Nezves and E1eC1z1caJ S11uJ11on

A I1hVI1I YAVANV NLIL L M NYL"AI?VIV

John P.18lOt8n0JOn Y.L!mk, 1t.
Department of Electrical and Computer Engineering
Rice Univerity
P.O. Box I82,Houston, Tx 772\
A new 00buIed paametr moel of the myelinated nerve fiber
i s presented. Te model assumes a tiaxial cable form, with separate
noal, paranoal and interoal regions, a myelin sheath indepndent
of the underlying interoal membrane and a periaxonal conductance
pathway. Um previous coaxial cable equivalent models, this new
tiaxial form alSo allows for inclusion of ionic channels into the inter
noal membI3ue. The mel uses ionic channel dynamics for te
mammal and 3mphibi an derived from rcent experimental studies. The
model reproduces conduction behavior seen in exprimental and previ
ous modeling efforts. Most imprtantly, the moel reprouces the
deplarizing 8l|I0nIta\ described by several i nvestigators whIch
may have signicance in determination of the bchavIorofthe nerve
fiber as m0Itpe action plcnuals are conducted or as multe stimuli
are impsed. This aftertential cannot be produced by me previous
coaxial models. The infuece of the characteristcs of the myelin
sheat, underlying interodal membrane ad the priaxonal conduc
tance pathway on the behavior of the neIVe fbe is explored.
Previous models of m myelinated nerve fber were based
on a coaxial cable equivalent rpresentation which assumed that
the interodal region of the nere fibr could be segmented into
passive lumped paralleled conductances and capacitances. This
assumption cannot be made if the periaxonal space is to be con
sidered and the myelin sheath is to be represented independent of
the interodal membrane. Therefore a traxial cable equivalent
form was assumed, consisting of an extacellular, intracellular
and periaxonal longitudinal conductive pathways and separate
transverse pathways representng the myelin and m membrane
beneat. A transverse conductance was also included, represent
ing the nodal gap region just outside the active node of Ranvier.
Nol PGenUal (mV)
ISO ~ . -
o
:
- -1- - - , - -_ I - - -1- - -
I I I I
-1- - - . ---1 - _1 __ _
i
Nol Gap PocaU_1 (mV)
0.1 -1---._-
O ~~
".1
- - - I ___ 1 _ _ _
, I I
I
-0.4 . - _| - __ . __ L__ _ _ -_
I i I
e.+

Transmembrane Potenllal (mV)

-+--- 1 -4--{

,, - - : ---1---
I
I
1 ___ 1 __ _
I I
TransmyeUn Poenllal (mV)
1-----<--
0- - -
10
Figure 1: Temporal plolof the top8gating action ptential at memid-
pont ofthe nerve fiber.
Recent studies have examined the dynamics of m ionic
channels in the amphibian and mammalian myelinated nerve
fi ber in the nodal as well as the interodal regions. In order to
produce a model based on the most recent exprimental data
available, these dynamics were included in the model. In the
cases where only experimental data was available, a Marquardt
non-linear parameter estimation method was used to ft the
steady-state and time constant relations describing te ionic
channel kinetics to the data.
An implicit integration technique was employed to solve
for m system of equarions reprsenting te nerve fiber. An
example of the behavior of the model can be seen in fgures I
and 2. Figure is an example of a propagating action ptential
vs. time as seen across the node, nodal gap, myelin and intero
dal membrane. ou the presence of most of the potential across
the myelin sheath which has the greater impedance. Also note
the presence of a depolarizing afterptential at the node and
across the interodal membrane. In fgure 2 the spatal distn'bu
tion of the action potential across the myelin sheath and the
underlying membrane can be seen. The fiber was stimulated at
the right end, propagaton is from right t left.
In conclusion, this new triaxial model allows for a detailed
functional representaton of the myelinated nerve fber. Both the
mammalian and amphibian nere fbers can be represented. The
model generates propagating action potentals that correlate with
measured potentials as well as previous modeling results. In
additon, the model prduces deplarizing aferpotentials as seen
in experimental studies which cannot be produced by previous
coaxial nerve fber models.
This work supported in pan by NSF Lrant EC&4064J6.
Traommranc Poential (mV)
......... ...
::ujII
I I ,
I I I
50 ! i
0
.1
s:lj(m)
0.6 0.8
FIgure 2: Spatial plot of the propagating CtIon ptential at O.3r after
stimaIation.
IEEE ENGINEERING IN MEDICINE & BIOLGY SOCIETY 10TH ANNUAL INTERATIONAL CONFERENCE--1923
CH2566-8/88/0000--1923 $1.00 0 1988 IEEE