Escolar Documentos
Profissional Documentos
Cultura Documentos
Preface
Imaging
Joel S. Schuman, MD
Guest Editor
0896-1549/04/$ see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.ohc.2004.02.001
From the Neocortex to the Eye, by Amiram Grinvald, PhD, Tobias Bonhoeffer, PhD, Ivo Vanzetta,
PhD, Ayala Pollack, MD, Eyal Aloni, MD, Ron Ofri,
MD, and Darin Nelson, PhD, reviews emerging technology for measurement of ocular blood flow and
oximetry. The potential of the technology described
is enormous, and this section promises to translate
what was once science fiction to science.
Joel S. Schuman, MD
University of Pittsburgh School of Medicine
UPMC Eye Center
203 Lothrop Street
Eye and Ear Institute, Suite 816
Pittsburgh, PA 15213, USA
E-mail address: schumanjs@upmc.edu
* Corresponding author.
E-mail address: huangd@ccf.org (D. Huang).
tical line to the OCT image. A corneal image produced by the Zeiss OCT 1 system (Fig. 1) has visible
motion artifacts because the image is acquired over
1 second. It also has coarse pixel grain because it
contains only 100 vertical lines. The 0.8-mm wavelength employed for retinal scanning is close to
visible wavelengths and cannot directly visualize
angle structures owing to scattering loss through the
limbus [22]. A widely useful anterior segment OCT
system would need to overcome both limitations.
This article presents the results using a high-speed
OCT prototype developed collaboratively between
Professor Joseph Izatt (Duke University), Professor
Andrew Rollins (Case University), and the authors at
the Cleveland Clinic Foundation. The system has a
scan rate of 4000 lines per second, which is fast
enough for biometric applications. It uses a longer
wavelength of 1.3 mm, which decreases scattering
through turbid tissues and allows visualization of
angle structures [23]. The system is useful in a range
of clinical applications from laser-assisted in situ
keratomileusis (LASIK) surgery to narrow angle
glaucoma. OCT of the anterior segment is still in its
infancy; therefore, it is discussed at the end of this
article in a section on future prospects.
Background
The precursor technology of OCT [1], optical
coherence domain reflectometry (OCDR), had its
earliest demonstration in biomedical applications
with the measurement of corneal thickness [24].
OCDR is an optical ranging technique, or a method
for measuring the distance between the measurement
0896-1549/04/$ see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/S0896-1549(03)00103-2
makes the accurate measurement of anterior chamber width and other biometric parameters possible.
Anterior chamber width measurement is clinically important for sizing angle-supported anterior
chamber intraocular lenses. With the increasing use
of refractive phakic intraocular lenses, accurate sizing
becomes an important issue. An intraocular lens that
is too large can press on the iris root and produce pupil
ovalization; an intraocular lens that is too small can
lead to lens movement, decentration, corneal endothelial damage, and iritis [49]. The traditional method
for sizing uses the external corneal diameter, which
is assumed to correspond to the internal anterior
chamber width. Typically, the intraocular lens length
is chosen to be the corneal diameter plus a constant
such as 1 mm.
When the wide-field CAS OCT system was used,
it was possible to measure the internal width of the
anterior chamber directly (Fig. 5). Anterior chamber
width as measured by OCT was compared with
corneal diameter measured by a Holladay cornea
gauge in 20 normal subjects. The anterior chamber
width was 12.53 F 0.47 mm (mean F SD). The
difference of the anterior chamber width from the
corneal diameter was 0.75 F 0.44 mm with a range
Future developments
Optical coherence topography is a versatile tool
for visualization and measurement of corneal and anterior segment anatomy. It has the potential for improving the functions currently served by Placido-ring
corneal topography, slit-scanning corneal topography,
ultrasound imaging, and ultrasound pachymetry.
Keratorefractive surgery, anterior chamber biometry, and angle assessment are some of the applications that should benefit from CAS OCT. The
commercialization and general availability of this
technology should increase applications through the
ingenuity of many practitioners.
Fig. 5. Anterior chamber (AC) imaged with the wide-field (15-mm) setting on the slit lamp mounted high-speed CAS OCT
system. The AC width is measured between angle recesses; its depth is measured from the corneal apex to lens apex.
References
[1] Huang D, Swanson EA, Lin CP, et al. Optical coherence tomography. Science 1991;254(5035):1178 81.
[2] Allemann N, Chamon W, Silverman RH, et al. Highfrequency ultrasound quantitative analyses of corneal
scarring following excimer laser keratectomy. Arch
Ophthalmol 1993;111(7):968 73.
[3] Pavlin CJ, Easterbrook M, Harasiewicz K, Foster FS.
An ultrasound biomicroscopic analysis of angle-closure
glaucoma secondary to ciliochoroidal effusion in IgA
nephropathy. Am J Ophthalmol 1993;116(3):341 5.
[4] Reinstein DZ, Silverman RH, Coleman DJ. High-frequency ultrasound measurement of the thickness of
the corneal epithelium. Refract Corneal Surg 1993;
9(5):385 7.
[5] Reinstein DZ, Silverman RH, Trokel SL, Allemann N,
Coleman DJ. High-frequency ultrasound digital signal
processing for biometry of the cornea in planning
phototherapeutic keratectomy [letter] [published erratum appears in Arch Ophthalmol 1993 Jul;111(7):926].
Arch Ophthalmol 1993;111(4):430 1.
[6] Reinstein DZ, Silverman RH, Trokel SL, Coleman DJ.
Corneal pachymetric topography. Ophthalmology
1994;101(3):432 8.
[7] Reinstein DZ, Silverman RH, Rondeau MJ, Coleman
DJ. Epithelial and corneal thickness measurements by
high-frequency ultrasound digital signal processing.
Ophthalmology 1994;101(1):140 6.
[8] Riley SF, Nairn JP, Maestre FA, Smith TJ. Analysis
of the anterior chamber angle by gonioscopy and by
ultrasound biomicroscopy. Int Ophthalmol Clin 1994;
34(3):271 82.
[9] Reinstein DZ, Silverman RH, Sutton HF, Coleman
DJ. Very high-frequency ultrasound corneal analysis
identifies anatomic correlates of optical complications of lamellar refractive surgery: anatomic diagnosis in lamellar surgery. Ophthalmology 1999;106(3):
474 82.
[10] Jalbert I, Stapleton F, Papas E, Sweeney DF, Coroneo
M. In vivo confocal microscopy of the human cornea.
Br J Ophthalmol 2003;87(2):225 36.
[11] Bauer NJ, Wicksted JP, Jongsma FH, March WF, Hendrikse F, Motamedi M. Noninvasive assessment of
the hydration gradient across the cornea using confocal Raman spectroscopy. Invest Ophthalmol Vis Sci
1998;39(5):831 5.
[12] Lemp MA, Dilly PN, Boyde A. Tandem-scanning
(confocal) microscopy of the full-thickness cornea.
Cornea 1985;4(4):205 9.
[13] Masters BR, Farmer MA. Three-dimensional confocal
microscopy and visualization of the in situ cornea.
Comput Med Imaging Graph 1993;17(3):211 9.
[14] Masters BR, Paddock S. In vitro confocal imaging of
the rabbit cornea. J Microsc 1990;158(Pt 2):267 74.
[15] Ichijima H, Petroll WM, Jester JV, Cavanagh HD.
Confocal microscopic studies of living rabbit cornea
treated with benzalkonium chloride. Cornea 1992;
11(3):221 5.
[16] Boker T, Sheqem J, Rauwolf M, Wegener A. Anterior chamber angle biometry: a comparison of Scheimpflug photography and ultrasound biomicroscopy.
Ophthalmic Res 1995;27(Suppl 1):104 9.
[17] Yaylali V, Kaufman SC, Thompson HW. Corneal
thickness measurements with the Orbscan Topography System and ultrasonic pachymetry. J Cataract
Refract Surg 1997;23(9):1345 50.
[18] Lattimore Jr MR, Kaupp S, Schallhorn S, Lewis RT.
Orbscan pachymetry: implications of a repeated measures and diurnal variation analysis. Ophthalmology
1999;106(5):977 81.
[19] Auffarth GU, Tetz MR, Biazid Y, Volcker HE. Measuring anterior chamber depth with Orbscan Topography
System. J Cataract Refract Surg 1997;23(9):1351 5.
[20] Boscia F, La Tegola MG, Alessio G, Sborgia C. Accuracy of Orbscan optical pachymetry in corneas with
haze. J Cataract Refract Surg 2002;28(2):253 8.
[21] Cairns G, McGhee CN, Collins MJ, Owens H, Gamble
GD. Accuracy of orbscan II slit-scanning elevation topography. J Cataract Refract Surg 2002;28(12):2181 7.
[22] Izatt JA, Hee MR, Swanson EA, et al. Micrometerscale resolution imaging of the anterior eye in vivo
with optical coherence tomography. Arch Ophthalmol
1994;112(12):1584 9.
[23] Radhakrishnan S, Rollins AM, Roth JE, et al. Real-time
optical coherence tomography of the anterior segment
at 1310 nm. Arch Ophthalmol 2001;119(8):1179 85.
[24] Huang D, Wang J, Lin CP, Puliafito CA, Fujimoto JG.
Micron-resolution ranging of cornea anterior chamber
by optical reflectometry. Lasers Surg Med 1991;11(5):
419 25.
[25] Koop N, Brinkmann R, Lankenau E, Flache S, Engelhardt R, Birngruber R. [Optical coherence tomography
of the cornea and the anterior eye segment]. Ophthalmologe 1997;94(7):481 6.
[26] Asiyo-Vogel MN, Koop N, Brinkmann R, et al. [Imaging of laser thermokeratoplasty lesions by optical low
coherence tomography and polarization microscopy
after Sirius Red staining]. Ophthalmologe 1997;94(7):
487 91.
[27] Maldonado MJ. Undersurface ablation of the flap for
laser in situ keratomileusis retreatment. Ophthalmology 2002;109(8):1453 64.
[28] Maldonado MJ, Munuera JM, Garcia-Layana A,
Moreno J, Aliseda D. Optical coherence tomography
(OCT) evaluation of the corneal cap and stromal bed
features after LASIK for high myopia. Presented at the
American Academy of Ophthalmology Annual Meeting. New Orleans, November 8 11, 1998.
[29] Maldonado MJ, Ruiz-Oblitas L, Munuera JM, Aliseda
D, Garcia-Layana A, Moreno-Montanes J. Optical
coherence tomography evaluation of the corneal cap
and stromal bed features after laser in situ keratomileusis for high myopia and astigmatism [In Process Citation]. Ophthalmology 2000;107(1):81 7
[discussion: 88].
[30] Muscat S, McKay N, Parks S, Kemp E, Keating D.
Repeatability and reproducibility of corneal thickness
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
0896-1549/04/$ see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.ohc.2003.12.001
Angle-closure glaucoma
Iris apposition to the trabecular meshwork is the
final common pathway of angle-closure glaucoma,
which represents a group of disorders. This condition
can be caused by one or more abnormalities in the
relative or absolute sizes or positions of anterior segment structures, or by abnormal forces in the posterior
segment that alter the anatomy of the anterior segment. Forces are generated to cause angle closure
in four anatomic sites: the iris (pupillary block), the
ciliary body (plateau iris), the lens (phacomorphic
glaucoma), and behind the iris by a combination of
various forces (malignant glaucoma and other posterior pushing glaucoma types). Differentiating these
affected sites is the key to provide effective treatment.
UBM is extremely useful for achieving this goal.
Angle occludability. Examining eyes with narrow
angles requires careful attention to the occludability
of the angle. Although provocative testing, such as
dark room gonioscopy, is useful for detecting the
angle occludability, it is now rarely used, because it
is subjective, time consuming, and prone to falsenegative results owing to the difficulty of standardizing the slit-lamp light intensity. With UBM, dark room
provocative testing can be performed in a standardized environment generating objective results by providing information on the state of the angle under
normal light conditions and its tendency to occlude
spontaneously under dark conditions (Fig. 2).
Fig. 2. Occludable angle with dark room provocative test. (A) The anterior chamber angle is slit-like opened (arrows) under
a lighted condition. (B) The angle is completely occluded (arrows) under a dark condition.
Pupillary block. Pupillary block is the most common type of angle-closure glaucoma. At the iridolenticular contact, resistance to aqueous flow from the
posterior to the anterior chamber creates an unbalanced relative pressure gradient between the two
chambers, pushing the iris up toward the cornea
(Fig. 3A). This abnormal resistance causes anterior
iris bowing, angle narrowing, and acute or chronic
angle-closure glaucoma. The other anterior segment structures and their anatomic relationships remain normal.
Laser iridectomy equalizes the pressure gradient
between the anterior and posterior chambers and
flattens the iris. The result is a widened anterior
chamber angle (Fig. 3B).
Plateau iris. A plateau iris configuration occurs
owing to a large or anteriorly positioned ciliary
body (pars plicata), which pushes the iris root mechanically up against the trabecular meshwork
(Fig. 4). The iris root may be short and inserted
anteriorly on the ciliary face, creating a narrow and
crowded angle. The anterior chamber is usually of
medium depth, and the iris surface looks flat or
slightly convex, just like in a normal eye. With indentation gonioscopy, the double-hump sign is observed. The peripheral hump results from the rigid
presence of the ciliary body holding the iris root; the
central hump represents the center part of the iris
resting over the anterior lens surface. The space between the two humps represents the area between
the ciliary processes and the endpoint of iridolenticular contact. These findings can be confirmed by
performing indentation UBM (Fig. 5), a special technique that imposes mild pressure on the peripheral
cornea with the skirt of a plastic eyecup so that one
can simulate indentation gonioscopy [6].
Phacomorphic glaucoma. Anterior subluxation of
the lens may lead to angle-closure glaucoma because
of the lens pushing the iris and ciliary body toward the
trabecular meshwork.
Malignant glaucoma. Malignant glaucoma, also
known as ciliary block or aqueous misdirection,
presents the greatest diagnostic and treatment challenge. Forces posterior to the lens push the lens iris
diaphragm forward, causing angle closure. UBM
clearly shows that all anterior segment structures are
displaced and pressed tightly against the cornea with
or without fluid in the supraciliary space (Fig. 6).
Other causes of angle closure. Iridociliary body
cysts can produce angle-closure glaucoma. The anterior chamber angle is occluded partially or intermittently owing to singular or multiple cysts (Fig. 7).
UBM is extremely useful in making the diagnosis
in these cases. Other entities, such as iridociliary tumor, enlargement of the ciliary body owing to inflammation or tumor infiltration, or air or gas bubbles after
intraocular surgery, may also present angle closure.
Open-angle glaucoma
The only type of open-angle glaucoma that shows
characteristic findings on UBM is the pigment dispersion syndrome. In this familial autosomal dominant
disease, mechanical friction between the posterior
iris surface and anterior zonular bundles releases iris
pigment particles into aqueous flow. These particles
are deposited on structures throughout the anterior
10
from dehiscence ciliary body detachment, which requires a different management approach [9].
Ocular trauma
Ocular trauma often limits the visibility of the
ocular structure owing to the presence of hyphema.
Accurate assessment of the structural damage and
locating small foreign bodies can be a challenging
task when clear direct visualization is not achieved.
UBM can be performed over a plano soft contact lens
to minimize the risk of further injury with eyecups
segment. The diagnostic triad consists of a Krukenberg spindle, radial transillumination defects of the
midperipheral iris, and pigment deposition on the
trabecular meshwork.
Typical UBM findings associated with this condition include a widely opened angle, an iris with
slight concavity (bowing posteriorly), and increased
iridolenticular contact (Fig. 8). As is true in pupillary
block, there is a relative pressure gradient between
the anterior and posterior chamber; however, because
the anterior chamber is the one that holds higher
pressure, this condition is called reverse pupillary
block [7]. Laser iridotomy eliminates this pressure
gradient, resulting in a flattened iris [8].
Abnormalities of the iris and ciliary body
Ultrasound biomicroscopy is helpful in differentiating solid from cystic lesions of the iris and ciliary
body (see Fig. 7 and Fig. 9). The size of these lesions can be measured, and the extent to which they
invade the iris root and ciliary face can be evaluated.
In hypotony cases, UBM can distinguish tractional
11
12
Physical resolution is often confused with measurement precision. Physical resolution specifies how
close together two objects can be located yet still
Fig. 12. Intraocular foreign body. (A) Foreign body (arrow head) with a material that consists of multiple cavities
inside (ie, wood and concrete) generates shadowing artifact
(arrow) by absorbing ultrasound power. The iris image is
masked by shadowing. (B) Hard and dense foreign body
(arrow head) (ie, glass and metal) creates comet tail artifact
(arrow) owing to multiple internal reflections. The iris image is disrupted by the comet tail artifact. (Adapted from
Laroche D, Ishikawa H, Greenfield D, et al. Ultrasound biomicroscopic localization and evaluation of intraocular foreign bodies. Acta Ophthalmol Scand 1998;76(4):491 5;
with permission.)
the theoretical size of the pixel. Measurement precision can be better than physical resolution by oversampling the signal.
Commercially available instruments provide lateral and axial physical resolution of approximately
50 and 25 mm, respectively. The resolution of the
Paradigm device is slightly better than that of the
OTI device. The theoretical lateral and axial measurement precision on the standard UBM monitor
(864 432 pixels) is approximately 6 and 12 mm.
Although UBM cannot distinguish two small objects
less than 25 mm apart along the axial scanning line,
it can still measure the distance between two objects
far enough apart ( > 25 mm, such as corneal thickness,
anterior chamber depth) with 12-mm precision.
13
Table 1
Parameters proposed by Pavlin et al [1]
Measurement accuracy
Name
Abbreviation
Description
Pavlin et al [2] reported good qualitative agreement of UBM images with histologic sections. Quantitatively, Maberly et al [18] showed good agreement
by measuring the distance from the anterior margin
of peripheral choroidal melanomas to the scleral spur
on UBM images and histologic sections.
Pierro et al [19] compared the corneal thickness
measured by UBM versus ultrasound and optical
pachymetry. The UBM measurement was similar to
the ultrasound pachymetry, whereas optical pachymetry showed a poor correlation with UBM and ultrasound pachymetry. Urbak [20] reported similar
results. Additionally, a specially prepared plastic material was measured with UBM and scanning electron microscopy. The axial and lateral accuracies of
UBM measurements were good and reliable.
Angle opening
distance
AOD
TIA q 1
Trabecular ciliary
process distance
TCPD
Iris thickness
ID1
Iris thickness
ID2
Measurement reproducibility
Iris thickness
ID3
ICPD
IZD
ILCD
ILA q 2
Distance between
the trabecular
meshwork and the
iris at 500 mm
anterior to the
scleral spur
Angle of the
angle recess
Distance between
the trabecular
meshwork and the
ciliary process at
500 mm anterior
to the scleral spur
Iris thickness at
500 mm anterior
to the scleral spur
Iris thickness at
2 mm from the
iris root
Maximum iris
thickness near the
pupillary edge
Distance between
the iris and the
ciliary process
along the line
of TCPD
Distance between
the iris and the
zonule along the
line of TCPD
Contact distance
between the iris
and the lens
Angle between
the iris and the
lens near the
pupillary edge
14
compromise. The UBM Pro 2000 (Paradigm Medical Industries, Salt Lake City, Utah) can measure
the AOD in a semi-automated fashion. It has dramatically improved overall reproducibility (coefficient of
variation, 7.3 to 2.5; Hiroshi Ishikawa, MD, unpublished data, 1998).
Fig. 15. Pavlins measurement parameters (see Table 1). (A) The angle opening distance (AOD) is defined as the length of the
line drawn from the point on the corneal endothelial surface 500 mm anterior to the scleral spur to the iris surface perpendicular to
the corneal endothelial surface. The trabecular iris angle (TIA, q1) is defined as an angle formed with the apex at the iris recess
and the arms passing through the point on the meshwork 500 mm from the scleral spur and the point on the iris perpendicularly
opposite. (B) The trabecular ciliary distance (TCPD) is defined as the distance between a point 500 mm from the scleral spur
and the ciliary process on the line that is perpendicular through the iris. The iris thickness (ID1) is defined along this line, as is
the iris ciliary process distance (ICPD). Iris thickness also can be measured 2 mm from the iris root (ID2) and at its thickest
point near the margin (ID3). The iris zonule distance (IZD) is defined as a part of theTCPD at a point just clearing the ciliary
process. The length of iris lens contact (ILCD) and the angle at which the iris leaves the lens surface (iris lens angle; ILA, q2) are
easily measured.
15
observer measuring greater corneal thickness, assuming that each observer would choose the same point as
an endothelial border. In general, repeated measurement by the same observer is reasonably reproducible.
Quantitative measurement methods
Methods proposed by Pavlin and colleagues
Pavlin et al [1] established various quantitative
measurement parameters as standards (Table 1,
Fig. 15). The position of the scleral spur is used as a
reference point for most of their parameters, because
this is the only landmark that can be distinguished
consistently in the anterior chamber angle region.
Iris concavity/convexity
Potash et al [24] introduced a parameter to evaluate
the dynamic configurational change of the iris. A line
is created from the most peripheral point to the most
central point of iris pigment epithelium. A perpendicular line is then extended from this line to the iris
pigment epithelium at the point of greatest concavity
or convexity (Fig. 16).
An improved method for assessing the anterior
chamber angle
There is one problem with AOD measurement,
Pavlins classical method of assessing the angle opening, which treats the iris surface as a straight line.
Fig. 17 shows two schematics of the angle, demonstrating exactly the same value for the AOD and the
trabecular iris angle (TIA). Nevertheless, it is obvious that the angle on the right is gonioscopically
narrower and more likely to be occludable than the
angle on the left; therefore, irregularities of iris contour and curvature need to be taken into account.
Ishikawa et al [25] defined the angle recess area
(ARA) as the triangular area bordered by the anterior
iris surface, corneal endothelium, and a line perpendicular to the corneal endothelium drawn to the iris
surface from a point 750 mm anterior to the scleral
spur (Fig. 18). In this way, the iris irregularity is properly accounted for in the measurement.
The semi-automated software in the UBM Pro
2000 also calculates the ARA. After the observer
selects the scleral spur, the program automatically
processes the image, detects a border, and calculates
the ARA. The program plots consecutive AODs from
the base of the angle recess to 750 mm anterior
to the scleral spur and performs linear regression
analysis of consecutive AODs, producing two figuresthe acceleration (or slope) and the y-intercept.
The acceleration describes how rapidly the angle
is getting wider, using the tangent of the angle instead
of degrees as the unit. In other words, the acceleration estimates the general shape of the angle, shallow
or wide. The y-intercept refers to the distance between the scleral spur and the iris surface along the
perpendicular to the trabecular meshwork plane. This
generalized value describes the angle opening at the
level of the scleral spur. Although these parameters
may seem similar to the AOD and TIA, there is a
fundamental difference between them. Because the
acceleration and the y-intercept are purely mathematical calculations based on linear regression analysis of the consecutive AODs, they can be negative
numbers, which is impossible for the physically measured AOD and TIA. A negative number for the acceleration means that the angle has an almost normal
configuration at its peripheral part and becomes very
shallow, or is attached to the cornea, at its central part
(ie, appositional angle closure starting at Schwalbes
line with space remaining in the angle recess)
(Fig. 19). A negative y-intercept means that the angle
recess is very shallow or is attached to the cornea at
its periphery, whereas it is relatively wide centrally
16
Fig. 17. Limitation of the conventional angle opening distance (AOD) measurement. (A) and (B) have exactly the same value
for the AOD and trabecular iris angle (TIA, q1). Nevertheless, the angle in (B) is gonioscopically narrower and is more likely
to be occludable than the normal-appearing angle in (A).
[28]. Ishikawa et al [25] measured the ARA, acceleration, and y-intercept under standardized dark and
light conditions and reported that the more posterior
the iris insertion on the ciliary face, the less likely the
provocative test would be positive. Esaki et al [29]
found that the anterior chamber angle opening in
normal Japanese eyes narrowed with age in a crosssectional study.
Ultrasound biomicroscopy also provides a powerful tool to evaluate the effect of drug instillation on
the anterior chamber angle, iris, and ciliary body.
Kobayashi et al [30] found that the angle opening
increased after the instillation of pilocarpine in eyes
with narrow angles but decreased in eyes with a wider
17
Fig. 18. Angle recess area (ARA). The ARA is defined as a triangular area bordered by the anterior iris surface, corneal
endothelium, and a line perpendicular to the corneal endothelium drawn from a point 750 mm anterior to the scleral spur to the
iris surface.
Fig. 19. Negative acceleration in ARA analysis. The linear regression analysis of ARA shows negative acceleration, meaning
that the angle almost has a normal configuration at its peripheral part and becomes very shallow or is apposed to the cornea at its
central part (ie, the appositional angle closure began at the level of Schwalbes line).
18
Fig. 20. Negative y-intercept in ARA analysis. The linear regression analysis shows a negative y-intercept, indicating that the
angle recess is very shallow or is attached to the cornea at its periphery, whereas it has a relatively wide angle recess centrally
(ie, plateau iris and synechial closure).
Summary
Ultrasound biomicroscopy technology has become
an indispensable tool in qualitative and quantitative
assessment of the anterior segment. Advances in software design and algorithms will improve theoretical understanding of the pathophysiology of anterior
segment disorders. Future applications of quantitative techniques will yield important information regarding mechanisms of angle closure, improving
understanding of the dynamic functions of the iris,
accommodation, presbyopia, and other aspects of
anterior segment physiology and pathophysiology.
References
[1] Pavlin CJ, Harasiewicz K, Foster FS. Ultrasound biomicroscopy of anterior segment structures in normal
and glaucomatous eyes. Am J Ophthalmol 1992;113:
381 9.
[2] Pavlin CJ, Sherar MD, Foster FS. Subsurface ultrasound microscopic imaging of the intact eye. Ophthalmology 1990;97:244 50.
[3] Pavlin CJ, Harasiewicz K, Sherar MD, et al. Clinical use
of ultrasound biomicroscopy. Ophthalmology 1991;98:
287 95.
[4] Tello C, Potash S, Liebmann J, et al. Soft contact lens
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
19
20
* Corresponding author.
E-mail address: cpuliafito@med.miami.edu
(C.A. Puliafito).
0896-1549/04/$ see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.ohc.2003.12.002
22
23
Case 3. A 61-year-old Caucasian man with worsening vision OD for 6 months presented for a second
opinion on the diagnosis of age-related macular degeneration. He had a past ocular history of inferotemporal branch retinal artery occlusion OD 5 years
before presentation. Best-corrected visual acuity was
20/100 OD and 20/20 OS. Confrontation fields were
intact OS, with loss of the superior hemifield OD.
Amsler grid testing showed sparing of central fixation
OD. Ocular motility was within normal limits, and a
24
25
26
Vein occlusion
Macular edema is a common complication of vein
occlusion. Although it can be detected by biomicroscopy or fluorescein angiography, blocking defects
by hemorrhage or media opacity may obscure detection of this vision-threatening complication. OCT has
the advantage of imaging macular edema and quantifying the increase in retinal thickness despite these
limiting factors. In addition to macular edema, subretinal fluid accumulation and neurosensory retinal
detachment can be detected. The presence of persistent edema unresponsive to initial therapy (ie, laser
photocoagulation) can establish the need for more
invasive intervention (ie, intravitreal triamcinolone).
OCT has a pivotal role in providing objective quantitative information to detect the presence of macular
edema that may be contributing to vision loss and to
monitor the effect of therapeutic interventions, such
as focal laser photocoagulation, intravitreal triamcinolone injection, adventitial sheathotomy, and pars
plana vitrectomy. It also can be helpful in differentiating patients whose vision is limited by photoreceptor damage/retinal atrophy and not macular edema.
Case 7. An 83-year-old Caucasian woman presented with worsening vision OD for 2 weeks. Ocular
history was significant for a central retinal vein occlusion OD diagnosed 2 months previously and presumed secondary to lymphoma. Best-corrected visual
27
28
Fig. 24. Cystoid macular edema and serous pigment epithelial detachment associated with choroidal neovascular
membrane OS.
29
30
Case 10. A 34-year-old Hispanic woman was referred for surgical treatment of a macula-involving
retinal detachment. Past ocular history was significant
for high myopia (-10 OU). At presentation, bestcorrected visual acuity was 20/60 OD and 20/20 OS.
Confrontation fields showed loss of the superonasal
field OD. Ocular motility was within normal limits,
and there was no afferent pupillary defect. Anterior
segment examination showed clear lenses OU. Intraocular pressures were 15 and 14, respectively. Dilated
fundus examination showed an inferotemporal retinal
detachment that extended into the macula OD. OCT
imaging to evaluate the cause of decreased vision to
the 20/60 level OD revealed a central thickness of
396 mm and CME despite foveal attachment (Fig. 31).
The patient was diagnosed as having a maculainvolving retinal detachment and underwent a scleral
buckle placement, cryotherapy, and perfluoropropane
gas injection. At the 3-month follow-up examination,
best-corrected visual acuity remained at 20/70 OD,
although the retina seemed to have undergone suc-
Summary
[7]
[8]
References
[9]
[1] Huang D, Swanson EA, Lin CP, Schuman JS, Stinson
WG, Chang W, et al. Optical coherence tomography.
Science 1991;254(5035):1178 81.
[2] Hee MR, Izatt JA, Swanson EA, Huang D, Schuman
JS, Lin CP, et al. Optical coherence tomography of the
human retina. Arch Ophthalmol 1995;113(3):325 32.
[3] Drexler W, Sattmann H, Hermann B, Ko TH, Stur M,
Unterhuber A, et al. Enhanced visualization of macular
pathology with the use of ultrahigh-resolution optical
coherence tomography. Arch Ophthalmol 2003;121:
695 706.
[4] Puliafito CA, Hee MR, Lin CP, Reichel E, Schuman
JS, Duker JS, et al. Imaging of macular diseases with
optical coherence tomography. Ophthalmology 1995;
102(2):217 29.
[5] Ip MS, Baker BJ, Duker JS, Reichel E, Baumal CR,
Gangnon R, et al. Anatomical outcomes of surgery for
idiopathic macular hole as determined by optical coherence tomography. Arch Ophthalmol 2002;120:29 35.
[6] Shahidi M, Ogura Y, Blair NP, et al. Retinal thickness
[10]
[11]
[12]
[13]
31
analysis for quantitative assessment of diabetic macular edema. Arch Ophthalmol 1991;109:1115 9.
Moss SE, Klein R, Klein BE. The 14-year incidence
of visual loss in a diabetic population. Ophthalmology
1998;105(6):998 1003.
Hee MR, Puliafito CA, Wong C, Duker JS, Reichel E,
Rutledge B, et al. Quantitative assessment of macular
edema with optical coherence tomography. Arch Ophthalmol 1995;113(8):1019 29.
Yang CS, Cheng CY, Lee FL, Hsu WM, Liu JH. Quantitative assessment of retinal thickness in diabetic
patients with and without clinically significant macular
edema using optical coherence tomography. Acta Ophthalmol Scand 2001;79(3):266 70.
Hee MR, Puliafito CA, Duker JS, Reichel E, Coker JG,
Wilkins JR, et al. Topography of diabetic macular
edema with optical coherence tomography. Ophthalmology 1998;105(2):360 70.
Martidis A, Duker JS, Greenberg PB, Rogers AH,
Puliafito CA, Reichel E, et al. Intravitreal triamcinolone for refractory diabetic macular edema. Ophthalmology 2002;109(5):920 6.
Rivellese M, George A, Sulkes D, Reichel E, Puliafito
CA. Optical coherence tomography after laser photocoagulation for clinically significant macular edema.
Ophthalmic Surg Lasers 2000;31(3):192 7.
Lattanzio R, Brancato R, Pierro L, Bandello F, Laccher
B, Fiore T, et al. Macular thickness measured by optical coherence tomography (OCT) in diabetic patients.
Eur J Ophthalmol 2002;12(6):482 7.
Imaging in glaucoma
Daniel M. Stein, BA, Gadi Wollstein, MD, Joel S. Schuman, MD*
UPMC Eye Center, Department of Ophthalmology, University of Pittsburgh School of Medicine, The Eye and Ear Institute,
Suite 816, 203 Lothrop Street, Pittsburgh, PA 15213, USA
more, numerous studies have shown that glaucomatous field abnormalities may be preceded by structural
changes of the ONH [11 13] and nerve fiber layer
[14 17].
Because glaucomatous damage is largely irreversible, it is imperative to identify accurately eyes
with early structural changes, because they are at risk
for continued injury. It has been suggested that the
earlier glaucoma is detected and treated, the greater
the likelihood that medical or surgical intervention
will delay or prevent the progression of glaucomatous
neuropathy and subsequent functional impairment
[18 20]. This assumption underscores the need for
accurate and reproducible quantitative evaluation of
the eye. Beyond early detection, quantitative imaging
devices might be a more sensitive way to detect glaucomatous progression when compared with clinical
qualitative assessments.
Work during the past two decades has resulted
in the development and implementation of several
imaging technologies designed to detect glaucomatous neuropathy at early stages of disease. This review outlines several of the most current imaging
technologies, including confocal scanning laser ophthalmoscopy (CSLO), scanning laser polarimetry
(SLP), optical coherence tomography (OCT), and
the retinal thickness analyzer (RTA). The technologic
underpinnings, sensitivity and specificity, and clinical
studies of each method are discussed.
* Corresponding author.
E-mail address: schumanjs@upmc.edu (J.S. Schuman).
Confocal scanning laser ophthalmoscopy is a realtime imaging technique that is used to produce three-
0896-1549/04/$ see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/S0896-1549(03)00102-0
34
Reproducibility
Measurements with the HRT have been found to
be highly reproducible in numerous studies [24 29].
Rohrschneider et al reported a standard deviation for
each measuring point of 30 F 6 mm in glaucoma
patients and 22 F 6 mm in normal subjects [26].
The coefficient of variation (COV) of the various
HRT measurements in this study ranged from 2.9%
to 5.2% in glaucomatous eyes to 3.3% to 4.6% in
normal eyes.
Weinreb et al [30] showed that reproducibility
was improved when multiple scans were averaged
into a single data set, recommending what is now
35
Fig. 2. HRT scan of normal subject (OD). On the upper left image, the topographic ONH map is shown. The red area marks the
cup, and the green and blue areas mark the neuroretinal rim. The upper right image, the reflectance image, is shown with the
tracing (green) of the optic disk margin and the outline (red) of the optic cup. All of the ONH sectors are marked with the green
checkmark, signifying that they are within the normal limits based on Moorfields regression analysis. At the bottom of the
figure, a graph depicts the surface height along the contour line on the ONH margin. Note the normal double-hump
appearance, with higher thickness in the superior and inferior regions.
36
Fig. 3. HRT scan of subject with glaucoma (OD). On the upper left image, note the extensive cupping of the ONH and the
thinnest area of neuroretinal rim in the inferotemporal region. Moorfields regression analysis marked the inferotemporal and the
nasal inferior sectors as abnormal and the temporal and the superior sectors as borderline. The typical double-hump
configuration in the lower graph was eliminated mainly owing to tissue loss at the ONH poles. These findings correspond with
the superior nasal visual field defect shown in Fig. 1 and show additional superior ONH damage that might precede the
perimetric appearance.
bilities of HRT in differentiating normal and glaucomatous eyes. Cup shape measure, rim area, and cup
volume seem to be the best HRT parameters in
discriminating between eyes, although considerable
overlap exist between groups [46,47]. Improved discrimination between groups has been found by combining several parameters and ONH segmentation
[48 53]. Two of these methods have been incorporated into the HRT II: Mikelbergs discriminating
analysis and Moorfields regression analysis [50,51].
When these methods are used, the specificity and
sensitivity to differentiate between normal subjects
and patients with early glaucoma reach 78% and
89%, respectively, with Mikelbergs discriminating
analysis and 96.3% and 84.3%, respectively, with
Moorfields regression analysis [50,51]. Implementing more sophisticated mathematical approaches
such as neural networks might provide further improvement [54].
The capability of the HRT to detect ONH glaucomatous changes before the appearance of a visual
field defect was evaluated by Kamal et al [55] who
examined a small group of ocular hypertensive individuals in whom reproducible visual field changes
subsequently developed. Various parameters were
found to be significantly different from those of normal subjects.
Chauhan et al [56] have described a new statistical technique for detecting changes in topography
over time with HRT using an analysis of variance
technique. When this technique is used, a higher
frequency of progression is detected by the HRT
when compared with visual field progression, which
might suggest higher sensitivity of this method for
longitudinal evaluation; however, the specificity remains uncertain [57].
Strengths and limitations
Strengths
Rapid simple operation of the device (HRT II)
Three-dimensional, topographic representation
of the ONH
No pupil dilation necessary
Advanced data analysis capability built-in to
machine
Limitations
The use of a reference plane is required.
Manual tracing of the ONH margin must be
37
38
Fig. 4. GDx VCC scan of normal subject (OD). This figure shows the fundus image (top left), a pseudocolor representation of the
nerve fiber layer thickness (middle left), and a map of areas that deviate from normal thickness values (bottom left). The nerve
fiber layer graph shows measured thickness values (green line) superimposed on the normative range (green band).
39
Vitreous opacities
Pons et al [69] studied the effect of vitreous
opacities on retardation measurements and showed
that artifacts produced by these irregularities could
falsely increase the value obtained for mean retinal
nerve fiber layer thickness.
Motion artifacts
The possibility that motion of the eye during
scanning could interfere with measurements was
investigated by Colen and Lemij [70]. Motion artifacts led to an increase in retardation and affected
several GDx parameters. This increase was shown to
be highly variable, and caution was recommended
when interpreting images with such artifacts.
Fig. 5. Visual field test of subject with glaucoma (OD).
Visual field shows an inferior hemifield defect. This field
corresponds to the following figure showing a GDx VCC
scan of a glaucomatous subject.
40
Fig. 6. GDx VCC scan of subject with glaucoma (OD). Note the blue shaded area in the superior region of the thickness map
that signifies a relatively thin nerve fiber layer. This thinning of the superior region is highlighted in the deviation map and
the nerve fiber layer graph showing thickness values outside of the norm in the superior region. This defect corresponds well
with this subjects visual field as shown in Fig. 5.
tic (ROC) curve, a measure of the overall discriminating ability of a test that incorporates sensitivity and specificity. The areas under the ROC for
the various GDx VCC parameters ranged from 0.75
to 0.83 versus 0.62 to 0.68 without the corneal
compensation. Similar findings in other studies confirm that the GDx VCC device has a better correlation
with visual function than the original GDx with fixed
compensation [88,89]. Another study showed that
adding individualized compensation to GDx scans
allowed the investigators to obtain retardation maps
that demonstrated defects that matched those ob-
distinguish normal versus glaucomatous eyes. Applying the Fourier-based linear discriminant function
gave a sensitivity of 84% for a given specificity of
92%. This result was a significant improvement over
the parameters given automatically by the device.
41
Limitations
Limited data are available regarding the VCC
system.
Device fundamentals
Optical coherence tomography is a noncontact
noninvasive imaging technology that uses light to
create high-resolution, real-time, cross-sectional tomographic images [95]. OCT is the optical equivalent
Fig. 7. OCT retinal nerve fiber layer thickness of a normal subject. At the top of the figure, a color-coded cross-sectional map
displays 256 adjacent A-scans. The uppermost red layer delineated with the white lines is the nerve fiber layer. Below this image,
a graph shows the patients nerve fiber layer thickness values (black line) in comparison with normative values. The circular
charts show average nerve fiber layer thickness values for the depicted regions, and the colors denote the correspondence with
the normative database.
42
the sample, which ultimately results in a two-dimensional, color-coded map based on detection of the
previously described interference signals. The device
can scan the macula, peripapillary, and ONH regions
(see Fig. 1 and Figs. 7 11). The peripapillary scan is
a circular scan optimally taken with a diameter of
3.4 mm centered at the ONH [96]. The macular
and ONH scans are composed from six linear scans
in a spoke pattern configuration equally spaced
30 degrees apart. The machine automatically identifies the ONH margin as the endings of the retinal
pigment epithelium layer, eliminating the need for
subjective definition of the margin by the operator.
Although it has been recommended that the images
be acquired after pupil dilation, satisfactory scans can
be achieved in most cases without dilation.
Fig. 8. OCT retinal nerve fiber layer thickness of a subject with glaucoma. The cross-sectional map shows diminution of the
nerve fiber layer when compared with the normal thickness. The thinning is pronounced for the most part in the inferior sectors
corresponding to the visual field defect shown in Fig. 1. Abnormal thinning is also evident in the upper sectors, possibly
reflecting structural damage not yet evident by perimetry.
43
Fig. 9. OCT macular map of a normal subject (OS). The uppermost image of the figure is a color-coded, cross-sectional map
along the vertical line, one of the six radial scans that compose the macular map. The central thinning corresponds to the foveola,
and the white lines mark the vitreoretinal and the retinal pigment epithelium boundaries of the retina. The lower left map is the
color-coded macular thickness map wherein blue signifies thinner retina and yellow-green thicker retina. The center map gives
the quantitative measurements in nine sectors.
44
Fig. 10. OCT macular map of a subject with glaucoma (OD). The uppermost image shows a marked diminution of the upper red
band, signifying damage to the nerve fiber layer. The color-coded map (lower left) demonstrates a thinning of the macula, most
extensively in the inferotemporal region, which corresponds to the visual field defect shown in Fig. 1. This finding also can be
appreciated from the numerical values in the lower center map.
45
Fig. 11. (A) OCT analysis of the ONH of a normal subject (OS). The image on the left is a inferonasal to superotemporal optical
cross-section map, one of the six radial scans that compose the ONH map. The edge of the retinal pigment epithelium/
choriocapillaris layer is marked by the blue cross, and a straight line connects between the margins. A parallel line located
anteriorly to this line separates the rim (above the line) and the cup (below the line). On the right, the contours of the ONH (red
circle) and optic cup (green circle) are displayed as created from the data obtained from all six radial scans (blue lines; yellow
line represents the scan depicted in the image on the left). (B) OCT analysis of the ONH of a subject with glaucoma (OD). Note
the widened optic cup and increased slope of the contour of the rim in the image on the left, representing axonal loss in the optic
cup of a glaucomatous patient. The contour map on the right demonstrates a large cup, with thinning of the neuroretinal rim most
pronounced in the inferotemporal region, which corresponds to the visual field defects in this patient (see Fig. 1).
46
Strengths
Provides a cross-sectional view of examined
tissue
Highest axial resolution
Multiple scanning regions
Automatic definition of ONH margin
Limitations
Limited transverse sampling
Reproducibility
Limitations
Pupil dilation required
Highly affected by media opacities
47
Summary
Structural assessment using the imaging technologies discussed herein provides reproducible quantitative measurements of posterior segment ocular
structures. These measurements have been found to
provide useful data for glaucoma detection in various
regions of the posterior segment. Further studies are
needed to evaluate the utility of these technologies
for pre-perimetric glaucoma detection and for monitoring glaucoma progression over an extended period.
References
[1] Leske MC, Heijl A, Hussein M, et al. Factors for
glaucoma progression and the effect of treatment:
the early manifest glaucoma trial. Arch Ophthalmol
2003;121(1):48 56.
[2] Goldberg I. Relationship between intraocular pressure
48
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
49
50
[60]
[61]
[62]
[63]
[64]
[65]
[66]
[67]
[68]
[69]
[70]
[71]
[72]
[73]
[74]
[75]
[76]
[77]
[78]
[79]
[80]
[81]
[82]
[83]
[84]
[85]
[86]
[87]
[88]
[89]
[90]
[91]
[92]
[93]
[94]
[95]
[96]
[97]
[98]
[99]
[100]
[101]
[102]
[103]
51
52
Department of Neurobiology, The Weizmann Institute of Science, The Grodetsky Center for Research of Higher Brain Functions,
Rehovot, 76100, Israel
b
Department of Neurobiology, The Max Plank Institute, Am Klopferspitz 18, 82152 Muenchen-Martinsried, Germany
c
Optical Imaging, 2 Bergman Street, Rehovot, 76705, Israel
d
Department of Ophthalmology, Kaplan Hospital, Rehovot, 76100, Israel
e
Koret School of Veterinary Medicine, The Hebrew University of Jerusalem, PO Box 12, Rehovot, 76100, Israel
ischemia. No currently existing technology can directly image oximetric parameters in the retina [1] in
a quantitative rapid manner [19]. There are no means
for the direct assessment of oxygen, the key parameter of tissue viability. No technology has been
able to demonstrate directly the metabolic functional
status of components of the retinal tissue. Once discovered, functional or metabolic signals will provide
necessary information on the viability and vitality of
diseased tissues using direct methods rather than
indirect methods.
Functional optical imaging, a noninvasive technology widely used in brain research, has recently been
adapted for functional retinal imaging. The authors
believe that functional parameters are more promising
for providing the earliest possible diagnostics rather
than anatomic parameters. This article briefly reviews
functional optical imaging in the neocortex and
describes the type of results obtained by applying
this technology to functional imaging of the eye.
Preliminary results have been published in abstract
form [20 24].
0896-1549/04/$ see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.ohc.2003.12.003
54
obtained under different neuronal processing conditions, differences that previously were thought to be
invisible. Unprecedented accurate functional mapping
of the functional architecture of the neocortex can be
performed. Using high-resolution digital imaging of
exposed cortex and removing extensive biologic
noise, it has been possible to create information-rich
images that reveal minute changes in the light intensity
reflected from the cortex, changes indicating intrinsic
metabolic activity of activated cortical domains (eg,
changes in oxygen consumption, blood flow, blood
volume, and light scattering). The imaging setup is
depicted in Fig. 1. Imaging begins with a sequence of
digital images of the target organ, taken with a video or
digital camera, just after the cortex has been activated
with a given sensory input. Subsequently, functional
maps are obtained by analyzing the small differences between the series of images owing to the
increased oxygen consumption, blood flow, and activity-dependent light-scattering signals. This stage is
followed by a biologic system specific noise reduction procedure that enables the extraction of small but
relevant signals from large artifacts (some are irrelevant physiological events that are removed), image
enhancement, and quantification.
Functional optical imaging based on intrinsic signals
Neocortical functional optical imaging based on
intrinsic signals has provided a new level of understanding of the principles underlying cortical development, organization, and function [28]. In the best
cases, it provides a spatial resolution of 20 mm for
mapping the layout of cortical columns in vivo. This
excellent resolution is much better than that required
to resolve cortical columns. The discussion herein
provides a brief overview of the types of applications
that have been pursued, and the general implications
of some findings for other neuroimaging techniques
based on hemodynamic responses, particularly functional magnetic resonance imaging (f-MRI).
In the mammalian brain, cells that perform a given
function or that share common functional properties
are often grouped together in cortical columns running
from pia to white mater [29 31]. Examples include
the orientation and ocular dominance columns of the
primary visual cortex. Without knowing the function
performed by each group of neurons, it is difficult to
discover the principles underlying the neural code and
its implementation. Attaining an understanding of the
functional organization of a given cortical area is a key
step toward revealing the fundamental mechanisms of
information processing therein. Of special importance
are experimental methods that allow visualization of
the functional organization of cortical columns, particularly methods providing high spatial and temporal
resolution. To date, single- or multi-unit extracellular
recording techniques have provided the best tools for
studying the functional response properties of single
cortical neurons and their synaptic interactions. Although multi-electrode techniques are more promising, the size and placement of these electrode arrays
pose severe problems. Several imaging techniques
have been developed that yield information about
the spatial distribution of active neurons in the brain.
Each of these techniques has significant advantages as
well as limitations. Although the 2-deoxyglucose
method permits postmortem visualization of active
brain areas, or even of single cells, it is a one shot
approach. Only a single stimulus condition in a single
animal can be assayed (the two-isotope method permits the mapping of activity resulting from two
stimulus conditions). Positron-emission tomography
and f-MRI offer spectacular three-dimensional localization of active regions in the functioning human
brain, but both methods have low spatial resolution,
two to three orders of magnitude worse than optical
imaging. Other imaging techniques have also been
applied in vivo with success but suffer from a limited
spatial resolution, a limited temporal resolution, or
both. Among these methods are radioactive imaging
of changes in blood flow, electroencephalography,
magnetoencephalography, and thermal imaging.
Overview of functional optical imaging in the
neocortex
Currently, the best method for imaging the functional architecture of the cortex is based on slow
changes in the intrinsic optical properties of active
brain tissue. This method permits visualization of
active cortical regions at a spatial resolution better
than 50 mm and is immune to some of the problems
associated with the use of optical imaging based on
extrinsic probes [32 35]. The sources for these activity-dependent intrinsic signals include changes in
the physical properties of the tissue affecting light
scattering, and changes in the absorption, fluorescence, or other optical properties of intrinsic molecules. The fact that, in many tissues, small intrinsic
optical changes are associated with metabolic activity
has been known since the pioneering experiments of
Kelin [36] and Millikan [37] on the absorption of
cytochromes and hemoglobin, respectively. The first
optical recording of neuronal activity was made almost 50 years ago by Hill and Keynes [38], who
detected light-scattering changes in active nerves.
Changes in the absorption or fluorescence of intrin-
55
Fig. 1. Optical imaging of functional maps in vivo. (A) The setup. Digital CCD images are taken of the exposed cortex of the
animal, which is sealed in an oil-filled chamber. The cortex is illuminated with light of 605 nm wavelength. During image
acquisition, the animal is visually stimulated with moving gratings projected onto a frosted glass screen by a video projector. The
acquired images are digitized in a camera controller, which transfers the data to the computer controlling the entire experiment.
Functional maps are subsequently analyzed and are displayed on a color monitor. To determine the quality of the maps during the
imaging sessions, the data can be sent to a second computer for detailed, quasi on-line analysis. (Data from Tso DY, Frostig RD,
Lieke E, Grinvald A. Functional organization of primate visual cortex revealed by high resolution optical imaging. Science
1990;249:417 20.) (B) Imaging functional architecture. An activity map for one orientation is obtained straightforwardly by
dividing the image captured during presentation of this orientation by the average of the images captured during presentation of
all orientations. The image of the cortical surface illuminated with green light to emphasize the vasculature is shown at the top.
Activity maps evoked by visual stimulation with horizontal and vertical gratings are shown at the bottom. Scale bar: 1mm.
(Data from Bonhoeffer T, Grinvald A. Iso-orientation domains in cat visual cortex are arranged in pinwheel-like patterns. Nature
1991;353:429 31.)
56
absorption
Changes in blood volume affecting tissue light
scattering
saturation level
Activity-dependent changes in tissue light scat-
tering
The first component of the intrinsic signal in the
neocortex originates from activity-dependent changes
in the oxygen saturation level of hemoglobin. This
change in oxygenation contains two different components. The first component is an early eventan
increase in the deoxyhemoglobin concentration resulting from elevated oxygen consumption of the neurons
owing to their metabolic activity. This component
causes a darkening of the cortex owing to increased
absorption by deoxyhemoglobin molecules and is
referred to as the initial dip. The second component
originates from an activity-dependent increase in
cortical blood volume, increasing the amount of light
absorbed by activated cortical regions in a wavelength-dependent fashion following the absorption
spectra of hemoglobin. The level of spatial regulation
of this component has been of great interest, because
most f-MRI studies use this component of the hemodynamic response. In previous studies of optical
imaging and imaging spectroscopy in the visual cortex
of cats and monkeys, the authors reported a much
weaker colocalization of cortical blood volume
changes with the electrically active columns than the
[Hbr] increase during the initial dip [27,46 49]. This
poorer colocalization is revealed by a lower signal-tonoise ratio in spatial maps of activation calculated
from the cortical blood volume component when
compared with those from the initial dip. This observation is especially true for maps in which the response to a single stimulus condition is normalized by
the response to a blank (single-condition maps). For
differential imaging, in which the response to a
stimulus condition is normalized by or subtracted
from the response to an orthogonal stimulus condition, the maps from the cortical blood volume are still
somewhat poorer than from the initial dip; however,
high-quality functional cortical blood volume maps
have been obtained in cat and monkey visual cortex
with differential imaging [27]. Such differential procedure removes much of the cortical blood volume
noise, because it nearly eliminates the contribution of
the nonstimulus-specific part of the vascular response.
These results indicate that part of the cortical blood
volume changes colocalize with areas of increased
neuronal activity.
The third component is a delayed eventan activity-related increase in blood flow causing a decrease in the deoxyhemoglobin concentration. The
blood rushing into the activated tissue contains higher
57
Fig. 2. Functional optical imaging has been applied to multiple cortical and noncortical areas in multiple species and multiple
sensory modalities. Such maps are depicted on the cover of several journals.
58
Fig. 3. Relationships among pinwheels, ocular dominance columns, and blobs in primary visual cortex of the macaque monkey.
(A) The optical map of ocular dominance. The dark bands represent columns dominated by input from the right eye. Scale bar:
500 mm. (B) The borders of the ocular dominance columns (black lines taken from A) were overlaid onto the discrete angle
map for orientation preference. The pinwheel centers were marked with circles. (C) The continuous angle map of orientation
preference. Same unsmoothed digital data shown in B are depicted using the color scale shown above the map. The oriented
color bars at the right indicate the color code. (D) The CO-rich blobs were marked on the histologic photograph. (E) Angle map
from B and two overlays showing the relationships among blobs, iso-orientation domains, and ocular dominance columns.
( F) Two adjacent fundamental modules, magnified from E. (G) The revised ice cube model illustrating the relationships found in
the above maps. Black lines mark the borders between columns of neurons that receive signals from different eyes. White ovals
represent groups of neurons responsible for color perception (blobs). The pinwheels are formed by neurons involved in the
perception of shape, with each color marking a column of neurons responding selectively to a particular orientation in space.
Note that both the blobs and the centers of the pinwheels lie at the center of the R or L columns. The iso-orientation lines
(appearing as a border between two colors) tend to cross borders of ocular dominance columns (black lines) at right angles. The
slice above the ice cube model depicts two adjacent fundamental modules (400 mm 800 mm). Each module contains a complete
set of about 60,000 neurons, processing all three features of orientation, depth, and color. (Data from Bartfeld E, Grinvald A.
Relationships between orientation preference pinwheels, cytochrome oxidase blobs and ocular dominance columns in primate
striate cortex. Proc Natl Acad Sci USA 1992;89:11905 9.)
59
60
back flow movies. An animation component integrated in the browser allows viewing the results of a
given imaging session with an interactive frame rate
and color table control. It also applies low-pass,
high-pass, and clipping operations to each image as
it is read.
Visualizing retinal blood flow
Invisible information is hidden in a red-free series
of images of the retinal vasculature, taken with the
stroboscopic illumination. Information on how blood
is flowing through the retinas veins, arteries, and
capillaries can be displayed as a movie that directly
shows individual clusters of red blood cells in motion. Fig. 4C shows one flow differential image taken
from the eye of a healthy human subject. The movies
can be viewed in a continuous loop by accessing
http://www.weizmann.grinvald.retinalflow. Fig. 4A
shows a red-free (green illuminated) digital image
of normal human macula. This image is one of a
series of six pictures obtained with the RFI over a
period of 100 ms. The overall retinal area imaged was
4.2 mm on a side (about 15 degrees). The yellow
square shows a large area of interest, 1140 mm on a
side (area shown in Figs. 4B and C). The red square
shows a region of interest 264 mm on a side. Fig. 4C
shows part of the corresponding flow image (yellow
in Fig. 4A), extracted using the RFIs analysis software. Interpretation of this flow image is straightforward. A black spot represents a blood cell or cluster
of red blood cells. In white spots, or gaps, red blood
cells are absent. The courses of several in-focus
capillaries are clearly traced out by trains of black
and white spots, which are elongated by constriction.
Larger blood vessels also have clear gaps and clusters, but these areas are more punctate. In the spaces
between in-focus vessels, the grainy texture of the
image reflects the distribution of red blood cells
moving through deeper, out-of-focus blood vessels.
Quantifying retinal blood motion
The ability of the RFI to visualize individual retinal red blood cells or red blood cell clusters directly,
without the injection of dye, opens up many new
possibilities in the investigation of retinal blood flow,
particularly in small arteries and veins and in capillaries. The flow-enhanced movies of the RFI clearly
reveal the motion of individual clusters of red blood
cellsor even single red blood cellsin vessels as
small as capillaries without the confounding consequences of intermediate assumptions and complicated
transformations. These movies provide a tool for the
61
Fig. 4. RFI direct visualization of blood flow in the human retina. (A) Red-free image of the fundus of a normal subject. Yellow
highlight shows the region magnified in panels B and C. (B) Red-free image at higher magnification. (C) Single frame of a flow
movie. Small red square shows region of additional magnification, illustrated in panel D. (D) Top row shows standard and
annotated (bottom row) flow images from the subregion corresponding to the red square in A, B, and C, obtained at 20-ms
intervals. The top row shows extracts from the raw flow image. Dark clusters of red blood cells appear at lower left, move
rightward, around a sharp curve, and then upward, and exit the frame at the right top. Some motion in other capillaries can also
be seen. In the bottom row, the same images have been annotated for emphasis. The main capillary has been traced in black, and
five gaps between red blood cell clusters have been assigned five colors to emphasize the motion of blood from one frame to the
next. Clusters (black) and gaps (white) move through the capillary shown here at an average rate of 1.6 mm/s.
Fig. 5. Direct automatic determination of flow rate map in the cat retina. (A) Flow rate map calculated from 200 frames, such as
the one shown in the right panel (B), were collected at 100 Hz using a continuous stroboscopic illumination. The velocity at each
pixel was calculated directly by the program. See text for more details.
62
Fig. 6. Semi-automatic quantification of flow in human retina. (A) Red-free image of a normal subject. This image shows several
arteries (red set of colors) and veins (violet set) that were selected for quantification. Each shade of color corresponds to a certain
velocity range. (B) Samples of the calculated velocity, as discussed in the text. Scale bar: 1mm.
Fig. 7. Enhanced visualization of hardly visible vessels. (A) Expanded view of the yellowed region shown in Fig. 4B. Some fine
venules and arterioles are visible, but very little of the capillary bed itself. (B) Same region flow enhanced. Regions with the
strongest flow signal are black, whereas weaker flow regions are white. The flow through larger vessels, although distinct, is less
regimented than the constricted flow through capillaries and appears in gray. (C,D) Similar results obtained from a larger area.
Scale bar: 1mm.
63
Automatic quantification
In tapetal animals, retinal reflectance is high
enough to permit retinal flow imaging with incandescent light or low-power flashes (continuous trains for
hours), permitting high-speed imaging of hundreds of
images per series or even live viewing of the blood
flow on a computer monitor. One can apply a fully
automatic quantification algorithm to such large frame
numbers. This algorithm looks for correlations among
image pixels at different time offsets. The points with
the closest correlation are matched, and the flow rate
is calculated from this pairing. Fig. 5 shows an
example from the cat retina.
Supervised quantification
By using the constraining paths given by the image
of the blood vessels, one can greatly improve the
signal-to-noise ratio of the time offset correlation
matching technique. Automatic measurement of flow
rates can be based on as few as six images. The user
loosely traces over the extent of vessels to be ana-
64
Fig. 9. Qualitative oximetric retinal maps for detection of ischemia in diabetic patients. Each of the three images includes a
comparison of the color image of the retina, the corresponding fluorescein angiogram, and the oximetric maps obtained by the
RFI. (A) In these three panels, it is evident that areas of leakage seen on the fluorescein angiogram are indeed ischemic. (B) The
darker color of the regions near the fovea suggests that the macula is somewhat anoxic. (C) Another example wherein the redfree and fluorescein angiogram clear sign for anoxia are not seen in another diabetic patient. Note that in B and C, some of the
abnormalities detected by the RFI are hardly seen or difficult to note in the corresponding angiographic images. These latter
results remain to be validated by direct oxygen tension measurements.
Fig. 10. Time course image series of the functional signal from cat retina. Near-infrared local changes (right) in retinal reflection
were evoked by local photic stimulation in the visible wavelength range. While a 1-s, local, 100-Hz train of green flashes was
delivered to the cats eye, video images of the feline retina were taken in the infrared at 400 ms/frame. Time increases from right
to left, row-wise from top to bottom. Stimulus duration is marked as a black bar. The control time series without phonic
stimulation is shown at the left. The time course of the signal and its wavelength dependency indicate that it is not the wellknown hemodynamic response. The origin of this metabolic signal remained to be explored. Each image width is 8mm.
Summary
Much work remains to be done to establish the
clinical usefulness of the RFI for early diagnosis and
treatment guidance. The discoveries obtained by functional optical imaging of the neocortex in the last
15 years and the recent RFI studies of the eyes of
normal subjects and patients with diabetic retinopathy, glaucoma, and age-related macular degeneration
suggest that functional optical imaging of the retina
is likely to become a multi-modality powerful clinical tool.
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
Acknowledgments
The authors thank their previous coworkers, R.
Frostig, E. Lieke, C.D. Gilbert, T. Wiesel, D. Tso, R.
Malach, D. Malonek, and A. Shmuel, for their contributions to the functional mapping of the neocortex,
and M. Belkin for his contribution to research with
the RFI.
[15]
References
[17]
[16]
65
66
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
[54]
[55]
[56]
[57]
[58]
[59]
[60]
[61]
mechanisms underlying the intrinsic signal and stability of cortical maps in V1 OF behaving monkeys.
J Neurosci 2000;20:8111 21.
Cohen LB, Keynes RD, Hille B. Light scattering and
birefringence changes during nerve activity. Nature
1968;218:438 41.
Grinvald A, Manker A, Segal M. Visualization of the
spread of electrical activity in rat hippocampal slices
by voltage-sensitive optical probes. J Physiol 1982;
333:269 91.
Bonhoeffer T, Grinvald A. Iso-orientation domains in
cat visual cortex are arranged in pinwheel-like patterns.
Nature 1991;353:429 31.
Malonek D, Tootell RBH, Grinvald A. Optical imaging
reveals the functional architecture of neurons processing shape and motion in owl monkey area MT. Proc R
Soc Lond 1994;258:109 19.
Shmuel A, Grinvald A. Functional organization for
direction of motion and its relation to orientation selectivity in cat area 18. J Neurosci 1996;16:6945 64.
Weliky M, Bosking WH, Fitzpatrick D. A systematic
map of direction preference in primary visual cortex.
Nature 1996;379:725 8.
Shoham D, Grinvald A. Visualization of the hand representation in macaque and human area S-I using
intrinsic signal optical imaging. J Neurosci 2001;21:
6820 35.
Ghose GM, Tso DY. Form processing modules in primate area V4. J Neurosci 1997;77(4):2191 6.
Wang G, Tanaka K, Tanifuji M. Optical imaging of
functional organization in the monkey inferotemporal
cortex. Science 1996;272:1665 8.
Bosking WH, Zhang Y, Schofield B, Fitzpatrick D.
Orientation selectivity and the arrangement of horizontal connections in tree shrew striate cortex. J Neurosci
1997;17:2112 27.
Basole A, White LE, Fitzpatrick D. Mapping multiple
features in the population response of visual cortex.
Nature 2003;424:986 90.
Masino SA, Frostig RD. Quantitative long-term imag-
[62]
[63]
[64]
[65]
[66]
[67]
[68]
[69]
[70]
[71]
67
Traditional electrophysiologic tests of vision involve stimulation of relatively large areas of the retina [1]. For the standard electroretinogram (ERG)
and flash visual evoked potential (VEP) tests, the
entire retina is illuminated, whereas for the pattern
ERG and VEP tests, the stimulus typically exceeds
15 degrees in diameter. The size of the stimuli used for
these tests presents a problem if the clinician is
interested in the local topography of the damage to
the retina or optic nerve. Although ERG and VEP
responses can be elicited with relatively small stimuli,
each retinal area must be tested separately. The time
required to obtain multiple responses to construct a
topographical map would be prohibitive. The multifocal technique was devised by Sutter to solve this
problem [2]. Using the multifocal ERG (mfERG) and
VEP (mfVEP) techniques, local responses are
recorded simultaneously from many regions of the
visual field. This article provides an introduction to
these techniques.
0896-1549/04/$ see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/S0896-1549(03)00101-9
70
Fig. 1. (A) The mfERG display. (B) The mfERG display as it might appear at any moment in time. (C) Examples of the
sequence of events at two locations. (D) A schematic of the continuous ERG signal. (E) mfERG responses, which are the
first-order correlations between the ERG signal (D) and the light sequence (C) of each hexagon. (Modified from Hood DC.
Assessing retinal function with the multifocal technique. Prog Retin Eye Res 2000;19:607 46; with permission).
The recording
Using the same electrodes and amplifiers employed for standard full-field ERG recording, a single
continuous ERG record is obtained (Fig. 1D), usually with the pupil dilated. The continuous record
is typically 4 to 8 minutes in length and is obtained
in 15- to 30-second segments for the subjects
comfort. The ISCEV guidelines [4] recommend a
low-frequency amplifier cutoff of 300 Hz and a
high-frequency cutoff of 3 or 10 Hz. The 10-Hz setting will produce changes in the waveforms but is
used more typically than the 3-Hz setting, because it
produces more stable recordings.
Deriving and presenting the multifocal
electroretinogram responses
The mfERG responses for stimulation of a control
subjects right eye are shown in Fig. 1E. These
responses are derived from the single continuous
ERG record (Fig. 1D). Technically, the mfERG responses are derived as the first-order kernels of the
cross correlation between the stimulation sequence
and the continuously recorded ERG [2,3,5 7].
The responses in Fig. 1E are positioned so that
they do not overlap; therefore, the scaling is not linear
71
Fig. 2. (A) The rings used for the analysis in panels B and C. (B) The mfERG responses summed by rings and expressed in mV.
(C) The mfERG responses in panel B divided by the area within each ring and expressed in nV/deg2. (Modified from Hood D,
Odel JG, Chen CS, Winn BJ. The multifocal electroretinogram (ERG): applications and limitations in neuro-ophthalmology.
J Neuroophthalmol 2003;23:225 35.).
72
Fig. 3. A model illustrating contributions of the different retinal cell types to the mfERG waveform. (Modified from Hood DC,
Frishman LJ, Saszik S, Viswanathan S. Retinal origins of the primate multifocal ERG: implications for the human response.
Invest Ophthalmol Vis Sci 2002;43:1673 85; with permission.)
73
Fig. 4. The full-field ERG records from a standard clinical protocol (left column) and the mfERG records (right column) are
shown for a patient with retinitis pigmentosa. (From Hood DC. Assessing retinal function with the multifocal technique. Prog
Retin Eye Res 2000;19:607 46; with permission).
74
Fig. 5. A schematic illustrating how the second-order kernel is derived. See text for details. (Modified from Hood D, Odel JG,
Chen CS, Winn BJ. The multifocal electroretinogram (ERG): applications and limitations in neuro-ophthalmology. J Neuroophthalmol 2003;23:225 35.)
after the flash differs from the response after the blank
frame, there is a second-order kernel and this kernel is
equal to the difference between the responses [3,5].
The second-order kernel is a measure of how the
mfERG response is affected by a preceding flash. It
is not a response of a particular subset of retinal cells.
The integrity of the outer and inner retina is important for a normal second-order kernel. In monkeys, blocking action potentials generated by the
ganglion cells and amacrine cells markedly reduce
but do not eliminate the second-order kernels [11,12].
In humans, damage to the ganglion cells does not
necessarily change the second-order kernels [14];
however, outer plexiform damage can eliminate the
second-order kernels in patients with degenerative
diseases of the receptors [3,16]. Consequently, a
diminished second-order kernel does not imply damage to a particular set of cells.
Clinical uses of the multifocal electroretinogram
Although the mfERG is used in clinics worldwide, many patients who could benefit from mfERG
are not tested. Other patients are tested using the
mfERG when full-field ERG tests would be as, if
not more, informative. The mfERG test is best reserved for situations in which the full-field ERG is
normal or is likely to be normal or uninformative. The
examples described in this section and elsewhere [3,7]
illustrate such cases.
Fig. 6. (A) The OS (left panel) and OD (right panel) 24-2 HVFs of a patient with paracentral scotomas. (B) The OS (left panel)
and OD (right panel) mfERG responses for this patient. The vertical and horizontal calibration bars indicate 100 nV and 60 ms,
respectively. Bold dark gray, thin black, and dashed light gray circles indicate radii of 5, 15, and 25 degrees, respectively. (C) The
OS (left panel) and OD (right panel) mfERG responses, expressed as response density, for the area within 5 degrees (dark gray),
between 5 and 15 degrees (black), and between 15 and 25 degrees (light gray). (Modified from Hood D, Odel JG, Chen CS, Winn
BJ. The multifocal electroretinogram (ERG): applications and limitations in neuro-ophthalmology. J Neuroophthalmol 2003;
23:225 35.)
75
76
Fig. 7. (A) The 30-2 HVF of a patient with retinitis pigmentosa. (B) The mfERG responses for this patient. (C) The mfERG
responses, expressed as response density (C) and as summed responses (D) for the area within 5 degrees (dark gray), between 5
and 15 degrees (black), and between 15 and 25 degrees (light gray).
77
Fig. 8. (A) The 10-2 HVF of a patient after surgery to repair a macular hole. (B) The mfERG responses for this patient. (C) A
three-dimensional representation of the mfERG results. (D) The responses from the affected eye (black) and the other eye (gray)
were averaged for regions within and outside of the defect. The dashed curve is the black curve scaled in amplitude.
78
Common mistakes
Two common mistakes are worth mentioning.
First, three-dimensional representations, such as that
shown in Fig. 8C, should never be presented without
the accompanying response array (Fig. 8B). Records
without any signal can produce three-dimensional
plots that have a peak in the center, that is, noise
by itself can appear as if there were a foveal response
present [4,7]. Second, eccentric fixation can mimic a
macular problem [7]. This effect is particularly troublesome in the patient who has a small central loss of
vision that could be caused by optic atrophy or outer
retinal disease. Optic atrophy, coupled with eccentric
fixation, can mimic the mfERG recorded when an
outer retinal defect is present. In cases of central loss,
it is important to have a way of determining where
the patient is looking [7].
Summary
The mfERG is a useful clinical tool [3]. It is particularly valuable for distinguishing damage to the
outer retina (choroidal, receptor, and bipolar damage)
from damage to the ganglion cells and optic nerve
when the full-field ERG is normal. In many cases, the
changes in amplitude and waveform can help identify
the particular level at which the retina is affected.
Technical problems must be overcome to obtain useful mfERG recordings. If the clinician is already
recording high-quality full-field ERGs, recording
high-quality mfERGs should not be difficult. Nevertheless, the difficulties involved in full-field ERG
testing are also encountered in the recording and the
analysis of mfERG, and other difficulties unique to
the multifocal technique arise. In general, mfERG
testing is best done at centers with an electrophysiologist trained in ERG and multifocal technology.
79
80
Fig. 9. (A) The display employed for the mfVEP recordings. (B) The electrode locations employed for the mfVEP recordings.
(C) Array of mfVEP responses obtained from the midline electrode placements (A D in panel B). (D) Array of mfVEP
responses obtained from the recording channels that produce the best responses, as defined by the highest signal-to-noise ratio.
81
Fig. 10. The averaged mfVEP responses for 30 normal control subjects.
82
Fig. 12. (A) The 24-2 HVF total deviation probability plot from the left eye of a patient with glaucoma. (B) The mfVEPs from
the left (red) and right (blue) eyes. The red, blue, and green circles in panels A and B indicate radii of 2.6, 9.8, and 22.2 degrees,
respectively. (C) The probability plots for the mfVEPs in panel B. See text for details. (Modified from Hood DC, Greenstein VC.
The multifocal VEP and ganglion cell damage: applications and limitations for the study of glaucoma. Prog Retin Eye Res
2003;22:201 51; with permission.)
83
Fig. 13. (A) The 24-2 HVF total deviation probability plot from a patient with optic neuritis and multiple sclerosis. (B) The
mfVEPs from the left (gray) and right (black) eyes. (Modified from Hood D, Odel JG, Winn BJ. The multifocal visual evoked
potential (VEP): applications and limitations in neuro-ophthalmology. J Neuroophthalmol 2003;23(3):225 35.)
84
Fig. 14. (A) The 24-2 HVF total deviation probability plots from a patient with optic neuritis and multiple sclerosis. (B) The
mfVEPs from the left (gray) and right (black) eyes are shown.
85
Summary
Multifocal VEP technology is still in its infancy.
Although many aspects of the recording and analysis will be improved with time, the mfVEP can be
used in the clinic today. It can help to identify
functional (nonorganic) problems, demyelinating diseases, glaucomatous damage, and poor field takers.
Given the software and techniques needed to record
and interpret the results, mfVEP testing should be
performed in centers experienced with this technique.
References
[1] Holopigian K, Hood DC. Electrophysiology. Ophthalmol Clin N Am 2002;16:237 51.
[2] Sutter EE, Tran D. The field topography of ERG components in man. I. The photopic luminance response.
Vision Res 1992;32:433 46.
[3] Hood DC. Assessing retinal function with the multifocal technique. Prog Retin Eye Res 2000;19:607 46.
[4] Marmor MF, Hood D, Keating D, Kondo M, Miyake
Y. Guidelines for basic multifocal electroretinography
(mfERG). Doc Ophthalmol 2003;106:105 15.
[5] Sutter E. The interpretation of multifocal binary kernels. Doc Ophthalmol 2000;100:49 75.
[6] Keating D, Parks S, Evans A. Technical aspects of multifocal ERG recording. Doc Ophthalmol 2000;100:
77 98.
[7] Hood D, Odel JG, Chen CS, Winn BJ. The multifocal
electroretinogram (ERG): applications and limitations
in neuro-ophthalmology. J Neuroophthalmol 2003;23:
225 35.
[8] Wu S, Sutter EE. A topographic study of oscillatory
potentials in man. Vis Neurosci 1995;12:1013 25.
[9] Hood DC, Wladis EJ, Shady S, Holopigian K, Li J,
86
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
Viswanathan S, Seiple W, et al. Identifying inner retinal contributions to the human multifocal ERG. Vision
Res 1999;39:2285 91.
Hasegawa S, Takagi M, Usui T, Takada R, Abe H.
Waveform changes of the first-order multifocal electroretinogram in patients with glaucoma. Invest Ophthalmol Vis Sci 2000;41:1597 603.
Fortune B, Johnson CA, Cioffi GA. The topographic
relationship between multifocal electroretinographic
and behavioral perimetric measures of function in
glaucoma. Optom Vis Sci 2001;78:206 14.
Sutter EE, Bearse MA, Stamper R. Monitoring retinal
ganglion cell function with the mfERG: recent advances. In: Vision science and its applications, Optical
Society of America Tech. Dig. Ser. Washington (DC):
Opt Soc Amer; 2001. p. 10 3.
Palmowski AM, Allgayer R, Heinemann-Vemaleken
B. The multifocal ERG in open angle glaucomaa
comparison of high and low contrast recordings in
high- and low-tension open angle glaucoma. Doc Ophthalmol 2000;101:35 49.
Sutter EE, Bearse MA. The retinal topography of
local and lateral gain control mechanisms. In: Vision
science and its applications, Optical Society of America Tech. Dig. Ser. Washington (DC): Opt Soc Amer;
1998. p. 20 3.
Bearse MA, Sutter EE, Stamper R. Detection of glaucomatous dysfunction using a global flash multifocal
electroretinogram (mERG) paradigm. In: Vision science and its applications, Optical Society of America
Tech. Dig. Ser. Washington (DC): Opt. Soc. Amer;
2001. p. 14 7.
Shimada Y, Li Y, Bearse Jr MA, Sutter EE, Fung W.
Assessment of early retinal changes in diabetes using
a new multifocal ERG protocol. Br J Ophthalmol
2001;85:414 9.
Fortune B, Bearse Jr MA, Cioffi GA, et al. Selective
loss of an oscillatory component from temporal retinal
multifocal ERG responses in glaucoma. Invest Ophthalmol Vis Sci 2002;43:2638 47.
Palmowski AM, Allgayer R, Heinemann-Vernaleken
B, et al. Multifocal electroretinogram with a multiflash
stimulation in open-angle glaucoma. Ophthalmic Res
2002;34:83 9.
Hood DC. Objective measurement of visual function
in glaucoma. Cur Opin Ophthalmol 2003;14:78 82.
Baseler HA, Sutter EE, Klein SA, Carney T. The topography of visual evoked response properties across
the visual field. Electroencephalogr Clin Neurophysiol 1994;90:65 81.
Baseler HA, Sutter EE. M and P components of the
VEP and their visual field distribution. Vision Res
1997;37:675 90.
Hood DC, Greenstein VC. The multifocal VEP and
ganglion cell damage: applications and limitations for
the study of glaucoma. Prog Retin Eye Res 2003;22:
201 51.
Klistorner A, Graham SL. Objective perimetry in glaucoma. Ophthalmology 2000;107:2283 99.
87
[54] Hood DC, Zhang X, Greenstein VC, et al. An interocular comparison of the multifocal VEP: a possible
technique for detecting local damage to the optic nerve.
Invest Ophthalmol Vis Sci 2000;41:1580 7.
[55] Fortune B, Hood DC. Conventional pattern-reversal
VEPs are not equivalent to summed multifocal VEPs.
Invest Ophthalmol Vis Sci 2003;44:1364 75.
[56] Zhang X, Hood DC. A principal component analysis of the multifocal pattern reversal VEP. Journal of
Vision 2004, in press.
[57] Graham SL, Klistorner AI, Grigg JR, Billson FA. Objective VEP perimetry in glaucoma: asymmetry analysis to identify early deficits. J Glaucoma 2000;9:10 9.
[58] Goldberg I, Graham SL, Klistorner AI. Multifocal objective perimetry in the detection of glaucomatous
field loss. Am J Ophthalmol 2002;133:29 39.
[59] Hood DC, Greenstein VC, Odel JG, et al. Visual field
defects and multifocal visual evoked potentials: evidence of a linear relationship. Arch Ophthalmol 2002;
120:1672 81.
[60] Hood DC, Zhang X, Winn BJ. Detecting glaucomatous damage with multifocal visual evoked potentials:
how can a monocular test work? J Glaucoma 2003;
12:3 15.
[61] Hood DC, Zhang X, Hong JE, Chen CS. Quantifying
the benefits of additional channels of multifocal VEP
recording. Doc Ophthalmol 2002;104:303 20.
[62] Zhang X, Hood DC, Chen CS, Hong JE. A signal-tonoise analysis of multifocal VEP responses: an objective definition for poor records. Doc Ophthalmol 2002;
104:287 302.
[63] Hood DC, Odel JG, Zhang X. Tracking the recovery of
local optic nerve function after optic neuritis: a multifocal VEP study. Invest Ophthalmol Vis Sci 2000;41:
4032 8.
[64] Kardon RH, Givre SJ, Wall M, Hood D. Comparison
of threshold and multifocal-VEP perimetry in recovered optic neuritis. In: Wall M, Wills RP, editors.
Perimetry update 2000: proceedings of the XVII International Perimetric Society Meeting. New York: Kugler; 2001. p. 19 28.
[65] Hood D, Odel JG, Winn BJ. The multifocal visual
evoked potential (VEP): applications and limitations
in neuro-ophthalmology. J Neuroophthalmol, in press.
[66] Miele DL, Odel JG, Behrens MM, Zhang X, Hood D.
Functional bitemporal quadrantopia and the multifocal
visual evoked potential. J Neuroophthalmol 2000;20:
159 62.
[67] Quigley HA, Addicks EM, Green WR. Optic nerve
damage in human glaucoma. III. Quantitative correlation of nerve fiber loss and visual field defect in glaucoma, ischemic neuropathy, papilledema, and toxic
neuropathy. Arch Ophthalmol 1982;100:135 46.
[68] Quigley HA, Dunkelberger GR, Green WR. Retinal
ganglion cell atrophy correlated with automated
perimetry in human eyes with glaucoma. Am J Ophthalmol 1989;107:453 64.
[69] Quigley HA. Open-angle glaucoma. N Engl J Med
1993;328:1097 106.
88
[70] Klistorner AI, Graham SL, Grigg JR, Billson FA. Multifocal topographic visual evoked potential: improving
objective detection of local visual field defects. Invest
Ophthalmol Vis Sci 1998;39:937 50.
[71] Fortune B, Goh K, Demirel S, et al. Detection of glaucomatous visual field loss using multifocal VEP.
Perimetry Update, in press.
[72] Hasegawa S, Abe H. Mapping of glaucomatous visual
Neuro-ophthalmology has many possible definitions, including the discipline of diagnosing diseases
affecting vision before the development of clearly
identifiable clinical signs. In addition, neuro-ophthalmology includes proving the absence of neurologic
disease, especially in patients with functional loss of
vision. Frequently, it is determined that a patient with
apparent neurogenic loss of vision actually has an
ophthalmic condition, such as occult maculopathy.
Three relatively new modalities are becoming technologically refined enough to be useful in clinical practice. These methods are optical coherence tomography
(OCT), multifocal electroretinography (mfERG),
and multifocal visual evoked potential recording
(mfVEP). Perhaps the most revolutionary modality is
mfVEP, because it has the potential of replacing
conventional visual field testing.
* Corresponding author.
E-mail address: THedges@tufts-nemc.org
(T.R. Hedges III).
0896-1549/04/$ see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/S0896-1549(03)00104-4
90
91
Fig. 2. Normal mfVEP responses (A) from a 45-year-old woman who had a temporal visual field defect in her left eye (B), which
remained when both eyes were tested simultaneously.
Multifocal electroretinography
Multifocal ERG, like mfVEP, permits an evaluation of local electrophysiologic responses across the
human retina. It provides objective data useful for the
early detection of focal retinal disease and the evaluation of treatment. The technique allows the simultaneous and topographical recording of the ERG
activity of hundreds of retinal locations from the posterior pole, and it can separate true variation among
retinal regions from a variability in the responses over
time in a single recording session of 4 to 8 minutes. It
is especially useful in patients who have localized
damage in the outer retina. Conventional ERG remains useful for detecting disorders that cause diffuse
92
Fig. 3. Multifocal VEP (A) from a patient with a well-circumscribed visual field defect (B) from a macular lesion.
Fig. 4. Multifocal VEP (A,B) from a man with bitemporal visual field defects (C,D) from a pituitary adenoma.
93
94
Fig. 5. Multifocal VEPs from a 30 year-old woman who had recovered most of the vision lost from optic neuritis associated with
multiple sclerosis. mfVEP for the left eye is normal (A) but shows increased latencies in the right mfVEP (B). Visual acuities
were 20/15 OD and 20/20 OS, and visual fields were relatively normal OS (C) and showed areas of mild depression OD (D).
95
Fig. 6. Multifocal VEP (A,B) from a 55-year-old woman who had right homonymous hemianopia (C) from apparent perinatal
hypoxic injury to the left occipital cortex since birth associated with hemiband atrophy of the optic nerves.
96
Fig. 7. Multifocal ERG recordings (A) showing generalized depression surrounding a relatively preserved central peak from a
52-year-old woman who had been on hydroxychloroquine for 8 years. She had visual acuities of 20/25 OD and 20/25 OS, normal
screening color plate testing, and pericentral visual field loss on automated perimetry (B) compared with a normal mfERG (C).
97
Fig. 8. Multifocal ERG showing reduced responses in the area corresponding with blind spot enlargement (A) from a 38-year-old
woman. (B) Visual acuity was 20/20 in both eyes, and there was mild haziness around an otherwise normal-appearing optic disk
and retina.
98
Fig. 9. Multifocal ERGs (A,B) from a 13-year-old boy with visual acuity of 20/60 OD and 20/40 OS. There were no fundus
abnormalities except for a dark choroid on fundus fluorescein angiography (C,D). Full-field ERG was normal.
99
Fig. 10. Optical coherence tomogram (A) and fluorescein angiogram (B) from a 34-year-old man with a history of facial
numbness and blotchy vision in the right eye for 1 week. Visual acuity was 20/15 OD and 20/10 OS, color vision normal, there
was no afferent pupillary defect, and visual field testing was unrevealing. OCT showed a discrete area of subretinal fluid. Review
of the fluorescein angiogram showed a subtle area of leakage in the macula.
100
Fig. 11. A 9-year-old boy with neuroblastoma was treated with occipital radiation and cisplatin and lost vision in both eyes
within 1 week. MRI was normal. (A, B) OCT showed subretinal and intraretinal edema. (C, D) Both optic disks showed indistinct margins.
101
Fig. 12. A 30-year-old nurse experienced memory loss and fatigue. MRI showed a craniopharyngioma. Postoperatively, visual
acuities were 20/20 in both eyes, but there were persistent left homonymous visual field defects (A). OCT showed diffuse
thinning of the nerve fiber layer in the contralateral right eye and loss temporally (B) and nasally (C) in the ipsilateral left eye.
102
Fig. 13. A 70-year-old woman had anterior ischemic optic neuropathy in her right eye with optic disk swelling and inferior visual
field loss. The left visual field (A) and acuity were normal. OCT showed nerve fiber layer thickening in the left asymptomatic eye
(B), and re-evaluation of the left optic nerve head showed mild swelling and a peripapillary flame hemorrhage (C). Three weeks
later, she had anterior ischemic optic neuropathy in the left eye with inferior visual field loss (D), more severe nerve fiber layer
swelling (E,F).
field of glaucoma [33]. In patients who have neuroophthalmic problems, OCT demonstration of retinal
nerve fiber layer thinning can be used to detect and
monitor optic atrophy from a variety of conditions,
including optic nerve head drusen [35], optic pits [36],
optic nerve trauma [37], and optic hypoplasia [38].
OCT also can be used to help rule out optic neuropathy when nerve fiber layer measurements are normal.
Optical coherence tomography can be used to
measure retinal nerve fiber layer loss in conditions
affecting the pregeniculate visual pathways and in rare
cases when retrogeniculate lesions have been present
long enough to cause transsynaptic degeneration.
The utility of OCT nerve fiber layer analysis for patients with lesions such as pituitary adenoma remains
to be determined. Currently, visual field testing (and
mfVEP) to detect optic nerve dysfunction before
nerve fiber layer loss occurs remains the most effective way to assist endocrinology and neurosurgical
colleagues in following patients with compressive
optic neuropathy (Fig. 12).
103
Fig. 14. Fundus photographs (A,B) and OCT showing submacular fluid (C,D) from a 30-year-old woman with idiopathic
intracranial hypertension. Visual acuities were 20/50 OD and 20/60 OS.
104
[10]
[11]
Summary
[12]
[13]
[14]
[15]
[16]
References
[1] Sokol S. Visually evoked potentials: theory, technique
and clinical application. Surv Ophthalmol 1976;21:
18 44.
[2] Sutter EE. The fast m-transform: a fast computation of
cross correlations with binary m-sequences. Soc Ind
Appl Math 1991;26:686 94.
[3] Baeseler HA, Sutter EE, Klein SA, Carney T. The topography of visual evoked response properties across
the visual field. Electroencephalogr Clin Neurophysiol
1994;90:65 81.
[4] Hood DC, Greenstein VC. Multifocal VEP and ganglion cell damage: applications and limitations for
the study of glaucoma. Prog Retin Eye Res 2003;22:
201 51.
[5] Graham SL, Klistorner AI, Grigg JR, Billison FA. Objective perimetry in glaucoma: asymmetric analysis to
identify early deficits. J Glaucoma 2000;9:10 9.
[6] Winn BJ, Hood DC, Shin EH, Odel JG. Interpreting the
multifocal visual evoked potential (mfVEP): the influence of preretinal factors. Presented at the North American Neuro-ophthalmology Society Annual Meeting.
Salt Lake City, April, 2003.
[7] Merle DL, Odel JG, Berhens M, Zhang X, Hood DC.
Functional bitemporal quadrantanopia and the multifocal visual evoked potential. J Neuroophthalmol 2000;
20:159 62.
[8] Hood DC, Odel JG, Zhang X. Tracking the recovery of
local optic nerve function after optic neuritis: a multifocal VEP study. Invest Ophthalmol Vis Sci 2000;41:
4032 8.
[9] Hood DC, Zhand X, Greenswtein VC, Kangovi S,
Odel JG, Liebmann JM, et al. An interocular compari-
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
105
[34] Katz BJ, Ward JH, Digre KB, Creel DJ, Mamalis N.
Persistent severe visual loss and electroretinographic
abnormalities after intravenous cisplatin therapy.
J Neuroophthalmol 2003;23:132 5.
[35] Roh S, Noecker RJ, Schuman JS, Hedges TR, Weiter
JJ, Mattox C. Effect of optic nerve head drusen on
nerve fiber layer thickness. Opthalmology 1998;105:
878 85.
[36] Theodossiadis GP, Theodossiadis P. Optical coherence
tomography in optic disc pit maculopathy treated by
macular buckling procedure. Am J Ophthalmol 2001;
132:184 90.
[37] Medeiros FA, Moura FC, Vessani RM, Susanna R.
Axonal loss after traumatic optic neuropathy documented by optical coherence tomography. Am J Ophthalmol 2001;135:406 8.
[38] Unoki K, Ohba N, Hoyt WF. Optical coherence tomography of superior segmental optic hypoplasia. Br J
Ophthalmol 2002;86:910 4.
[39] Hayreh SS. Anterior ischemic optic neuropathy. V. Optic disc edema an early sign. Arch Ophthalmol 1981;99:
1030 40.
[40] Samuels B. Histopathology of papilledema. Am J Ophthalmol 1938;21:1242 8.
[41] Corbett JJ, Jacobson MD, Mauer MD, Thompson HS.
Enlargement of the blind spot caused by papilledema.
Am J Ophthalmol 1998;105:261 5.
[42] HoyeVJ, Berrocal AN, Hedges TR, Amaro-Quireza ML. Optical coherence tomography demonstrates subretinal macular edema from papilledema.
Arch Ophthalmol 2001;119:1287 90.