 247f
biophysics of, 144, 146 clinical evaluation, 205–207
in neurons, 4 euphoria, 219
electrical charges, 53, 63 Evidence-Based Medicine
electrical devices lack of data on monitoring methods, 13–14
motors, 115 level 1 clinical trials, difficulty conducting, 159
series for recording neuronal activity, 92f evoked potentials, hypothetical use to localize
electrical field, effects of lack of control, 322f homuncular representation, 84f
electrical potentials, creation, 54f excitation, as neuronal response to DBS, 142
electrical stimulation external landmarks, registration of, 31
of brain, 145f extracellular action potentials
brain injuries from, 147 failure to record, 221
potential mechanisms underlying effects, 145f intraoperative microelectrode recordings of, 2
schematic representation of, 145f extremities, assessment of muscle tone in upper,
electric current, controlling flow, 66f 204–205, 205f
electricity, 48–51 eye movements
electrocautery, 115 impaired, 218–219
electrode popping, 111 schematic representation of stimulation effects, 212f
electrode recordings, 63–85
electrodes, 31 facia dystonia, assessment of, 213
configuration, 156, 214 false localization, 21
microstimulation and type, 137–138 Faraday cage, 116
schematic representation of regional divisions, feedback circuits, 89
185f fiducials, 33, 35
schematic representation of systems, 76f filtering, 94–97
translation methods for tissue volume accessible effect of entering filter with cutoff moved, 96f
by, 40f effects on signals entering filter, 95f
See also trajectory of electrode microelectrode recordings, 88
electromagnetic fields, 106 potential effects on neuronal action potential, 98f
electromagnetic interference. See interference finger movements, impact of corticospinal tract
electromotive force, 52 on independence, 203, 204f
electronic device, brain as, 48 flicker noise, 93
electronics, 51–52 fluctuating current, 55–61
of recording and stimulating systems, 52–54 flumazenil, 222
electrons to reverse midazolam effects, 44
Index / / 395
fluorescent lights, 115, 120 schematic of saggital section through, 359f, 360f
noise from, 109 sensorimotor anatomy of, 11f
secondary dystonia due to, 2 G protein-coupled channel, 68f
foil shield, grounding, 116 graded potentials, 69f
forms for macrostimulation action vs., 67–75
for dyskinesia, 386–387 grnneuromod.com, 377
for dystonia, 381–383 grounding, 117–119, 117f
for Parkinson’s disease, 377–378 connected electrical devices, 112f
for tics, 384–385 ground loop currents, noise from, 111
for tremor disorders, 379–380
forms for microelectrode recordings half-cell potential, 53
for globus pallidus interna DBS, 371–373 hallucinations, 222
for subthalamic nucleus DBS, 368–370 halo, 35
for ventral intermediate thalamus, 374–376 hand gestures, for mouth and tongue
Fourier transform of signal, frequencies within movement, 133f
waveforms, 60 harmonics, 97
frame-based stereotactic systems, 35–36 HDE (humanitarian device exemption), 327
frameless stereotactic systems, 34f, 35–36 head homuncular region
internal landmarks referenced to, 33 avoiding, 254
frequency-dependent impedance, on waveform joint rotations of jaw to identify representations,
shape, 61f 132, 133f
“funny feeling” sensation, 211–212, 270, 275 hematoma. See intracerebral hematoma;
intracranial hematoma
globus pallidus externa high-frequency-pause neurons, 180, 181f
high-frequency-pause neurons in, 134 high-pass filter, 96–97, 97f
indication of neuronal density at recording site, hissing, 119, 121
180 from neuronal activity in optic nerve axons, 134
neuronal activity consistent with, 286 homuncular representation, hypothetical use of
photomicrograph of, 274f evoked potentials to localize, 84f
recording site in, 298 horizontal arrays, 139–140
schematic of coronal section through, 358f hum
schematic of saggital section through, 359f, 360f in audio monitor, 108
globus pallidus interna, 10, 172–186 frequency of, 119
clinical evaluation during stimulation, 220 humanitarian device exemption (HDE), 327
cognitive and psychological adverse effects, 321 Huntingdon’s disease, secondary dystonia
Deep Brain Stimulation algorithm, 357–367 due to, 2
Deep Brain Stimulation surgery, 9 hyperkinetic syndromes, 182
extracellular action potentials, 181–182, 183f clinical evaluation, 210–211
form for microelectrode recordings, 371–373 hyperpolarization, 69
lateral, regional anatomy around, 179f
mean discharge frequency, 165 impedance, 55, 58, 60
neuronal activity consistent with, 286 acute changes, and capacitance, 22
overactivity, and Parkinson’s disease, 5 effects on constant voltage, 154
photomicrograph of, 274f frequency-dependent, on waveform shape, 61f
physiologically defined or symptomatically and impedance matching, 90–94
defined optimal target, 173–176 relationship to inductive reactance, 57f
reconstructing regional anatomy according to risks and effects of mismatched, 91
physiology, 183–185 schematic representation of effects, 153f
recording site in, 298 tungsten microelectrodes and change in, 137
regional neuronal physiology, 179–183 indifferent electrical contact, 21–22
schematic of coronal section through, 358f induction, 55
396 / / I ndex
inductive reactance, 56, 90 joint rotations

relationship to impedance, 57f of jaw, 132, 133, 133f
injury current, from trauma to neuron, 75 neurons responding to, 252
injury discharge, 128
interference knee reflex, schematic representation of, 224f
from AC power lines, 107, 107f
in “on the fly” recording, 82 landmarks, for electrode targets, 31
internal capsule, posterior limb, 167–168, 183, language, clinical assessment of, 213–214
187, 197–198 lateral ventricles, avoiding penetration of, 30–31
avoiding DBS lead placement too close to, 173 lead placement, consequences of too shallow,
border detection by bracketing, 358 242, 242f
identifying anterior border, 286 LED bulbs, 115
neuronal activity absence when passing through, lenticular fasciculus, photomicrograph of, 274f
163f ligand-gated ion channels, 67, 68f
photomicrograph, 274f light, patient experiencing flashes of, 220
stimulation of, 213 lipophilic agents, 44
internal landmarks, registration of, 31 liquid crystal displays (LCDs), 109
interneuronal communication, neurotransmitters local field-potential recordings, 18–20, 77–79
for, 4 effect of synchronization, 19f
intracerebral hematoma, 221 effects of spatial resolution, 83f
absence of neuronal extracellular action practical principles, 82–84
potentials and, 325 local field potentials, 328
quiet region and, 136 determination of signal generator’s location, 79f
intracranial air, and misplaced leads, 316 effects of spatial resolution, 87f
intracranial hematoma mechanisms underlying, 78f
and brain shift, 43 recording system for, 93
clinical manifestations, 221–222 low-pass filter, 95f
lack of neuronal recordings from, 234
intraoperative clinical assessments, components, 201 macrostimulation, 1, 22–23, 142
intraoperative neurophysiological monitoring intraoperative form for dyskinesia, 386–387
as diagnostic test, 325–327 intraoperative form for dystonia, 381–383
ethical concerns, 23–27 intraoperative form for Parkinson’s disease,
importance of, 2–3 377–378
purposes to test stimulation, 142 intraoperative form for tics, 384–385
signals used, 64–67 intraoperative form for tremor disorders,
intraoperative neurophysiological monitoring 379–380
preparations, 30–47 vs. microstimulation, 175
anesthesia and sedation, 43–45 minimizing risk of adverse effects, 191
brain shift issue, 42–43 purpose of, 148–149
postimplantation studies, 45–46 responses to, 274
targeting methods, 31–35 through DBS lead, 155–156
translocation methods, 36–41, 37f through indifferent or reference contact of
ions, 49 bipolar microelectrode, 138–139
current allowing change in concentrations, 69 magnetic field, 49, 55–56, 56f
flow from negatively charged stimulation magnetic force, iron-filling experiment
electrode tip, 50, 51f demonstrating lines, 56f
isolation transformers, 102, 102f magnetic resonance imaging (MRI)
axial section of scan, 33f
jaw for determining electrode trajectory, 30
joint rotations, 132, 133, 133f mapping, neurophysiological, anesthetic agents
tightening, corticobulbar fibers and, 300 interfering with, 45
Index / / 397
mean stimulation parameters, problems from micro-otomy effect, 20

testing only with, 161 micropallidotomy effect, 175, 359
mechanical artifact, 113f microstimulation, 2, 20–22
mechanical sources, artifact generation from, 110 and electrode type, 137–138
medial globus pallidus interna, 164 vs. macrostimulation, 175
DBS leads inplanted in, 165–166 minimizing risk of adverse effects, 191
MRI scan depicting trajectories through, 178f and optimal target for subthalamic nucleus, 160
medial lemniscus, placing DBS lead too close, 159 purpose of, 148–149
medial subthalamic nucleus, 169 micro-subthalamotomy effect, 268
risk of stimulating current propagation, 164 midazolam, 35, 44, 221
median stimulation parameters, problems from morphine and, 45
testing only with, 161 misinformation, 3
medications monopolar recordings, 75, 102–103, 155
delusions or hallucinations from withholding, 222 mood, intraoperative stimulation influence in,
and neuronal activity suppression, 325 219f
side effects, 4–5 morphine, 45, 221
Meige’s syndrome, 213 motor homunculus, 87f
mereological fallacy, 4 mouth movement, hand gestures for, 133f
microdrive chatter, 111 Movement Disorders Society-United Parkinson’s
microelectrode recordings, 17–18, 125–141 Disease Rating Scales Part III, 203, 205
for DBS lead placement, 1 multiple sclerosis, 207
documentation for Parkinson’s, 262–263, multipolar recordings, 19
278–280, 282–285, 292–296 muscle contraction, 212
documentation for tremor, 228, 231, 233, muscle tone, assessment in upper extremity,
248–251, 255–256 204–205, 205f
epistemic status of, 12–17
example of high-frequency, high-density, 127f, negative current, 50
169f negative feedback, 74
example of low-frequency, low-density, 126f, 167f neurohumoral paradigm, 4
example of moderate-frequency, neuroimaging target, 12
moderate-density, 127f, 168f neurological disorders, 3–6
examples of regular and irregular extracellular neuronal action potentials
action potentials, 130f potential effects of filters on, 98f
form for globus pallidus interna DBS, 371–373 recording, 75–77
form for subthalamic nucleus DBS, 368–370 neuronal activities
form for ventral intermediate thalamus, 374–376 lack when passing through internal capsule
neuronal density within single site, 127–128 posterior limb, 163f
principles, 79–82, 125–126 parameters, and anatomical localization of
reports in case studies, 227 recording site, 126–134
microelectrode recording systems neuronal-behavioral correlations, 131–134
contacts, 75 examples from microelectrode recording sites,
schematic of typical, 26f 132f
screenshot made during DBS surgery, 104f neuronal current, sequence manipulation, 67
microelectrodes neuronal discharge frequency, 126
cardioballistic effects on, 111 neuronal extracellular action potentials, failure to
for identifying sensorimotor regions, 8–9 record, 221
malfunctioning, 136 neuronal regularity, 128
materials for, 20 neuronal switch, schematic representation of, 67f
output impedance relative to input impedance of neurons, 63–64
amplifier system, 92 in basal ganglia-thalamic-cortical system, 144
typical tip, 88 as battery, 65, 65f
398 / / I ndex
neurons (Cont.) paresthesias

density within single microelectrode recording discordant and concordant, in response to
site, 127–128 stimulation, 11f
depolarization as consequence of membrane inability to infer somatotopic representation
tear, 113f from, 246
electrical activity in, 4 from placing DBS lead too close to medial
injury current from trauma to, 75 lemniscus, 159
lysis, 75 possible mechanisms of, 149f
recording electrical activity, 50 schematic representation of low-threshold, 219f
schematic representation of, 64f threshold increase on, 246
separation of charges from, 64 from ventral caudal thalamus stimulation, 220
switch, 66f Parkinson’s disease, 2, 3, 173
systems activation, 150–152 case studies, 262–276, 278–290, 292–300,
neuropharmacology, ix 302–310
neurophysiological mapping, anesthetic agents case studies, MRI scan, 272f, 273f
interfering with, 45 clinical assessment of, 202–205
neurosurgeon, 319 examples of high-frequency-pause neurons
and neurophysiologist, 27 indicating globus pallidus interna, 135f
neurotransmitters, ix, 4, 142, 144 globus pallidus interna neuronal activity, 182
noise, 93–94, 106, 107–111 globus pallidus interna overactivity and, 5
from capacitive coupling, 110 high-frequency-pause neurons and, 181f
defined, 107 intraoperative macrostimulation form, 377–378
from electron movement, 108 neuronal discharge patterns, 180
from ground loop currents, 111 stimulation of zona incerta, and symptom
preventing, 112–115 improvement, 168
reducing, 119–123 testing tremors in, 204
removal with common-mode rejection, 100f patella deep tendon reflex, schematic
shielding to control, 115–117, 116f representation of, 224f
and signal of interest, 98f patient
notch filter, 97, 99f age, and depth of optimal DBS trajectory, 191
artifacts from speaking, 122
obsessive-compulsive disorder, 2 bioelectrical signals in, 118
oculomotor nerve, 8 regional anatomy visualization, 171
ohms, 53 stress in operating room, 215
Ohm’s law, 54, 55 pedunculopontine nucleus, 328
water-propulsion analogy, 55f perinatal injury, secondary dystonia due to, 2
omission bias, 23 persistent posturing, assessment of, 208
operational amplifier (op am), 88, 89f pharmacological agents
optic tract, 134, 286 side effects, 4–5
extracellular action potentials within, See also medications
182–183 pharmacological therapies, vs. Deep Brain
optimal target Stimulation (DBS), 3–6
for globus pallidus interna, 173–176 phosephene production, 149
identifiable, 14–15, 15f phosephenes, threshold, 139f
physiologically defined for subthalamic nucleus, for distance estimates, 138f, 139f
160–162 physical movement, noise from, 122–123
for ventral intermediate thalamus, 190–192 platinum-iridium microelectrodes, 20, 137–138,
optogenetics, 328–329 146
oralis thalamus, ventral lateral posterior, 197 polar coordinates, orthogonality in, 36
orthodromic conduction, 77 posterior commissures (PC), 32, 32f
oxidation, 50 posterior ventral caudal thalamus, 197
Index / / 399
postive current, 50 return path, for current in brain, 155

postoperative X-rays, 46 risks after surgery, 24
postsynaptic potentials, 150–151 rotational translocation systems, 37
graded, 70 “Runaway Trolley Car dilemma,” 23
synchronization, 77
potassium (K+), 49, 64 safety
power, 79 biosafety, 146–148
preparations. See intraoperative stimulation, 156–157
neurophysiological monitoring preparations sagittal plane
principles angle of entry, 164, 176, 177, 177f, 193
method validity for seeing target, 323–324 DBS lead in, 166
need for, 318 schematic representation of, 165f
possible applications, 327–329 schematic representation of effects of excessively
success definition, 318–320 shallow trajectory, 178f
target assessment relative to nontarget areas, sampling issues, 79
321–322 saturation of capacitor, 60
target assessment relative to proposed sedation, 43–45
intervention, 320–321 semi-microelectrode recordings, 17–18
of target definition, 320 epistemic status of, 12–17
visualization of target, 322–323 semi-microelectrode recording system, schematic
probe’s view, 31 of typical, 26f
propofol, 35, 44 semi-microelectrodes
protein channel, 67 for identifying sensorimotor regions, 8–9
pseudotransitivity, fallacy of, 5 indifferent contact on, 138–139
psychiatric disorders, 3–6 sensitivity, of electrode location testing, 326
Purkinje cells, 150 sensorimotor hormunculus, in lateral globus
putamen pallidus interna, 182
neuronal activities, 180 sensorimotor region, of subthalamic nucleus, 7
photomicrograph of, 274f sensory stimulation, 133
recording site in, 298 shearing force, 82
schematic of coronal section through, 358f of array, 140
schematic of saggital section through, 359f, 360f shielding, 115–117, 116f
short circuits, 118
quality control, 25 signal ground, 118
signal of interest, and noise, 98f
radiation, 106 signal-to-noise ratio, impact of, 3
reactance, 91 silent region, 136–137
recordings, contamination with AC power-line sine waves, 80f, 81f
noise, 96 sodium (Na+), 49, 64
Redox reaction, 50, 146 opening conductance channels, 144
biosafety of, 146–148 somatotopic organization, in subthalamic
reduction, 50 nucleus, 166–167
reference plane, for Cartesian system, 32, 33f spatial resolution, 87
refractory period, 74 spatial summation, 71, 151
regenerative process, of action potentials, 72, 73f schematic representation of, 71f, 152f
regional anatomy, visualizing patient’s, 170 specificity of electrode location testing, 326
reports spectrogram, 81f
blank forms, 227 speech
See also forms clinical assessments of, 213–214
resting tremor, assessing, 206 macrostimulation effects on, 236
reticular activating system, and consciousness, 222 stimulation affecting, 218
400 / / I ndex
spherical coordinate system, 37f superficial reflexes, 225f

spike discriminator, 104f, 120f injury to corticobulbar and corticospinal tract
square wave pulse, 59 and, 224
stainless steel stimulating electrodes, 146 supraharmonic, 97
step-down transformers, 109 surgery, risks after, 24
step-up transformers, 109 symptoms, DBS relief of, 160
stereotactic systems, frame-based and frameless, synchronization, local field potentials dependent
35–36 on, 18
stimulating electrodes, precise, accurate
placement of, 1 tactile sensation, 197
stimulation tardive dyskinesia, secondary dystonia due to, 2
average, and optimal benefit, 225–226 tardive dystonia, secondary dystonia due to, 2
safety, 156–157 target
stimulus electrical field, effects of lack of control, algorithm to determine where to search next for, 26
321f importance of localization, 7–12
stress method validity for seeing, 323–324
delusions or hallucinations from, 222 visualization methods, 322–323
to patient in operating room, 215 target-centered systems, 38f
stroke, 3 rotational, 39f
subarachnoid space, air entry in, 42 temporal summation, 70, 71, 151
subcortical aphasia, 189 schematic representation of effects, 152f
substantia nigra pars compacta, 164 tension pneumocephalus, 164–165
substantia nigra pars reticulata, 164, 169 MRI scan, 275f, 315f
microelectrode recording example, 169f thalamostriate veins, 31
microelectrode skirting of, 166f thalamus
neuronal activities, 307 anterior, neuronal activity, 167
subthalamic nucleus, 150, 159–171 anterior ventral caudal, 196–197
clinical evaluation during stimulation, 215–219 DBS surgery for, 9
cognitive and psychological adverse effects, 321 dorsal, 195
deep brain stimulation algorithm, 329–334, 330f photomicrograph of, 274f
form for microelectrode recordings, 368–370 schematic of saggital section, 336f, 337f
lateral, 168–169 sensorimotor anatomy of ventral intermediate
lateral vs. medial, 215 nucleus of, 9f
medial, 169 ventral lateral posterior oralis, 197
medial, risk of stimulating current propagation, ventral oral posterior, 336
164 See also ventral caudal thalamus; ventral
microelectrode traversal, 166f intermediate nucleus of thalamus
regional anatomy of, 7f, 268f therapeutic window, 215, 359
regional anatomy of electrode trajectory, demonstratining reasonable, 161
162–169 goal of providing sufficiently wide, 319
regional anatomy reconstruction to conform threshold, minimum, vs. threshold producing
with physiology, 170–171 persistent side effects, 192
sensorimotor region, 159 tic disorders
sensorimotor region neuronal activity, 130 clinical evaluation, 211
size, 7, 159 intraoperative macrostimulation form, 384–385
as small DBS target, 87–88 timing, importance in DBS, 6
somatotopic organization in, 166–167 tinnitus, 3
spatial accuracy and precision required for DBS tissue activation, volume surrounding single
of, 8f cathode, 17
success, definition, 318–320 tissue volume accessible by electrode system,
sulci, trajectories planned to bypass, 31 translation methods, 40f
Index / / 401
tongue movement, hand gestures for, 133f Unified Parkinson Disease Rating Scales, 377
tonic contractions US Food and Drug Administration, 2, 191
from DBS lead in globus pallidus interna, 172 humanitarian device exemption (HDE), 327
detecting change in thresholds, 268
increasing risk of, 164 vagus nerve, 5–6
low stimulation threshold for, 358 variance, 131
schematic representation of low-threshold, 218f ventral caudal thalamus
torticollis, estimating degree of departure from anterior, 196–197
neutral position, 208f homuncular organization, 338f
Tourette’s syndrome, 2, 211 as optimal target, 335
tragedy of partial improvement, 24 paresthesias from DBS lead too close to tactile
training, value of experience-based, xiii region, 337–338
trajectory of electrode paresthesias from stimulation, 220
anatomy for ventral intermediate nucleus of posterior, 197
thalamus, 193–195 ventral intermediate nucleus of thalamus, 149,
angle, and traversal of several structures, 184 187–200
consequences of too shallow, 276f anatomy, 187, 188f
decision tree for, 170–171 clinical evaluation during stimulation, 220–221
effects of angle, 163f criteria for physiologically defined or
estimating densities in, 252 symptomatically defined optimal target,
excessively shallow on sagittal plane, 178f 190–192
for globus pallidus interna, 176–179 Deep Brain Stimulation (DBS) algorithm,
MRI for angle, 30 335–356, 336f
neuronal recording site density within, 129–131 form for microelectrode recordings, 374–376
physiology in subthalamic nucleus, 167–169 homuncular organization, 189f
regional anatomy for subthalamic nucleus, reconstructing regional anatomy according to
162–167 physiology, 198–199
risk from tangential, 254 regional anatomy of electrode trajectory,
schematic representation with excessive shallow 193–195
areas, 194f regional neuronal physiology of structures,
shallowness in sagittal plane, 359 195–198
transformers, 108–109, 108f targeting lateral region of, 10
transient neuronal activities, 127–128, 128f ventral lateral posterior oralis thalamus, 197
with decreasing amplitude from NA+ ion influx, ventral oral posterior thalamus, 336
129f ventral tier, microelectrode entry in, 252
representative example, 180, 195f visual analog rating scales, 377
translocation visual disturbances, DBS lead placement to
horizontal method of, 41f prevent, 173
methods of, 36–41, 37f visualization of regional anatomy, 184
tremor, 2 voltage, constant, vs. constant current, 153–154
clinical evaluation, 205–207 voltage-gated channels, 70, 72
demonstration in clinic for types of, 206f deactivation, 74
intraoperative macrostimulation form, 379–380 voltage sensitive channels, 70
primary orthostatic, 10 volts, 52
quantitating amplitude, 207f
testing in Parkinson’s disease, 204 waveform shape, frequency-dependent
tungsten microelectrodes, 20, 137–138, 146 i mpedance on, 61f
turnkey systems weakness
danger from, 26–27 schematic representation of asymmetric facial,
troubleshooting, 86 224f
“tyranny of partial benefit,” 320 signs of, 223
402 / / I ndex
“weird” effect in patient, 177, 220 X-rays

worst-case scenario, in postoperative DBS care, of DBS lead fracture, 46f
anticipating, 161 postoperative, 46
wrist in dystonia, estimating degree
of departure from neutral position, zero-voltage ground, 118
209f zona incerta, 168

Intraoperative Neurophysiological Monitoring For Deep Brain Stim

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Intraoperative Neurophysiological Monitoring For Deep Brain Stim

Enviado por

Direitos autorais:

Formatos disponíveis

INTRAOPERATIVE

Erwin B. Montgomery Jr., MD

1. Importance of intraoperative neurophysiological monitoring 1

2. Preparations for intraoperative neurophysiological monitoring 30

3. Basic concepts of electricity and electronics 48

5. Microelectrode and semi-microelectrode recordings: Electronics 86

6. Noise and artifact 106

7. Microelectrode recordings: Neuronal characteristics and

8. Microstimulation and macrostimulation 142

9. The subthalamic nucleus 159

10. The globus pallidus interna nucleus 172

11. The ventral intermediate nucleus of the thalamus 187

12. Clinical assessments during intraoperative

13. Cases 227

14. Future intraoperative neurophysiological monitoring 317

Appendix A. Subthalamic nucleus deep brain stimulation algorithm  331

Appendix B. Ventral intermediate thalamic deep brain stimulation

Appendix C. Globus pallidus interna deep brain stimulation algorithm  359

Appendix D. Microelectrode recording form for subthalamic nucleus

Appendix E. Microelectrode recording form for globus pallidus interna  373

Appendix F. Microelectrode recording form for ventral intermediate

Appendix G. Intraoperative macrostimulation for clinical effect

Appendix H. Intraoperative macrostimulation for clinical effect in

Appendix I. Intraoperative macrostimulation for clinical effect

Appendix J. Intraoperative macrostimulation for clinical effect on tics  386

Appendix K. Intraoperative macrostimulation for clinical effect on

have the effect of reinforcing misplaced presumptions. Paraphrasing Claude Bernard,

the impression that intraoperative neurophysiological monitoring is routine, when it is

THE GROWING IMPORTANCE OF DEEP BRAIN STIMULATION AND

DEEP BRAIN STIMULATION’S FUNDAMENTAL DIFFERENCE FROM

Fundamentally different from, and therefore an important supplement to, available

IMPORTANCE OF TARGET LOCALIZATION AND THE ROLE OF

FIGURE 1.2 Schematic representation of the regional anatomy of the subthalamic nucleus

CORONAL SOMATOTOPIC ARRANGEMENT AXIAL SOMATOTOPIC ARRANGEMENT

FIGURE 1.4 Sensorimotor anatomy (homunculus) of the ventral intermediate nucleus of the

presents a useful example. It involves a homuncular representation consisting of the

FIGURE 1.5 Possible mechanism of discordant and concordant paresthesias in response to

THE EPISTEMIC STATUS OF MICROELECTRODE OR

The different methods of intraoperative neurophysiological monitoring being

Volume of Optional Sites

Confidence Area in the Horizontal Plane

A second premise, which is derived from physiological experience and computational

MICROELECTRODE AND SEMI-MICROELECTRODE RECORDINGS

Identifying and analyzing extracellular action potentials arising from individual

semi-microelectrode recordings generally do not allow such recognition, because the

LOCAL FIELD-POTENTIAL RECORDINGS

same source tend to result in increased synchronization and increased magnitude of

One can accomplish microstimulation by passing an electrical current through a micro-

Indifferent or reference electrode

FIGURE 1.10 Schematic representation of a bipolar microelectrode. The active electrode is the

In the context of intraoperative neurophysiological monitoring, macrostimulation

Use of constant current stimulation helps to mitigate the effects of changes in

plays a role in a physician’s approach to DBS. Chapter 14 contains detailed discussion

FIGURE 1.11 Schematic of a typical microelectrode and semi-microelectrode recording sys-

commercial FDA-approved systems improve considerably on the “home-built” sys-

The question whether invasive methods of intraoperative neurophysiological

The intraoperative neurophysiologist must understand the methods involved in

FRAME-BASED AND FRAMELESS STEREOTACTIC SYSTEMS

The interoperative neurophysiologist must understand the mechanisms for manipu-

FIGURE 2.5 An example of a spherical coordinate system in which a vector of length L and

(a) (f) (g)

(c) (d) (e)

Another primary translocation system employs a rotational system whose cone

Appendix A. Subthalamic nucleus deep brain stimulation algorithm 331

Appendix B. Ventral intermediate thalamic deep brain stimulation

Appendix C. Globus pallidus interna deep brain stimulation algorithm 359

Appendix D. Microelectrode recording form for subthalamic nucleus

Appendix E. Microelectrode recording form for globus pallidus interna 373

Appendix F. Microelectrode recording form for ventral intermediate

Appendix G. Intraoperative macrostimulation for clinical effect

Appendix H. Intraoperative macrostimulation for clinical effect in

Appendix I. Intraoperative macrostimulation for clinical effect

Appendix J. Intraoperative macrostimulation for clinical effect on tics 386

Appendix K. Intraoperative macrostimulation for clinical effect on

Chapter 3 covered electricity’s fundamental characteristics, and chapter 4 covered its