Electrophysiological and histological mapping of the cortical area of central vision in the dog

MISSING IMAGE

Material Information

Title:
Electrophysiological and histological mapping of the cortical area of central vision in the dog
Physical Description:
ix, 109 leaves : ill. ; 29 cm.
Language:
English
Creator:
Ofri, Ron, 1960-
Publication Date:

Subjects

Subjects / Keywords:
Research   ( mesh )
Brain Mapping -- veterinary   ( mesh )
Visual Cortex -- physiology   ( mesh )
Visual Cortex -- anatomy & histology   ( mesh )
Electrophysiology   ( mesh )
Retina -- physiology   ( mesh )
Glaucoma   ( mesh )
Dogs -- physiology   ( mesh )
Department of Veterinary Medicine thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Veterinary Medicine -- Department of Veterinary Medicine -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1993.
Bibliography:
Bibliography: leaves 95-108.
Statement of Responsibility:
by Ron Ofri.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002296910
oclc - 49808376
notis - ALQ0149
System ID:
AA00009056:00001


This item is only available as the following downloads:


Full Text











ELECTROPHYSIOLOGICAL AND HISTOLOGICAL MAPPING
OF THE CORTICAL AREA OF CENTRAL VISION IN THE DOG



















By

RON OFRI


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY CF FLORIDA


1993














ACKNOWLEDGEMENTS


I would like to express my sincere thanks and

appreciation to everyone who has helped me complete this

project. Special thanks go to Dr. William W. Dawson for his

support and guidance during these past 4 years. Many thanks go

to Dr. Don A. Samuelson for his support and encouragement.

Special thanks go to Dr. Kirk N. Gelatt for his contribution

of the Beagle dogs for this study, and to Drs. Dennis E.

Brooks and Mark B. Sherwood for their advice and support.

I would like to express my sincere gratefulness to Ms.

Kim Foli and Minnie Hawthorne for their help in conducting the

experiments. Thanks go to Dr. R. Reep, Ms. Patricia Lewis,

Heidi Wearne, Aga Agnajelski and Mae Chisholm for their help

with the histological studies. Many thanks go to Dr. A. Webb

for his help with the anesthesia protocol, and to Dr. S.

McGorray for her help with the statistical analysis of the

data. I am also grateful to Mr. M. Freeland for his help with

the electronic equipment and Ms. Joanne Clarke for all of her

help.

Last, but not least, I would like to thank my wife, Iris,

for her support, encouragement and understanding that made it

all possible.














TABLE OF CONTENTS



ACKNOWLEDGEMENTS . . ii

KEY TO ABBREVIATIONS . . v

LIST OF TABLES . vi

LIST OF FIGURES . . vii

ABSTRACT . . viii

CHAPTER 1 INTRODUCTION . ... 1

Location of the Primary Visual Cortex 1
Organization of the Visual Cortex . 4
Incoming Signal . 4
Representation of the Visual Field 4
Area Size and Magnification Factor 8
Projection of the Area Centralis 9
Number of Maps .. . 10
Histology of the Visual Cortex . 12
Architectonics of the Cerebral Cortex .. 12
Distinguishing Features of the Striate and
Parastriate Cortex. . 16
Visually Evoked Potentials . 18
VEP's in the Normal Dog . 18
Flash VEP vs. Pattern VEP . 19
Origin of the Signal . 22
Stimulus Parameters and Their Effect on the
PVEP . . 26
VEP's and Glaucoma . 33

CHAPTER 2 MATERIALS AND METHODS. . 36

Experimental Animals . 36
Surgical Protocol . 37
Electrophysiological Mapping of the Area of Central
Vision . . 40
Electrode Placement . .. 40
The Stimulus . .. 40
Signal Processing . ... 41
Histological Studies of the Area of Central
Vision . . 43


iii








Marking the Location of the Area of Central
Vision . 43
Processing and Staining . .. 44
Photography .. . 46
Potential Problems in the Experimental Design 46
Use of Halothane as a General Anesthetic
Agent . . 46
Determining the Location of the Canine Area
Centralis . 48
Pain to the Animal and its Possible Influence
on the Signal . 49

CHAPTER 3 RESULTS . . 51

Results in a Typical Dog . 51
Results in Normal Beagles . .. 51
Results in Pre-glaucomatous Beagles 59
Results in Greyhounds . .. 63
Comparison of the Three Experimental Groups 66
Normal and Pre-glaucomatous Beagles 66
Beagles and Greyhounds . 67
Comparison of the Three Mapping Techniques 76
Accuracy of the Three Mapping Techniques 76
Linkage Between the Three Mapping Techniques 77
The Canine AMF . . 79

CHAPTER 4 DISCUSSION . . 80

The Cortical Area of Central Vision in the Dog 80
Location of the Area of Central Vision on the
Surface of the Canine Cortex 80
Glaucoma and the Location of the Area of
Central Vision . 80
Comparison of Results in Beagles and
Greyhounds . 84
Histological Location of the Cortical Area of
Central Vision . 85
Comparison to the Cat . 86
The Recorded Signal . 90
Significance of this Study. . 91

REFERENCE LIST . . 95

BIOGRAPHICAL SKETCH . . 109














KEY TO ABBREVIATIONS


AMF--areal magnification factor

ANOVA--analysis of variance

BAEP--brainstem auditory evoked potentials

cpd--cycles per degree

c.v.--cresyl violet

D--diopter

EPSP--excitatory post synaptic potential

FERG--flash electroretinogram

FVEP--flash visual evoked potentials

IAP--interaural plane

IPSP--inhibitory post synaptic potential

LGN--lateral geniculate nucleus

PERG--pattern electroretinogram

POAG--primary open angle glaucoma

PVEP--pattern visual evoked potentials

RMS--root mean square

V I--visual area I

V II--visual area II

V III--visual area III

VEP--Visual Evoked Potentials














LIST OF TABLES


Table Pace

1. Experimental animals ...................................36

2. Results in normal Beagles ............................... 58

3. Results in pre-glaucomatous Beagles .....................60

4. Results in Greyhounds ................................... 65

5. Expected and observed frequencies of accuracy in the
anterior-posterior axis ................................... 77

6. Expected and observed frequencies of accuracy in the
medial-lateral axis ........................................ 77

7. Expected and observed frequencies of shortest distances .78

8. Retinal cell counts and visual function ................. 89














LIST OF FIGURES


Figure Page

1. Representation of the visual hemifield on the cat
cortex .................................................. 7

2. A stereotaxic map of the visual areas in the cat ..... 11

3. VEP recordings in man ............................... 21

4. The canine area centralis ............................. 42

5. Distribution of the signals on the cortical surface .. 53

6. Location of the cortical area of central vision in
normal Beagles ......................................... 54

7. Histological location of the cortical area of central
vision .................................................. 55

8. Location of the cortical area of central vision in
pre-glaucomatous Beagles ................................ 61

9. Location of the cortical area of central vision in
Greyhounds ............................................... 64

10. The distribution of the mean signal RMS amplitudes
in Beagles ........................... ............... 69

11. The distribution of the mean signal latencies in
Beagles ................................................. 71

12. Rate of change in mean signal RMS amplitude and
latency in Beagles ..................................... 74

13. A diagram of the canine cerebral cortex ............. 81

14. A representation of the canine medial occipital lobe
in the dog (transverse section) ......................... 82


vii














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

ELECTROPHYSIOLOGICAL AND HISTOLOGICAL MAPPING
OF THE CORTICAL AREA OF CENTRAL VISION IN THE DOG

By

Ron Ofri

December, 1993


Chairperson: Dr. D. A. Samuelson
Major Department: Veterinary Medicine

The visual cortex has been mapped in more than a dozen

animal species to date. However, most of the species in which

such mapping has been conducted are rodents or nonhuman

primates. With the exception of the cat, and possibly the

sheep, these animals are rarely presented as patients in

veterinary clinics. The visual cortex has yet to be mapped in

the dog, the species most frequently seen by veterinarians.

This gap in our understanding of the canine has hampered

clinicians and surgeons attempting to diagnose and treat

neurological and ophthalmological disorders in the dog. It has

also greatly complicated the work of researchers performing

electrophysiological experiments in the canine visual system,

and severely limited the use of the dog as a model for

ophthalmological and neurological disorders.


viii








The aim of this study was to map the area of central

vision in the canine cortex. After surgically exposing the

cortical surface, the dog was paralyzed and its retinal area

centralis stimulated by a pattern of alternating bars. An

electrode mounted on a stereotaxic device was used to perform

a search grid and record the resulting electrical signals from

the cortical surface. Histological studies were conducted to

confirm the cytoarchitecture and myeloarchitecture of the

recording site, and correlate them to the electrophysiological

findings.

Based on the results of these studies, it was determined

that the primary cortical area of central vision in dogs lies

at the junction of the marginal and endomarginal sulcii. The

stereotaxic coordinates of this region in the Beagle are 11.3

0.4 mm anterior to the interaural line and 8.3 0.1 mm

lateral to the midline. In Greyhounds the respective figures

are 15.6 0.1 and 8.5 0.6. A significant difference in the

location of the cortical area of central vision between the

two breeds was found only along the anterior-posterior axis

(ps0.04). No significant differences in the location of the

cortical area of central vision were found between normal and

glaucomatous Beagles (p>0.53). There was a high degree of

correlation (a=5%) between the electrophysiological and

histological locations of the area of central vision. The

possible implications for comparative electrophysiological and

anatomical research are discussed.













CHAPTER 1
INTRODUCTION


Location of the Primary Visual Cortex


Mapping of the visual cortex in man was first performed

by Holmes (1) in World War I, using wounded British soldiers

who suffered brain injuries. The first animal whose visual

cortex was mapped was the cat, which was studied by Talbot in

1940 (2). Since then, mapping of the visual cortex has been

carried out in the owl monkey (3), squirrel monkey (4),

macaque monkey (5) and bushbaby (6). Nonprimates whose visual

areas have been mapped include the tree shrew (7), rabbit (8),

guinea pig (9), sheep (10), rat (11), mouse (12), hedgehog

(13), marmoset (14), degu (15), golden hamster (16) and grey

squirrel (17).

The early works performed on the visual cortex in man and

cat revealed that the primary visual area, so called because

it is the main recipient of lateral geniculate nucleus (LGN)

input, occupies the region known as Brodmann's area 17 (1, 2).

This area is a histologically defined region in the occipital

lobe, possessing unique cytoarchitectural and

myeloarchitectural characteristics, the boundaries of which

were determined for humans in the early 20th century (18). In

man, this area lies on the medial aspect of the occipital

1








2

lobe, along the banks of the calcarine sulcus, bounded

anteriorly by the cuneate sulcus and posteriorly by the

collateral sulcus (19). In the cat, it occupies the posterior-

medial portion of the cortex, extending from the crown of the

lateral gyrus on the dorsal surface to the superior bank of

the splenial sulcus on the medial surface (20). This area has

also been called the striate cortex, due to the large

concentration of heavily myelinated fibers in layer IV, or V

I (Visual I), a physiological term coined by Woolsey and

Fairman in 1946 (21). It should be noted that a significant

degree of individual variability has been shown in the

location of the striate cortex in man (22). However, modern

magnetic resonance imaging techniques show no zonal variation

in the patterns of the cuneate and collateral sulci of the

medial occipital lobe in man (23). In standard human

ophthalmology texts the cuneate and collateral sulci are still

referred to as the borders of area 17 (19).

A second visual area, known as V II, was discovered by

Talbot in 1942, while working on the cat (24). Since this area

lies as a concentric ring lateral to the striate cortex, it is

also called the parastriate cortex (25). In all mammalian

species studied to date, this region corresponds

histologically to Brodmann's area 18, which is adjacent to

Brodmann's area 17 (26).

Since Talbot's discovery of a second visual area in the

cat, multiple visual areas have been discovered in almost








3

every species studied. In the cat, V III occupies Brodmann's

area 19, and since it is also shaped as an outer concentric

ring surrounding V I and V II, it has also been called the

peristriate cortex (25). Together, these three visual areas

occupy most of the feline occipital lobe (26). Additional

visual areas have been discovered in the Clare-Bishop area of

the feline cortex (27), while in man it has been shown that

visual processing takes place as far away as the temporal and

parietal lobes (28). Together with V II these multiple visual

areas are known collectively as the extrastriate visual cortex

(26).

The existence of these multiple cortical visual areas

reflects the extent of the development of the mammalian visual

system. For example, three cortical visual areas have been

identified in the hedgehog, a primitive insectivore (13). Four

such areas have been located in the mouse (29). In the cat, on

the other hand, a dozen extrastriate visual areas, mostly

lying lateral to V I, have been identified (30). These areas

differ in their size, neural organization, region of the

visual field which they map, and the type of information

processed in them. It should also be noted that the mammalian

sensory and auditory organs also have multiple cortical

representations of their stimuli (31). However, since in all

species studied to date the projection of the fovea or area

centralis lies on the border of areas 17 and 18, this

discussion is limited to the description of V I and V II.












Organization of the Visual Cortex


Despite recent advances, V I and V II still remain the

most thoroughly investigated visual areas. Since the area of

central vision is projected onto their common border, they

constitute the focus of interest of this study. There are

numerous physiological differences in the representations and

processing carried out by these two areas.


Incoming Signal


V I receives its input from the LGN (32, 33). More

specifically, these thalamic afferents synapse in layer IVc of

V I (19). This input consists of the entire contralateral

visual hemifield (20). In some species, including the cat, the

input also contains part of the ipsilateral hemifield near the

vertical meridian (20). On the other hand V II receives most

of its input from V I (32, 34). This input doesn't include the

temporal periphery of the visual field; therefore, a smaller

portion of the visual field is represented in V II (26).


Representation of the Visual Field

The vertical meridian is represented at the lateral

border of V I (20), which in primates lies along the banks of

the lunate sulcus (35). The area of central vision is

represented at the center of the vertical meridian. Medial

movement (i.e., away from the lunate sulcus) on the surface of








5

the striate cortex represents peripheral movement (i.e., away

from the area centralis) in the visual field (20, 26). On the

surface of V I, adjacent loci of the cortex map onto adjacent

loci of the contralateral visual field (36) in a simple,

point-to-point manner. This phenomenon is called a 1st order

transformation (37), and has been found in all species studied

to date.

In non-primates V II lies laterally to V I (it is antero-

lateral to V I in primates), bordering it along the vertical

meridian (38). In lower mammals, V II's organization is a

mirror image of V I (7, 13, 17). Therefore, movement from V I

into V II across the vertical meridian would cause a reversal

in the topographic arrangement of the receptive fields. In

higher mammals, V II also borders V I along the vertical

meridian (Figure 1), but the organization is more complex.

Here, adjacent loci in the visual field project onto

non-adjacent cortical loci (39). The horizontal meridian

serves as the dividing line along which adjacent visual field

loci are split. This organization has been called a 2nd order

transformation (37). A further complication in the mapping of

V II's surface is the existence of regions of topographical

irregularity, which defy the normal point-to-point mapping of

the visual field (26). This phenomenon has been described both

in primates and non-primates (39).










0)* 0 -) 0
Sm r- 4 .
M)C 0 0 C
r. rq 44 .. -I t. o
4*H -* 4)
HO 0 c
o) o 0 p r-
*0 W) -4 > 0 14) 0
,4 ) .4J V 4 )
w< 4 rj toM tn


^* 00 "


-H0 ) to r .
44 4 *44 4
GM 'g- m .4


-, pjO C4
0X* 0 4 M

S0 r-4 4 0 4 0
0-H-H 0 0 0 4 9 0

A *0 k



4)CH4>0 10 *Hr-
*H4J taro IZ.




-4 r-4 He i .J 4 p k
V 4V ) I ,V a)) 0





H 0 0 0 g r-
> > > E O .4 0 L
44 4J 0 9r-4 a)4W 4)
*-ri X0r-4 Q W ro r. tv P





00W-H-Hi E-H, 44
t)oH 0 N H > r. -A


04 4M 44 4 A 0
r- to rO 0 00 4 )V
V *-rq 0 to -4 *) :J
44 P 0 0 0 0 $4 0 U V
4 V S) M V 44 t o



V ) Q ) 4 9) 4 M -r4 0 9 in
9: 0 It O4 4U


a r-i --, ^- o + a- 0V)

0
E4 r*H


0)0
H$I
44 4











7




















,* .







0 o
%. I







ql 0 0





, I ,, o
S^ \ \ \



















SS *























2io










Area Size and Magnification Factor


The magnification factor is a term coined in 1961 (35) to

quantify the disproportionate amount of cortical area devoted

to processing visual information from the area centralis. In

the retina, the area centralis is defined by the increased

density of the ganglion cell and photoreceptor (depending on

the species) population. In the cortex, on the other hand, an

identical cellular organization serves to represent both the

area centralis and the peripheral visual fields (32). The

increased detail with which the visual input from the area

centralis is analyzed is not a result of thicker layers and

increased cell density, as in the retina. Rather, this

increased cortical visual discrimination is a result of the

increased cortical area devoted to representing the area

centralis. In the hedgehog, for example, the surface area of

the striate cortex is 20 mm2, 50% of which is devoted to

representing the central 350 of the contralateral visual

hemifield (13). In the cat, the respective figures are 380

mm2, 50% of which is devoted to the central 200 of the visual

field (20).

The areal magnification factor (AMF) is measured in mm2

of cortical area devoted to each degree2 of the visual field.

When a spherical polar coordinate system is used to describe

the topography of the visual field, the visual direction

relative to the center of gaze is expressed by two angles,

azimuth and elevation (20). Thus the shaded area in Figure la








9

represents a visual field 5,000 degree2 in size (1000 along

the vertical axis times 500 along the horizontal axis).

The AMF for the cortical representation of the feline

area centralis in V I is 3.6 mm2 of cortex/degree2 of visual

field (20). In V II, the magnification is usually much

smaller. In the cat, the AMF for the representation of the

area centralis in V II is 0.75 mm2/degree2 (39), or less than

20% of the AMF value for V I. Since the peripheral 40 of the

visual field are not represented at all in V II, it can be

said that this region is an extreme example of AMF=0

mm2/degree2 (39).

An obvious implication of the difference in AMF values

for V I and V II is that these areas differ in size. If V I

represents more of the visual hemifield at a greater

magnification, it stands to reason that it will be larger than

V II or any other visual area (Figure 1). Indeed, V I is

usually described as a core, and V II as its adjoining belt

(26). Since this belt is narrow, it should be noted that in V

II the vertically oriented meridians are much longer than the

horizontally oriented ones (39).


Projection of the Area Centralis

As stated previously, the representations of the visual

field in V I and V II border each other along the vertical

meridian (Figure 1). The area centralis or fovea lies, by

definition, in the center of the vertical meridian. In other








10

words, the representation of the area centralis is on the line

demarcating the border between V I and V II (38, 39, 40). This

characteristic location has been found in all mammals studied

to date, including the rat (11), hedgehog (13), grey squirrel

(17), tree shrew (7), rabbit (8), domestic cat (20, 38, 39)

and various nonhuman primates (3, 4, 26). In all of these

species, the representation of the area centralis invariably

lies on the vertical meridian, at the lateral border of V I,

next to V II. The fact that this principle holds true in such

distantly related mammalian species, but not, for example, in

the turtle, makes it reasonable to assume that this trait

originates with a common ancestor of all euthrian mammals

(13).

The stereotaxic coordinates of the representation of the

area centralis in the feline cortex are P3-L5, i.e., 3 mm

posterior to the interaural plane (IAP) and 5 mm lateral to

the midline (Figure 2) (40). Anatomically, it is located on

the crown of the lateral gyrus, near the junction of the

lateral and posterior lateral gyri (20).


Number of Maps


As stated earlier, the extent of a species' visual

resolution and evolution is reflected not only in the surface

area of the striate cortex and the AMF, but also in the number

of visually responsive areas of that species (26, 41). In the

cat areas 17, 18 and 19 (which borders area 18 along the




























16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16 18 20
Millimetres


Figure 2.


A stereotaxic map of the visual areas in the cat.
The heavy line (2700-900) is the vertical meridian
of the visual field and separates V I medially from
V II laterally. The 1800 meridian between C and D
separates V II from V III. 00 marks the location of
the area centralis. Heavy lines indicate areas in
which folding of cortex made reliable representation
of the cortical surface impossible.
(A) the bottom of the splenial sulcus. (B) the
medial edge of the left hemisphere. (C) the bottom
of the post-lateral sulcus. (D) the bottom of the
lateral sulcus.
Y axis-the anterior-posterior axis, where IAP
marks the division between the anterior and
posterior planes.
X axis-the medial-lateral axis, where 0 marks the
division between the medial and lateral planes.


From: Bilge M, Bingle A, Seneviratne KN, Whitteridge D. A map
of the visual cortex in the cat. J Physiol (Lond) 1967;
191; 116-118P (ref #40).








12

horizontal meridian) (42) each contain one representation of

the visual hemifield, but the picture is more complicated in

other areas. Areas 20 and 21 in the cat each contain two

representations of the visual field (43). The lateral

suprasylvian region in the cat contains six retinotopically

organized units (27, 30). Based on many behavioral studies in

nonhuman primates, it is assumed that these extra-occipital

areas deal with "higher" visual processing, such as shape and

location discrimination and facial recognition (28, 44, 45).

It is hoped that once the central canine visual cortex has

been accurately localized, future investigation of these

"higher" visual processing areas will be attempted in the dog.


Histology of the Visual Cortex


Architectonics of the Cerebral Cortex


Ever since the ground-breaking studies by Brodmann in

1905 and by C. Vogt and 0. Vogt in 1919, it has been customary

to describe the structure of the cerebral cortex in

architectonic, rather than histological, terms (18, 46). In

other words, local variations in the array, numbers and gross

form of the cells forming the individual cortical layers,

rather than a specialized cell type, characterize the various

cortical regions of the brain (46). Architectonically, there

are two ways to describe the structure of the cerebral cortex:

(1) cytoarchitecture, referring to cell bodies stained by

Nissl stains, and (2) myeloarchitecture, referring to








13

variations in fiber myelination as revealed by myelin stains

(47). Later works used electrophysiological methods to

determine the correlation between Brodmann's architectonic

divisions (18) and the physiological functions of these areas

(2, 21, 24). These methods revealed a remarkable correlation

between the early histological and the more recent physio-

logical maps of the cortex. As stated previously, the

architectonically defined areas 17, 18 and 19 correspond

precisely with the physiologically defined visual areas V I,

V II and V III, respectively.

Cytoarchitecturally, several types of cells can be found

in the cerebral cortex. Stellate, also known as multipolar or

granular, cells function as cortical intraneurons (48).

Morphologically they can be divided into fusiform cells (which

are vertically oriented) and basket cells (which are

horizontally oriented). These cells have either spinous or

non-spinous dendrites, which extend into adjacent layers, as

do their axons (19). Pyramidal cells, on the other hand,

project out of the cortex, to the white matter or subcortex

(19). As the name implies, they are shaped like a pyramid,

with one apical dendrite, several basal dendrites extending

from the lower angles and an axon continuous with its base

(19). Generally speaking, pyramidal cells are typical of motor

regions of the cortex, while granular cells dominate in the

sensory areas (19). Myeloarchitecturally, two types of fibers

can be found in the cerebral cortex: horizontal fibers








14

connecting various regions and layers of the cortex, and

vertical fibers which are mostly thalamic afferent fibers

(19).

In 1905, working with eight mammalian orders and using

cytoarchitectural methods, Brodmann divided the cerebral

cortex into 6 layers, or laminae (18). Ten years later, using

myeloarchitectural methods, Vogt and Vogt arrived at a similar

conclusion (46). Furthermore, Brodmann's and Vogt's divisions

overlapped to a great extent, proving that either method is

adequate to describe the cortical architecture (46). Based on

their findings, the 6 layers can be described as follows (18,

19, 46, 47, 48):

Layer I. This outermost layer has been called the plexi-

form, molecular or zonal layer. Cytoarchitecturally, this is

a cell-poor layer in the adult, though during fetal

development it is populated by superficial granular cells

(49). Only few small stellate cells possessing wide

horizontally spreading axons are to be found in this layer

after birth. Myeloarchitecturally, the layer contains a

relatively small amount of tangential nerve fibers.

Layer II. The external granular layer contains small,

dense pyramidal cells, characterized by apical and basal dend-

rites spreading more or less directly into the adjacent

layers. Few stellate cells may also be found in this layer.

Myeloarchitecturally this layer is the poorest in horizontal








15

fibers, and has therefore also been called the dysfibrous

layer.

Layer III. The external pyramidal layer is populated by

pyramidal cells, characterized by a graded increase in their

size, i.e., medium size cells near layer II and larger cells

near layer IV. The apical dendrites of these cells reach layer

I, while their basal dendrites reach layer IV. Their axons

descend all the way into the white matter.

Myeloarchitecturally layer III shows an increasing amount of

horizontal fibers, corresponding to the increasing gradient of

cells size.

Layer IV. The internal granular layer is populated mainly

by small stellate cells, characterized by a starlike pattern

of spineless dendrites. This layer has rather well demarcated

borders with layers III and V, and its width and cell density

form the main criterion for differentiating between the

different cortical areas. Myeloarchitecturally this layer

contains the outer stripe of Baillarger, also called the stria

of Gennari, formed by the large number of fibers which

terminate and synapse here.

Layer V. Called the ganglionic or inner pyramidal layer.

This layer is populated by large pyramidal cells, the axons of

which descend into the white matter as far down as the spinal

cord. This layer has been sub-divided by Brodmann into layer

Va, which contains larger pyramidal cells, and layer Vb,

containing medium size cells. Ramon and Cajal went as far as








16

counting these subdivisons as two individual layers (50).

Myeloarchitecturally, layer V had been called the interstriate

layer, since it contains less horizontal fibers than the two

adjacent stripes of Baillarger.

Layer VI. This innermost layer has been called the multi-

form or spindle cell layer. As the first name implies, this

layer is populated by an assortment of spindle, pyramidal and

stellate cells which gradually thin out towards the white

matter. Myeloarchitecturally this layer contains an opposing

gradient of an increasing number of horizontal fibers which

eventually fuse with the white matter.


Distinguishing Features of the Striate and Parastriate Cortex


Area 17 in the cat is characterized by an unusually thin

layer II, which is almost indistinguishable from layer III

(48). In humans layer II is essentially similar to that of the

cat, although layer III in humans is populated by a somewhat

greater concentration of pyramidal and stellate cells (19).

However, as stated previously, the characteristic

features of area 17 are to be found mainly in layer IV and the

other deep layers. As is the case with all other sensory

cortical areas, area 17 contains a large number of granular

cells. This is especially true of layer IV, where the cells

are densely packed, and which is wider in area 17 than in any

other cortical region (19). Another unique aspect of area 17

is the prominence of the outer stripe of Baillarger, which in








17

fact is almost visible to the naked eye (19). It should be

noted that in primates an inner stripe of Baillarger, located

throughout layer VI of area 17 in most mammals, is present

only near the border with area 18 (47).

In his original work, Brodmann divided lamina IV of area

17 into three sublayers, the outer two of which contain large

numbers of myelinated fibers (18), and therefore form the true

outer stripe of Baillarger (47). Layers IVa and IVb are

populated by large stellate and fusiform cells, with which

cortical inter-connecting (association) axons synapse (48).

Layer IVc is populated mainly by small stellate cells, and it

is here that thalamic afferents from the LGN synapse (19).

Layer V is thin and sparsely populated by large pyramidal

cells (47). It therefore contrasts sharply with the two

adjacent layers which contain large numbers of cells and

myelinated fibers.

Area 18, which forms a concentric ring lateral to area

17, differs markedly from the striate cortex in both its

myeloarchitecture and cytoarchitecture. Layer III of the

parastriate cortex is populated by large pyramidal cells and

contains many coarse, obliquely oriented fibers (42, 47).

Therefore, as opposed to area 17, this results in a wide and

prominent layer III, quite distinct from layer II (20). In

fact it is so wide that sometimes it is divided into two

sub-layers (48). Layer V is also more prominent in area 18,

containing fewer large cells (48), but possessing a large








18

concentration of radially oriented fibers (20, 39). The fibers

in layer IV of area 18 are also radially oriented, as opposed

to the horizontal and vertical orientation in area 17, thus

making for a prominently myelinated band in layers IV and V

(20, 39). On the other hand, layer VI of area 18 is narrower

and has a more sparse cell population than that of area 17

(20). It also contains fewer fibers than the striate cortex

(39). This, coupled with the fact that layer V displays

prominent myelination, makes for a wider myelinated zone in

area 18, one in which the inner stripe of Baillarger is

virtually indistinguishable from the outer stripe (39, 47,

48).


Visually Evoked Potentials


The visually evoked potential (VEP) is an electrical

signal produced by the visual centers of the cortex as a

result of retinal stimulation.


VEP's in the Normal Dog


In undilated and non-anesthetized dogs, flashes of white

light elicit a signal containing a total of three positive and

two negative peaks: P1, N1, P2, N2 and P3 (51). The first VEP

in the dog can be recorded 2 days after birth (52). Peak

latencies reach maturity at age 11 (for Pl) to 38 (for N1 and

P2) days (53). All peak-to-peak amplitude measurements reach

mature levels 58 days after birth (53). This development








19

roughly parallels the maturation of the canine ERG. A

primitive ERG can be recorded in dogs at age 3 weeks, and by

8 weeks of age the canine ERG is fully developed (54).

In clinically normal, nonsedated dogs, the mean latencies

of these five peaks are 14.3, 29.2, 54.5, 78.0 and 98.1 msec,

respectively (51). Of the five peaks, P2 is usually the most

prominent. Mean amplitudes for P1-N1, N1-P2, P2-N2 and N2-P3

are 7.19, 13.30, 6.13 and 5.88 AV, respectively (51). In an

earlier study, using dogs anesthetized with methoxyflurane,

only P2 was identified, but its latency and amplitude were

similar to those reported above (55).


Flash VEP vs. Pattern VEP


Generally speaking, two types of light stimuli can be

used to elicit visual potentials. Flash VEP (FVEP) is the

result of a high intensity, short duration flash of

unpatterned light. A patterned stimuli, using checkerboards,

gratings, dots or other shapes, can also be used to elicit a

visual response. In this study, a pattern reversal stimulus

was used, in which a field of vertical dark and light gratings

is alternated with its complementary field thus maintaining a

constant luminance and eliciting a pattern VEP (PVEP).

As might be anticipated, there are significant

differences between the waveforms generated by such different

stimuli (Figure 3). The two most significant differences

between the two waveforms are the absence of P3 in PVEP, and








20

the fact that the first peak in PVEP's is negative rather than

positive (56, 57). However, it is difficult to describe a

"typical" PVEP, because the waveform varies greatly with the

luminance, reversal rate, size and location of the stimulated

area, and other parameters. In the rat, the peak-to-peak

amplitudes of the FVEP are 128 jV for P1N1, 190 for N1P2, 158

for P2N2, 140 for N2P3 and 165 jV for P3N3 (56). The values

for the rat PVEP are 27 AV for N1P1, 39 for P1N2, 2.3 for N2P2

and 25 AV for P2N3 (56). The latencies of the 6 FVEP peaks in

the rat are approximately 25, 33, 48, 70, 100 and 165 msec.,

respectively (56). The latencies of the 5 PVEP peaks are

approximately 45, 67, 85, 87 and 112 msec, respectively (56).

Generally speaking, when stimulating the same retinal

areas, the FVEP has a different and wider distribution over

the scalp when compared to the PVEP (59). However, closer

analysis of the various peaks in the rat VEP reveals that the

PI, N2 and N3 peaks of FVEP, and the P1 and N3 peaks of the

PVEP have a smaller surface distribution of maximal amplitude,

indicating, perhaps, a more localized site of origin (56). In

human PVEP the 5 peaks are identified as N75, P100, N100, N150

and P200, based on their mean latencies (60).

It has already been established that flash ERG (FERG) and

pattern ERG (PERG) have different origins in the retina. The

pattern ERG is considered to be an indicator of inner retinal

function, while FERG is regarded as an indicator of outer

retinal function (61). Separate origins have also been

























1


Figure 3.


VEP recordings in man
Top-PVEP recordings elicited by 15' checks, 120 in
diameter.
Bottom-FVEP recordings elicited by a flash stimulus
in the same subject. Luminance and field size were
identical to those of the pattern stimulus. N1 is
prominent in the FVEP signal but reduced in the PVEP
signal. On the other hand, the N2-P2 amplitude is
much larger in the PVEP signal. Vertical line: 10V;
horizontal line: 100msec.


From: Sokol S. Visually evoked potentials: Theory, techniques
and clinical applications. Surv Ophthalmol 1976; 21;
18-42 (ref #58).








22

suggested for FVEP and PVEP. Differences between FVEP and PVEP

in people suffering from optic neuritis have led researchers

to conclude that the 2 types of VEP are transmitted by

completely separate fibers as early as the optic nerve (62).

It has been suggested that PVEP is generated by activity in

the X-type ganglion cells while FVEP is generated by activity

in the Y-type ganglion cells (63). However, despite these

suspected different origins, responses recorded in the LGN to

flash and pattern stimuli showed no significant differences in

their amplitudes and latencies. These differences became

apparent only in the higher visual centers of the brain (64).


Origin of the Signal

The gross VEP

When recording VEP's the signal may be "contaminated"

with artifacts such as muscular and cardiac activity, and

respiration. In the dog it has been reported that the VEP

signal may also be contaminated with far-field ERG activity

(65). However, most of these artifacts can be eliminated by

careful selection of sites for the reference and ground

electrodes, and by restricting the recording to signals

time-locked with the stimulus.

Another possible source of signal contamination is from

the pathways leading from the retina to the visual cortex.

While most of the retinal axons relay their information to the

striate (and in some species to the extrastriate) cortex via

the LGN, some axons take a less direct route. These fibers run








23

from the retina to the superior colliculus, and from there,

via the pulvinar or the lateral posterior nucleus of the

thalamus, to extrastriate visual areas (26). This pathway is

primarily involved in location of objects in the visual field,

shifting of visual attention and control of eye movements

(66). Electrical activity in this secondary pathway may also

interfere with the recording of the VEP (67). Again, this

interference may be partially eliminated by judicious

placement of the electrodes over the occipital lobe.

When the VEP is recorded with surface electrodes placed

over the occipital lobe, a significant part of the recorded

signal is a result of activity of the central retina. There

are two reasons for this phenomenon. The first is the

magnification factor discussed earlier. Because a

disproportionate area of the cortex is devoted to processing

information from the fovea or area centralis, this region

contributes a significant part of the signal (58). The second

reason is the anatomy of area 17. As stated earlier, the area

centralis representation is located at the crown of the

lateral gyrus in the cat (20), or the junction of the dorsal

and tentorial surfaces of the calcarine sulcus in nonhuman

primates (68), i.e., on the exposed dorsal surface of the

occipital lobe. Representations of the more peripheral visual

fields, on the other hand, lie deep within the calcarine

sulcus in nonhuman primates (68) or the splenial sulcus in

cats (20). Therefore, the recording electrodes are physically








24

located directly over the representation of the area

centralis, while representations of the peripheral fields are

far less accessible.

The neuronal basis of VEP's

Visual evoked potentials represent the sum of the

electrical activity of cortical neurons following visual

stimulation. This activity may be the result of synaptic

depolarization causing excitation (EPSP) or hyperpolarization

causing inhibition (IPSP). Furthermore, in the same cell, a

primary EPSP event may be followed by an IPSP, or vice versa

(69). Therefore, some intracellular recordings yield results

very similar to surface VEP recordings, while in other cases

the two recordings produce different signals.

While the generator sites for other evoked potentials,

such as auditory brainstem evoked potentials, have been

precisely located, this has not been the case for VEP's.

Nevertheless, a few general statements can still be made.

Inhibitory post synaptic potentials dominate during the

positive phases of surface recorded VEP's, while EPSP's

prevail during the negative surface waves (70). Both EPSP and

IPSP are tuned to the preferred direction or orientation of

the stimulus (71). In cats the first positive deflection of

the VEP wave reflects the activity of afferent geniculo-

cortical on-center fibers and of about 1/3 of the cortical

neurons in layer IV (69). The following large positive

deflection is a result of a GABA-mediated primary IPSP in a








25

large population of cortical neurons (69). The negative compo-

nents of the VEP wave reflects a late, less synchronized EPSP

of a large proportion of cortical cells (69).

In electrophysiological studies of man and nonhuman

primates, it is customary to identify the VEP peaks based on

their latency rather than the order of their appearance (72),

though the waveforms are sometimes similar to those of non-

primates. It has been reported that when the FVEP of nonhuman

primates and cattle are compared, a homology is found between

P15 and P1, N40 and N1, P65 and P2, and N95 and N2 of the

monkey and cow respectively (73). Similar homologies have been

proposed with the P40, N70, P100, N130 and P170 of the FVEP in

man (72), and with P1, N1, P2, N2 and P3 in the canine FVEP

(74). Based on these homologies, it has been suggested that in

both dogs and cattle P1 originates at a subcortical source,

primarily the thalamocortical radiations (73, 74). N1

represents excitatory postsynaptic potential from thalamo-

cortical fibers onto neurons in lamina IVA and IVCb of the

striate cortex (73, 74). P2 represents inhibitory postsynaptic

potentials from stellate cells onto other neurons in lamina

IVCb of the striate cortex (73, 74). N2 represents stellate

cell input to layer III of the striate cortex (73, 74). P3 is

the result of poststriate activity (73, 74). However, as

mentioned earlier, the question of the location of the

generator sites has not been unequivocally resolved. Some

researchers have suggested that the P2 peak of the feline FVEP








26

is a retinotectal pathway signal transmitted to the

extrastriate cortex (62). Other studies concluded that the

origin of P1 in rats is cortical rather than subcortical (56).


Stimulus Parameters and Their Effect on the PVEP

Visual Refraction

A great effort was made in these experiments to ensure

the high optical quality of the stimulus image. The equipment

design made it possible to visualize the image on the canine

funds and to focus it in order to achieve maximum image

quality. This quality control was necessary because the PVEP

is more sensitive to small refractive errors than the ERG

(75). A one-diopter (D) error can reduce the PVEP amplitude by

30%, and a four-D error can cause the wave to approach zero

amplitude (75). Refraction also affects the latency of the

PVEP, and this effect is proportional to the spatial frequency

of the stimulus (76).

Retinal location and field size

Due to the limited objective of these experiments, the

stimulating pattern was centered on the canine area centralis.

However, the location of the stimulus in the retinal field can

have a considerable effect on the signal recorded, though this

effect is, once again, spatial frequency dependent (58).

Maximal PVEP amplitudes are recorded when narrow gratings (<20

minutes of arc) are used to stimulate the central 2-3 degrees

of the retina in man (77, 78). Using these same patterns to

stimulate parafoveal regions results in a significant








27

reduction of PVEP amplitudes (77, 78). The parafoveal and

peripheral areas of the retina generate larger PVEP signals as

the spatial frequency of the stimulus is decreased, especially

outside the 6-12 degree isopter (78, 79). These phenomena are

probably a result of the distribution of cones and rods in the

retina, and the fact that lower spatial frequency stimuli tend

to produce more of a luminance, rather than pattern, response

(58).

Photoreceptor concentration also explains the finding

that in the more peripheral retina, larger stimulating fields

are needed to evoke the same response that smaller fields

evoke in the central retina (77). In the central retina PVEP

amplitudes increase linearly with the diameter of the

stimulated area, though the use of an annular stimulus can

cause a reduced response, probably due to center-surround

antagonism (80).

Spatial frequency

It is well established that PVEP amplitudes are related

to the spatial frequency of the stimulus (81-87). Many

researchers describe the spatial frequency-PVEP amplitude

function as a tuning function (85, 86). This function displays

a maximum PVEP amplitude with stimuli 10-20 minutes of arc in

size, with a drop in amplitude when the grating size is

increased or decreased (81, 82). More careful analysis of the

data suggests that this tuning function may be bimodal (i.e.,

containing 2 peaks) (85, 86). This result, and the varying








28

reports on the most "effective" spatial frequency, may be due

to variations in electrode positioning, since different

cortical regions, and even individual neurons, may possess

different spatial tuning characteristics (88).

However, when wide gratings (>30 minutes of arc) are used

as stimuli the PVEP signal becomes more luminance- rather

than pattern-, dependent (58), and some researchers suggest

that at these frequencies the PVEP amplitude levels off rather

than decrease (83). These conflicting results are probably due

to the fact that the effect of spatial frequency on PVEP is

also dependent on the temporal frequency and contrast of the

stimulus (84, 88).

In man it has been shown that the P200 peak is most

sensitive to changes in spatial frequency, while the P100 is

hardly affected by these changes (87). In rats, on the other

hand, both the latency and amplitude of P1 are reportedly

affected by the spatial frequency of the stimulus (84).

Stimulus contrast

Changing the contrast of the stimulus has different

effects on the amplitude and latency of the PVEP. At low

contrast levels, the relationship between contrast and PVEP

amplitudes is linear (89, 90), with a shallow slope (88). At

higher contrasts, the function has been described either as

nonmonotonic (89), or a linear relation of the amplitude to

log contrast (90, 91), with a steep slope (88). On the other

hand, the latency of the signal decreases as the stimulus








29

contrast (84, 92) or log contrast (91) is increased. In rats,

the contrast function of the amplitude of the second harmonic

of PVEP also consists of two linear segments: one for low

contrasts with a shallow slope and one for high contrasts with

a steep slope (84). The fact that these two distinct

components have been found in both rats and primates suggests

the existence of 2 separate contrast mechanisms (84).

Stimulus luminance

The effect of luminance on the PVEP is closely related to

the spatial and temporal frequencies of the stimulus. It is

generally agreed that luminance is best perceived at

intermediate spatial frequencies, suggesting the existence of

a bandpass filtering mechanism (93, 94). PVEP recordings in

rhesus monkeys have shown that the perception of gratings

composed of luminance contrast peaks at four cycles per degree

(cpd), and drops at lower and higher spatial frequencies (94).

Similar results were obtained in cats and rats, but it is also

reported that in these species the signal latency increases as

a function of spatial frequency (84, 95).

The effect of the temporal frequency of the stimulus on

luminance perception is more controversial. Some researchers

claim that similar bandpass filtering of PVEP signals takes

place when the temporal frequency of the stimulus is changed,

and that contrast perception peaks at a frequency of 5-15 Hz

(93). Others suggest a lowpass tuning phenomenon in the








30

temporal domain, with contrast perception dropping at

frequencies >3 Hz (94).

Background luminance can also affect the PVEP recording.

In high background luminance peak N100 is dominant, while N150

is barely noticeable (96). As the background luminance is

decreased, the situation is reversed, and N150 becomes the

dominant negative peak (96).

Temporal frequency

As stated previously, the effect of the temporal

frequency of the stimulus on PVEP is also dependent on the

spatial frequency and contrast. When plotting PVEP amplitudes

vs. temporal frequency in nonhuman primates, a sharp tuning

curve is obtained, with the largest signal recorded at a

frequency of approximately 15 Hz (88). However, even at this

"optimal" frequency, the signal will depend on the spatial

frequency (88). On the other hand, the signal's phase lag, or

latency, increases with increasing temporal frequency and

this increase is more pronounced at higher spatial frequencies

(88). It has also been reported that at very low temporal

frequencies (<2 Hz.) the PVEP becomes more transient, rather

than steady-state, in its nature (92). Thus, the PVEP shifts

from a transient, bandpass type of temporal response at low

spatial frequencies, to a sustained, low-pass response at high

spatial frequencies (97). This characteristic has indeed been

found to be similar to the properties of the receptive field

of a single X-type feline ganglion cell (98).










Stimulus wavelength

Electrophysiological recordings can be used to measure

spectral sensitivity in humans viewing alternating bars of

equal brightness but different wavelengths. Pattern VEP

recordings are much more sensitive than ERG recordings at

detecting subtle wavelength differences (99). Both the PVEP

amplitude and latency measurements provide a spectral

sensitivity curve which is in good agreement with the standard

CIE photopic function and with psychophysical studies, with

retinal sensitivity peaking at 570 nm (99).

The effect of chromatic changes in the stimulus on the

PVEP is, once again, dependent on the temporal and spatial

properties of the grating pattern. Chromatic sensitivity is

highest at low spatial and temporal frequencies, suggesting

the existence of a lowpass filtering mechanism (94). The

greatest chromatic sensitivity in man was recorded at a

temporal frequency of 0.2 Hz (97). At lower frequencies,

chromatic perception drops, and disappears at 0.01 Hz (97).

Electrode placement

It is hoped that the results of this work will assist in

the development of a protocol for PVEP recordings in the dog,

similar to that which exists in the cat (60). As previously

noted, one of the most important considerations in carrying

out successful PVEP recordings is the precise location of the

recording and reference electrodes. Displacement of the active

cortical electrode by as little as 5 mm can cause a decrease








32

of up to 25% (depending on the spatial frequency of the

stimulus) in the amplitude of the PVEP in the dogs (Dawson,

personal observations). In humans, the respective figures are

2.5 cm and 75% (100). In the cat, the latency of cortical

responses to flash stimuli was not significantly affected when

different active electrode positions were used, but the

amplitude of the signals changed by as much as 280% (60).

Positioning of the reference electrode may affect the

contamination of the PVEP signal by far-field potentials such

as ERG (55, 65, 96).

Visual evoked potentials and psychophysical testing

Visual evoked potential recordings are frequently highly

correlated to psychophysical test results, and it is this fact

that makes the VEP such as an effective research and

diagnostic tool (58). In humans, PVEP refraction agrees quite

well with Snellen chart refraction (58). The PVEP refraction

technique has been refined so that both sphere and cylinder

corrections can be quickly determined (101). The spatial

tuning curve of the PVEP in rhesus monkeys parallels the

psychophysical sensitivity falloff reported in these nonhuman

primates. Electrophysiologically- and behaviorally-determined

visual resolutions in the dog also produce similar results

(102).

The psychophysical contrast threshold in man is in close

agreement with the threshold of P100 (88-90), which is the

most contrast sensitive component of the human PVEP (92). The








33

spectral sensitivity of man as determined by VEP's and

psychophysical testing is also very similar (58). This

similarity has been documented both with FVEP's (103) and

PVEP's (99). Peak VEP sensitivity occurs at 570 nm under

photopic conditions (99) and at 500 nm under scotopic

conditions (58).

Visual evoked potentials are also very sensitive to

cerebral lesions in both rats (79) and humans (104). In rats,

even small, focal lesions have a significant effect on the

VEP, though the effect varies for FVEP's and PVEP's (79). This

difference in the effect on the two types of VEP's once again

hints at the existence of different generation sites for these

two potentials (79). In man FVEP's and PVEP's are also

differentially affected by cerebral lesions (104). Latency of

the signal is the parameter most altered in FVEP's, while

amplitude of the signal is the parameter most affected in

PVEP's (104). Pattern VEP's recorded using a multichannel

system accurately pinpoint the location of cerebral lesions in

man (104).


VEP's and Glaucoma


Visual evoked potentials are regarded as an especially

useful tool in the early diagnosis of glaucoma, since

frequently they display abnormalities long before other

clinical changes, such as optic disc cupping, become visible

(105-107). Patients suffering from ocular hypertension also








34

have a high percentage of abnormal PVEP recordings, even

though their visual fields are still normal (108-110).

When a flash stimulus is used to elicit FVEP in

glaucomatous patients, signal latency (111) and P2 amplitude

(112) are unchanged. However, the amplitude of P1 is reported

to decrease significantly, and the extent of this decrease is

correlated with the stage of the disease (112). When the flash

frequency is altered, it is revealed that the loss is evident

mostly at high (>13 Hz) flicker rates (110, 113). It has been

suggested that this is because at the early stages of glaucoma

there is a preferential loss of Y-type retinal ganglion cells

which are indeed more responsive to high frequency stimu-

lation (113). However, other researchers report that these

changes in the signal can be reversed if the IOP is lowered,

thereby suggesting the existence of a different mechanism

(111).

Most PVEP parameters, including amplitude, latency,

contrast threshold and/or slope, are also abnormal when a

pattern is used as stimulus (107). These changes are, once

again, dependent on the spatial and temporal frequencies of

the stimulus used. The changes in the latency of the PVEP

signal in glaucomatous patients is most pronounced at low

spatial- and high temporal-frequencies (109). The latency of

the average signal increases by 8 msec, with more pronounced

changes as the disease progresses (109). Loss of contrast

sensitivity (106) and PVEP amplitudes (105) are also most








35

evident at low spatial and high (8 Hz) temporal frequencies.

A 50% drop in PVEP amplitudes can be recorded as early as 2

weeks after glaucoma is experimentally induced in nonhuman

primates (105). By 16 weeks, signals drop to nearly 1/3 their

normal amplitude (105). In people suffering from advanced

glaucoma, the PVEP waveform can become so distorted that

identification of the various peaks and analysis of the signal

is all but impossible (108).

As is the case with the flash VEP, the existence of two

segregated pathways is the probable cause for the spatial and

temporal tuning properties of the PVEP in glaucoma (106). In

both nonhuman primates and cats, Y- (or alpha-) type retinal

ganglion cells are sensitive to stimuli possessing a low

spatial and a high temporal frequency (114, 115). These

neurons, which project to the magnocellular layers of the LGN,

possess large diameter axons which have been shown to be

preferentially damaged in induced glaucoma in nonhuman

primates (116). It is worth noting that while magnocellular

and parvocellular systems have yet to be documented in the

dog, the preferential loss of large diameter axons has been

demonstrated in the optic nerve of chronically glaucomatous

Beagles (117).














CHAPTER 2
MATERIALS AND METHODS


Experimental Animals


A total of 12 dogs were used in the study. The age, sex,

breed, eye disease status and stimulated eye of all dogs used

in this study are summarized in Table 1.



TABLE 1
EXPERIMENTAL ANIMALS

Dog Age Sexa Breedb Statusu Stimulated
designation (in years) eye

Oa 7.6 F B N left
Kd 10.0 F B N right
Jl 5.3 M B N right
Sd 6.0 M B N left
Oe 10.6 F B N right
Bo 7.8 M B N right
Be 4.9 M B N left
Pa 3.1 F B G left
Se 1.8 M B G right
Kf 3.3 F H N left
Br 4.0 M H N left
Bn 4.7 F H N right

,Sex: F--female M--male
bBreed: B--Beagle H--Greyhound
CStatus: N--normal G--Glaucomatous

Nine of the 12 dogs were Beagles. Beagles were chosen

since it is the canine breed most frequently used in research

and stereotaxic mapping of the Beagle's brain is published

(118). Two of the Beagles were pre-glaucomatous dogs from a








37

colony with hereditary primary open angle glaucoma (POAG)

(119). These dogs were used to study the possible effect of

glaucoma on the location of the cortical area of central

vision in the dog, or, if as expected, no effect was found, to

increase the size of the sample. Mapping was also conducted in

three Greyhounds, for inter-breed comparison purposes.

Greyhounds were chosen due to their distinctly different head

size and shape. Recordings were conducted only on normal or

pre-glaucomatous dogs.

It has been reported that VEP measurements in dogs

stabilize by age 100 days (53); all dogs used in this study

were much older. Every dog used in the study underwent a

clinical and ophthalmological examination on the day of the

experiment.


Surgical Protocol


Use of dogs in this study was approved by the

Institutional Animal Care and Use Committee of the University

of Florida as project #0106.

Dogs used in this study were deprived of food the

previous evening, but had unrestricted access to water.

Animals were premedicated with 0.01 mg/kg of glycopyrrolate

I.M., to reduce secretions and stabilize cardiac function. The

animal was then catheterized and injected with 15 mg/kg

thiamylal, I.V. The dog was intubated and placed on halothane

anesthesia, using a mixture of 25% oxygen and 75% nitrous








38

oxide. The initial halothane concentration was 3%, and was

reduced to 1% after the animal reached a surgical plane of

anesthesia.

The animal was placed on a hot water blanket, and a

rectal thermometer was used to monitor body temperature. An

EKG was used to monitor cardiac function, and the eyelids were

sutured to prevent the cornea from drying during the

operation. The dog's head was then mounted in a stereotaxic

device, and local injections of lidocaine were used to relieve

any pain that might result from use of the ear-bars.

A lateral rostrotentorial craniotomy operation was

performed to expose the presumed location of V I in the

occipital lobe (120). A horseshoe-shaped skin incision was

made, extending from the ear to the dorsal midline, and

curving cranially to the lateral canthus of the eye. The

underlying muscles were severed along the nuchal crest, and

reflected ventrolaterally. Compression, ligation and

coagulation were used to minimize bleeding. The muscles were

scraped off the skull using a periosteal elevator. An electric

drill was then used to create the initial hole in the bone.

This hole was then expanded with a Lampert rongeur until an

opening with the following borders was obtained: 0.5 cm

rostral to the nuchal crest and lateral to the dorsal midline;

rostrally, at the orbit and frontal sinus; and ventrally at

the temporal musculature, zygomatic arch and coronoid process

of the mandible (120). Bleeding from exposed sinuses was








39

controlled with vitamin K injections (2 mg/kg, I.M.) given 24

hours pre-op (121), and with gelfoam and bonewax inserted into

the sinuses during surgery. The dura was reflected using the

tent-flap method and cut with iris scissors. The edges of the

wound were treated with local injections of lidocaine, but no

drug came in contact with the exposed cortical surface.

Following the operation, the dog was taken off the

anesthesia, and its muscles relaxed with an injection of

pancuronium (0.1 mg/kg, I.V.). A respirator, delivering the

same oxygen-nitrous oxide mixture, was used to ventilate the

dog. End-tidal CO2 levels were monitored and maintained at a

range of 4.4-4.7%. Additional injections of pancuronium were

administered when needed, approximately one injection every 45

minutes.

Great care was taken to ensure that the cortical tissue

remained viable following its exposure. After reflecting the

dura, a plastic chamber was mounted over the exposed brain and

filled with warm mineral oil to prevent the pia from drying

(17). To prevent brain edema once the dura has been reflected,

dilute mannitol (0.625%) was added to the fluids which the dog

was receiving. This was supplemented by I.V. injections of

dexamethasone and mannitol as needed. No dog received more

than 50 ml of mannitol (25%) and/or 6 mg of dexamethasone

(121).













Electrophysiological Mapping of the Area of Central Vision


Electrode Placement


The contralateral pupil was dilated using tropicamide -and

covered with a contact lens in which a platinum wire has been

embedded. This lens served as an active electrode for the

simultaneous recording of PERG signals; its reference

electrode was placed at the medial canthus (122). A 2 mm

diameter silver ball tipped wire (32, 123) mounted on the

stereotaxic device served as the active electrode for the

unipolar recording (8, 13) of PVEP's. As the area centralis

was being stimulated, this active cortical electrode was

moved, using the control dials of the stereotaxic device, in

a search grid-like manner on the exposed surface of the

cortex. The increments of movement were 2 mm. This distance

was used by Woolsey and Talbot in their mapping studies (8),

and was shown by Hubel and Wiesel to correspond to the size of

one aggregate receptive field in the cortex (32). The cortical

reference electrode was attached to a screw mounted on the

skull just rostrally to the bone lesion. The dog was grounded

through the earbars of the stereotaxic device.


The Stimulus


Using a modified Bausch & Lomb table mounted

ophthalmoscope, a computer generated counterphased image was








41

projected on the funds of the eye. Visual control was used to

carefully position and focus the image on the area centralis

of the retina (Figure 4). The image consisted of square wave

contrast, vertical, dark and light bars. The stimulated

retinal field was 150 in diameter, based on a mean retinal

axial length of 18 mm (124) and a mean optic disc diameter of

1.8 mm (125). The stimulated field was surrounded by a

luminated field 450 X 750 in size.

The width of an individual bar was 16 minutes of

arc/phase. Bars were alternated in a square-wave fashion 0.5

cycles (1 bar width) at 2.5 Hz. The stimulus had a mean

luminance of 1.4 log Fl, with 70% contrast of the bars. The

luminance of the surround field was 0.3 log units higher than

the luminance of the light bar of the dark/light bar pair, to

suppress the effect of any scattered light. The room was

darkened.


Signal Processing


The analogue/digital converter used to record the VEP's

was time-locked to the phase shifting of the stimulus, so that

any signals not synchronized with the stimulating pattern were

reduced. As described previously, the active cortical

electrode was moved at increments of 2 mm on the surface of

the cortex while the area centralis was being stimulated. At

each new recording site, a sample of 100 signals was

collected. Signals picked up by the electrode were amplified





















































Figure 4. The canine area centralis.
Picture is of a typical, normal, Beagle. For
detailed description of the area, see the text.








43

X10,000 with a 2-300 Hz (3dB) bandpass and digitized. The 100

signals were then summed to reduce random noise, and the

result was displayed on a screen and stored on a magnetic disk

for future analysis.

Electrical signals produced by counterphasing of light

and dark grating bars were indicators of the functional nature

of the area centralis pathways as the retina and visual cortex

responded to stimulation. The signals' root mean square (RMS)

amplitude at each recording site was calculated off-line and

provided an unbiased estimate of the magnitude of the

responses (126). The simultaneous recording of PERG signals

was intended to rule out any variables which may affect the

VEP signals. As long as the ERG signals were consistent, any

variation in the VEP signals could have resulted only from the

movement of the active electrode on the surface of the cortex.


Histological Studies of the Area of Central Vision


Marking the Location of the Area of Central Vision


Following the electrophysiological recordings, the dog

was euthanized with an overdose of intravenous pentobarbital

sodium. Photographs were taken of the site which yielded the

signal with the highest amplitude. A Hamilton microliter

syringe (33 ga., 1/2 in.) was then mounted on the same

stereotaxic device, and used to inject 2.0 pL of fluoro-gold

at the same site (127). Fluoro-gold is a fluorescent

retrograde axonally transported tracer that provides intensive








44

labelling which does not fade over long periods of time (128).

This tracer was chosen as a marker because it is an extremely

stable compound and therefore compatible with numerous other

histochemical procedures (129).

Since numerical values for signal RMS amplitude and

latency were not available at this stage of the experiment, it

was decided to inject the tracer at the site generating the

signal with the highest amplitude. The site was chosen by

looking at the signals on the oscilloscope, and it was felt

that the signal with the highest amplitude could be identified

more reliably than the signal with the shortest latency. This

decision was later validated by the statistical analysis of

the electrophysiological results (see Chapter 3). The tracer

was injected 1 mm below the surface of the cortex. A nine cm2

block of cortex surrounding the injection site was then

removed from the brain.


Processing and Staining


The block of cortical tissue which was removed from the

brain was trimmed, and the white matter was notched for

orientation purposes. After peeling off the pia mater, the

specimen was fixed by immersion fixation (13, 17, 36) in a 4%

paraformaldehyde solution containing 5% sucrose (by weight).

Shrinkage resulting from this fixation method (13, 17, 36) has

been estimated at 25%, and accounted for.








45

Following fixation, the block was frozen (in dry ice),

and sectioned coronally at 40g using a sliding freezing

microtome. Each specimen was serially sectioned approximately

290 times. The sections were arranged in 6 parallel series in

trays containing diluted (1:10) fixative and mounted on

gelatin (0.5%) subbed slides (3 sections per slide). Sections

in alternate series were processed for the fluoro-gold tracer

or stained for either the cyto-architecture or the myelo-

architecture of the cortical tissue.

The stains used in this study were chosen based on

mapping studies in other species (16, 35). Standard cresyl-

violet (c. v.) stain was used for the cyto-architectural

series (16), and standard Weil stain was used for the myelo-

architectural series (35). The slides were then viewed through

a light microscope to find those sections containing the cyto-

and myelo-architectural characteristics of the area of central

vision as presented in Figure 1 (see Introduction).

Slides in series undergoing fluoro-gold processing were

dehydrated and defatted (127). The slides were then mounted

with fluoromount and cover slipped. Fluoro-gold labelling of

the sections was viewed by fluorescence microscopy (UV light).

Since the tracer was injected at the site which yielded the

highest amplitude signal, sections containing the fluoro-gold

stain were interpreted to represent the electrophysiologically

determined area of central vision. The distance between the








46
electrophysiologically- and histologically-determined areas of

central vision was then calculated, based on a distance of 40A

per section.


Photography


Sections containing the histological characteristics of

the area of central vision or the fluoro-gold tracer were

photographed. Neutral density filters were used, when needed,

to keep the exposure time under 0.5 seconds. Magnification was

X12.5. Sections stained with c. v. or Weil stains were

photographed both in black and white and color film. Black &

white prints were obtained using Kodak Technical Pan film (100

ASA). For comparison purposes, pictures were taken both with

and without a yellow filter. Color slides were obtained using

Kodak Ektachrome 50 film. A Kodak Ektachrome P800 (400 ASA)

film was used to photograph the sections containing the

fluoro-gold tracer.


Potential Problems in the Experimental Design

Use of Halothane as a General Anesthetic Agent


It is important to remember that in this study, halothane

was used during surgery, but not during the recording session

itself. During the craniotomy, the dogs were anesthetized

using a low (<1.5%) concentration of halothane. However, after

the cortex was exposed the use of halothane was discontinued.

At least 90 minutes (needed to set up the electrodes and solve








47
electrical noise problems) would then pass before the

beginning of the recordings. While this may not be sufficient

time for all of the halothane to be eliminated from the

cortex, other studies have shown that near normal

electrophysiological recordings can be conducted after this

time period (39, 43, 65, 130-134).

It is known that inhalation anesthetics affect, sometimes

profoundly, cortical activity. However, this does not suggest

that all brain or retinal activity ceases during anesthesia.

The brainstem, for example, is hardly affected by halothane

(132). The agent is recommended for use as an anesthetic in

recording brainstem auditory evoked potentials (BAEP's) (132).

In the retina (an extension of the central nervous system)

halothane has been found to have no effect on the decay time

course of rod and cone late receptor potential (133). In the

canine retina, DC--recorded a-- and b--waves are only mildly

changed in the presence of 2% halothane (131). Even the

c-wave, which is usually significantly reduced by halothane in

the dog, has on occasion been potentiated in the presence of

halothane (131), as has the b-wave of the rabbit ERG (135).

The effect of halothane on the electrophysiology of the

visual cortex is controversial. While some researchers claim

the agent has profound effects on VEP recordings, others

report the effect to be minor (39, 43, 65, 130, 134, 136).

Halothane/nitrous oxide anesthesia is reported to have a very

moderate effect on qualitative properties, such as the








48
preferred orientation and direction of stimuli movement, of

neurons in the visual cortex of cats (134). Indeed, Tusa et

al. used an anesthetic protocol very similar to the one used

in this study when mapping the retinotopic organization of

areas 18-21 of the cat's visual cortex (39, 43). In their

experiments, halothane was used for the surgical preparation

of the animal, and then the animal was paralyzed for the

recording session. Other researchers who recorded VEP's in

dogs went a step further and performed their recordings while

the animal was still under halothane anesthesia (65, 130).

Since the clearance time of halothane from the canine central

nervous system is less than 40 minutes (136), and dogs used in

these experiments were taken off halothane 90 minutes prior to

the recordings, it is reasonable to assume that halothane did

not interfere significantly with the cortical signals recorded

from these dogs.


Determining the Location of the Canine Area Centralis


The anatomically-determined area centralis of the dog is

a horizontally-oriented oval region (Figure 4) (137). It is

located in the tapetal funds, approximately one disc diameter

dorsal, and 2 disc diameters temporal, to the optic disc (138,

139). Fundoscopically, the area appears to be free of visible

blood vessels, but it does contain a capillary network (115).

As expected, this area contains the greatest concentration of

cones (5.1 per 50g field) in the canine retina (140), and the








49
greatest concentration (>5,000/mm2) of ganglion cells (137).

Therefore, the decision on where to project the stimulus field

was made based on the relation to the optic disc and the

absence of visible blood vessels.


Pain to the Animal and its Possible Influence on the Signal


All possible precautions were taken to minimize pain to

the animals used in this study. The animal was anesthetized

using halothane and a mixture of 75% nitrous oxide and 25%

oxygen. Mounting of the animal in the stereotaxic device and

all surgical procedures were performed under surgical

anesthesia. After the craniotomy, and prior to the removal of

the animal from the halothane, the regions in contact with the

ear-bars and the edges of the wound were injected with 2%

lidocaine hydrochloride. No tissue manipulations were

undertaken once the halothane administration had stopped.

During the entire recording session, the animal was ventilated

with the same nitrous oxide-oxygen mixture.

Once the animal was removed from halothane, it may have

experienced some discomfort due to its inability to change its

position, the use of the stereotaxic device and the surgical

wound. However, it is believed that the use of nitrous oxide

and lidocaine kept this discomfort at a minimal level. Close

attention was paid, throughout the experiment, to any signs of

pain. Core temperature, heart rate, end-tidal CO2 levels and,

obviously, cortical signals, were constantly monitored as








50
indicators of pain. In none of the recording sessions did we

note any significant changes in any of these indicators.














CHAPTER 3
RESULTS


Results in a Typical Dog


The distribution of signals recorded from one

representative normal Beagle (Bo) is shown in Figure 5, along

with landmarks of the cortical surface. As explained earlier,

the distance between two adjacent signals was 2 mm. It can be

seen that signals with the highest RMS amplitude (up to 59 AV)

and shortest latency (as low as 29 msec) are clustered around

the junction of the marginal and endomarginal sulcii. The

decrease in signal amplitude and increase in signal latency as

the electrode was moved from this region can be appreciated.

The stereotaxic coordinates of the sites generating the signal

with the highest RMS amplitude and the signal with the

shortest latency in this dog are presented in Table 2 and

Figure 6. The fluoro-gold tracer was injected at the site

generating the signal with the highest amplitude, and served

as a marker for the histological analysis (Figure 7).


Results in Normal Beagles

The location of the cortical area of central vision in

normal Beagles was determined by analysis of the electro-



























o) 0)
U V Ur
0 0>

3-0 -- i



41 ) -I. (04
me ,
H C4 0 *H r-

*- 0 0 0 0 *H


0 M 14 t 0 0P
m (1 VH 5m
o <0 kO go w o-
V 0* -4 '-4 > Q



-IC 0 -0 0

Mo k 4 4) M1
CO 40 M -I M
*4 4) 0 0 -*H I
0 0o II to 10

$4 1 10 I $4
O I 1- 0
Cr I 01 C 0 )
H 0 0 u 3 ,0 H
$4 30 4- ) -H m


V 40 $4 V 4W
w 0 ( 0 p 04 V

*4.-1 0 4O0 0 0








*rz

























(


I


E
Co
CA .Va
40


/


/


N\


4


h


CYv


4
4
4
4


4
4


0


i *
0
a (S
A 0.


(^













0
CMq





E
Et
r-.


L

-4-,
c
0


Figure 6.


Location of the cortical area of central vision in
normal Beagles.
The location of the cortical area of central vision
in seven normal Beagles was determined by the RMS
and latency of the electrophysiological signal and
by the cytoarchitecture and myeloarchitecture of the
histological slides. The three sets of coordinates
determined in each dog have been connected by lines
to form a triangle.
X axis--Distance (in mm) lateral to the midline.
Y axis--Distance (in mm) anterior to the IAP.
(D)--Dog Oa (0)--Dog Kd (A)--Dog Jl
(+)--Dog Sd (X)--Dog Oe (0)--Dog Bo
(v)--Dog Be (0)--Mean coordinates in normal Beagles


Normal Beagles




















4 6 8 10 12 14
lateral (mm)






































Figure 7.


Histological location of the cortical area of
central vision. Magnification x 12.5.
a) A 40A coronal section stained using c.v.
stain to illustrate the cytoarchitecture of the
region. The striate cortex, at the right, is a
granular region with large pyramidal cells in layers
III and V, and with small, densely-packed cells in
layer IV. The parastriate cortex at the left is also
granular, with layer III containing a greater number
of large pyramidal cells, but layers IV and V have
far fewer cells than in the striate cortex. The
needle tract resulting from the injection of fluoro-
gold can be seen at the left corner.
b) A 40g coronal section stained using Weil stain to
illustrate the myeloarchitecture of the region. In
the striate cortex on the left the heavily
myelinated outer stripe of Baillarger is visible in
layer IV, and the inner stripe of Baillarger is seen
in layer VI. The parastriate on the right is far
less myelinated. The needle tract resulting from the
injection of fluoro-gold can be seen at the right
corner.





























































Figure 7--continued








57

physiological signal recorded, and by the cytoarchitecture and

myeloarchitecture of the histological sections of the brain.

Sites with the highest RMS amplitude signal, the shortest

latency signal and the typical histological characteristics

were designated as the cortical area of central vision as

determined by each technique. The results in seven normal

Beagles are presented in Figure 6. The coordinates of the

cortical area of central vision in these seven dogs, as

determined by each technique, together with the mean results

for this experimental group, are presented in Table 2.

Two methods of statistical analysis were used to evaluate

the three mapping techniques. Results obtained by each mapping

technique were compared to the results obtained by the other

two techniques using multiple Student's t-tests. There were no

significant differences between the results obtained by any of

the three mapping techniques at neither the anterior-posterior

axis (p>0.18) nor at the medial-lateral axis (p>0.64). An

analysis of variance (ANOVA) test yielded similar results.

There were no significant differences between the three mean

coordinates, at neither the anterior-posterior axis (p=0.32)

nor at the medial-lateral axis (p=0.85). Parametric methods of

analysis were used based on the assumptions that the

population of Beagles is normally distributed and that the

sample was randomly chosen.

A second method of analysis was to calculate the

distances between the three sets of coordinates obtained in











TABLE 2
RESULTS IN NORMAL BEAGLES


Dog Mapping
technique


Coordinates"
AP ML
axis axis


10.7
7.3
8.5
9.2
9.0
12.2
7.9
11.2
9.3
15.0
13.4
10.6
15.0
7.2
10.8
12.0
11.7
14.4
10.9
8.1
15.4


8.8
10.1
8.2
5.9
4.3
7.9
6.5
5.7
8.3
11.5
8.6
6.7
11.8
11.9
8.0
6.3
5.1
7.9
7.2
9.9
13.2


Distance between
coordinates
(in mm)c


AL=3.6
HL=2.2
HA=2.3
AL=1.6
HL=4.8
HA=3.6
AL=3.4
HL=3.2
HA=2.3
AL=3.3
HL=3.4
HA=6.5
AL=7.8
HL=5.3
HA=5.7
AL=1.2
HL=3.9
HA=2.9
AL=3.9
HL=8.0
HA=7.5


Mean" A 11.5 1.0 8.3 0.9 AL=3.5 0.8
L 9.7 0.9 7.9 1.1 HL=4.4 0.7
H 11.6 1.0 8.6 0.8 HA=4.4 2.1
1--------------------------


-Mapping technique used:
A--Analysis of RMS amplitude of signal
L--Analysis of latency of signal
H--Analysis of histological data
bCoordinates (in mm) along the anterior-posterior (AP) and
medial-lateral (ML) axes. All of the coordinates given are
lateral to the midline and anterior to the IAP.
CDistance between the areas of central vision as determined
by:
AL--the RMS amplitude and the latency of the signal
HL--histology and the latency of the signal
HA--histology and the RMS amplitude of the signal
dMean coordinates one standard deviation.








59

each dog, i.e., the lengths of the three sides of the triangle

described by these coordinates (Figure 6). Results of these

calculations are presented in the last column of Table 2. In

no case were two coordinates in a given dog more than 8 mm

apart, and the mean distance between any two coordinates was

4.1 mm. Multiple Student's t-tests yielded no significant

differences between the three sets of distances calculated

(p>0.44). An ANOVA test revealed no significant differences

between the three mean distances (p=0.47), implying that the

triangle described by the three mean coordinates is

statistically equilateral.

Based on the overall mean of the coordinates obtained by

all three mapping techniques, it was determined that the

cortical area of central vision in normal Beagles lies 10.9

0.6 mm anterior to the IAP and 8.3 0.2 mm lateral to the

midline (results are presented as mean one standard

deviation).


Results in Pre-qlaucomatous Beagles


The location of the cortical area of central vision in

pre-glaucomatous Beagles was determined by analysis of the

electrophysiological signal recorded, and by the

cytoarchitecture and myeloarchitecture of the histological

sections of the brain. Sites with the highest RMS amplitude

signal, the shortest latency signal, and the typical

histological characteristics were designated as the cortical








60

area of central vision as determined by each technique. The

results in two pre-glaucomatous Beagles are presented in

Figure 8. The coordinates of the cortical area of central

vision in these two dogs, as determined by each technique,

together with the mean results of this experimental group, are

presented in Table 3.



TABLE 3
RESULTS IN PRE-GLAUCOMATOUS BEAGLES


Dog Mapping CoordinatesD Distance between
techniques AP ML coordinates
axis axis (in mm)c

Pa A 13.4 10.1 AL=7.6
L 7.7 5.1 HL=4.9
H 11.8 7.7 HA=2.9
Se A 11.2 7.3 AL=1.7
L 12.8 7.9 HL=3.0
H 13.1 10.9 HA=4.1

Meand A 12.3 1.1 8.7 1.4 AL=4.7 3.0
L 10.3 2.6 6.5 1.4 HL=4.0 1.0
H 12.5 0.7 9.3 1.6 HA=3.5 0.6

aMapping technique used:
A--Analysis of RMS amplitude of signal
L--Analysis of latency of signal
H--Analysis of histological data
bCoordinates (in mm) along the anterior-posterior (AP) and
medial-lateral (ML) axes. All of the coordinates given are
lateral to the midline and anterior to the IAP.
CDistance between the areas of central vision as determined
by:
AL--the RMS amplitude and the latency of the signal
HL--histology and the latency of the signal
HA--histology and the RMS amplitude of the signal
dMean coordinates one standard deviation.













Glauc. Beagles


6 8 10 12
lateral (mm)


Figure 8.


Location of the cortical area of central vision in
pre-glaucomatous Beagles.
The location of the cortical area of central vision
in two pre-glaucomatous Beagles w as determined by
the RMS and latency of the electrophysiological
signal and by the cytoarchitecture and
myeloarchitecture of the histological slides.
The three sets of coordinates determined in each dog
have been connected by lines to form a triangle.
X axis--Distance (in mm) lateral to the midline.
Y axis--Distance (in mm) anterior to the IAP.
(X)--Dog Pa (*)--Dog Se
(v)--Mean coordinates in pre-glaucomatous Beagles


V

0
"E

4.,
0








62

Two methods of statistical analysis were used to evaluate

the three mapping techniques. Results obtained by each mapping

technique were compared to the results obtained by the other

two techniques using multiple Student's t-tests. There were no

significant differences between the results obtained by any of

the three mapping techniques at neither the anterior-posterior

axis (p>0.49) nor the medial-lateral axis (p>0.32). An ANOVA

test to compare the three mean coordinates could not be

conducted due to the small sample size of this experimental

group.

A second method of evaluation was to calculate the

distances between the three sets of coordinates obtained in

each dog, i.e., the lengths of the three sides of the triangle

described by these coordinates (Figure 8). Results of these

calculations are presented in the last column of Table 3. In

no case were two coordinates in a given dog more than 7.6 mm

apart, and the mean distance between any two coordinates was

4.1 mm. Multiple Student's t-tests yielded no significant

differences between the three sets of distances calculated

(p>0.73). An ANOVA test of the three mean distances could not

be performed due to the small sample size of this experimental

group.

Based on the overall mean of the coordinates obtained by

all three mapping techniques, it was determined that the

cortical area of central vision in pre-glaucomatous Beagles








63

lies 11.7 0.7 mm anterior to the IAP and 8.2 0.9 mm lateral

to the midline.

Results in Greyhounds


The location of the cortical area of central vision in

Greyhounds was determined by analysis of the electro-

physiological signal recorded, and by the cytoarchitecture and

myeloarchitecture of the histological sections of the brain.

Sites with the highest RMS amplitude signal, the shortest

latency signal and the typical histological characteristics

were designated as the cortical area of central vision as

determined by each technique. The results in three Greyhounds

are presented in Figure 9. The coordinates of the cortical

area of central vision in these three dogs, as determined by

each technique, together with the mean results of this

experimental group, are presented in Table 4.

Two methods of statistical analysis were used to evaluate

the three mapping techniques. Results obtained by each mapping

technique were compared to the results obtained by the other

two techniques using multiple Student's t-tests. There were no

significant differences between the results obtained by any of

the three mapping techniques at neither the anterior-posterior

axis (p>0.88) nor at the medial-lateral axis (p>0.30). An

ANOVA test yielded similar results. There were no significant

differences between the three mean coordinates, at neither the

















Greyhounds


4 6 8 10 12 14
lateral (mm)


Figure 9.


Location of the cortical area of central vision in
Greyhounds.
The location of the cortical area of central vision
in three Greyhounds was determined by the RMS and
latency of the electrophysiological signal and by
the cytoarchitecture and myeloarchitecture of the
histological slides. The three sets of coordinates
determined in each dog have been connected by lines
to form a triangle.
X axis--Distance (in mm) lateral to the midline.
Y axis--Distance (in mm) anterior to the IAP.
(X)--Dog Kf (*)--Dog Br (v)--Dog Bn
()--Mean coordinates in Greyhounds


e-N
E
E

0

,4--
C
0










TABLE 4
RESULTS IN GREYHOUNDS


Dog Mapping Coordinatesb Distance between
techniques AP ML coordinates
axis axis (in mm)c

Kf A 16.9 5.5 AL=6.6
L 17.7 12.1 HL=4.3
H 14.9 8.9 HA=3.9
Br A 15.2 8.9 AL=2.5
L 13.9 6.8 HL=0.5
H 13.6 7.2 HA=2.3
Bn A 14.2 7.7 AL=3.8
L 15.3 9.3 HL=3.4
H 18.6 9.9 HA=4.9

Meand A 15.4 0.8 7.4 1.0 AL=4.3 1.2
L 15.6 1.1 9.4 1.5 HL=2.7 1.1
H 15.7 1.5 8.7 0.8 HA=3.7 0.8

aMapping technique used:
A--Analysis of RMS amplitude of signal
L--Analysis of latency of signal
H--Analysis of histological data
coordinates (in mm) along the anterior-posterior (AP) and
medial-lateral (ML) axes. All of the coordinates given are
lateral to the midline and anterior to the IAP.
CDistance between the areas of central vision as determined
by:
AL--the RMS amplitude and the latency of the signal
HL--histology and the latency of the signal
HA--histology and the RMS amplitude of the signal
dMean coordinates one standard deviation.


anterior-posterior axis (p=0.99) nor at the medial-lateral

axis (p=0.59).

A second method of analysis was to calculate the

distances between the three sets of coordinates obtained in

each dog, i.e., the lengths of the three sides of the triangle

described by these coordinates (Figure 9). Results of these

calculations are presented in the last column of Table 4. In

no case were two coordinates in a given dog more than 6.6 mm








66

apart, and the mean distance between any two coordinates was

3.6 mm. Multiple Student's t-tests yielded no significant

differences between the three sets of distances calculated

(p>0.40). An ANOVA test revealed no significant differences

between the three mean distances (p=0.27), implying that the

triangle described by the three mean coordinates is

statistically equilateral.

Based on the overall mean of the coordinates obtained by

all three mapping techniques, it was determined that the

cortical area of central vision in Greyhounds lies 15.6 0.1

mm anterior to the IAP and 8.5 0.6 mm lateral to the midline.




Comparison of the Three Experimental Groups


Normal and Pre-qlaucomatous Beagles


Results of the three mapping techniques (i.e., analysis

of RMS amplitude, latency and histology) were compared in

normal and pre-glaucomatous Beagles. Using the Student's t-

test, there were no significant differences in the location of

the cortical area of central vision between the two groups, at

neither the anterior-posterior axis (p=0.72, p=0.80 and

p=0.67, respectively) nor at the medial-lateral axis (p=0.84,

p=0.54 and p=0.69, respectively). Based on these results, it

was determined that the cortical area of central vision in

Beagles lies 11.3 0.4 mm anterior to the IAP, and 8.3 0.1 mm

lateral to the midline.








67

The location of the cortical area of central vision in

Beagles is presented in Figures 10 and 11. Figure 10 is a

potential map showing the distribution of the mean signals'

RMS amplitudes recorded from the cortical surfaces of nine

Beagles. The decrease in signal amplitude as one moves away

from the cortical area of central vision can be appreciated

(see also Figure 5). Figure 11 is a latency map showing the

distribution of the mean signals latencies recorded from the

cortical surfaces of nine Beagles. The increase in signal

latency as one moves away from the cortical area of central

vision can be appreciated (see also Figure 5).

Figure 12 presents the rate of change in the mean

signal's RMS amplitude and latency in nine Beagles as the

electrode was moved from the cortical area of central vision

along the anterior-posterior and medial-lateral axes. Since

simultaneous PERG recordings did not alter significantly

during the experiments, changes seen in the PVEP amplitude and

latency can only be due to changes in the cortical recording

site.


Beagles and Greyhounds


No significant differences in the location of the

cortical area of central vision along the medial-lateral axis

were found between normal Beagles and Greyhounds (p=0.58,

p=0.48 and p=0.96, respectively). Nor were any such differences









10 WO W
C S C 2

(o < (D
a)
4J


5H c *
Cp r4 + 0 3
U))
0 ."0 14o

H *H.M
to V 4M 0 He
0 > O $4 C- 4)
0 A r-4 p
Ml 4) tJ"-l a 0<
Im9 4* j 4-4 4
*H 4 0 0 *0


4)0
4J r 4 1 r.

4- r-q C
ai C o Q)
*H0 C *H 4) 4C 0

4 )4 *M m .0
rq.4 4- 4) iC V I)
aO H'OeO0 .v40
O *t C +| [ (0


$4 4J 4) 4)0
*Hp 430W4.) 4 3 a
HO to U+H 4 W


~ 4) k44H 0 go4)
g.H O) Sa) 0

- 0 V 0 0M 9 00
C O 0 r4 4- 0 4 it
0V 0 H0 m r
-l r + O Cto .H C+
-A S -AQ- 0 4
Wx V4 0 a>-) 0
-H Hn ,4 H 0
9: Ol 01 *4 r- I4 0 o
r. 4 1H 4 O
4) r-i to vn *> w *A
0 : V 0)
aC **caM 0-a) -
S *M ff4&>,CE-*W^
4)-H 4) ..! ) )-I 0
S 0 4)0 0 H3
V 91- u Va IV O -Hp

id g; w a) t
44 4o i 9 0 m


r. .V -4 )V 4) r r f
M0 W C O -r-
o V M 0 'a M 0 o


*-) aI W I t o 000r
*) iCICI Uai) c
:3 M H 0 *H r-lq 0 *H
.0.44(* X- XiOhgO.4Hr
A tox -HH x o r-i
i t I 0 0a a)
4 4 H O --1C 0 DQ
C0000 $4 -44- 0
0r 0 COr- 0f p O+0 H'-
rl 4 *k- 0 -r +1 -r4
Z c ) ) V+ r-i r-I
4) V +# H (HD
43^ccmi43eC *^*ex
to 44)S-) .9 Q44
E-1l H H i0 .> E-4 10 CO A 0


r*I
0



t-
*r-l
44































47

I-


01V 0


09 O0




















































oV 0


09 O9









10 r> qC
4) 4J C


0 0 E) r.



, .) H 0A
0 4 to


) 0 W
ro 04.) I H
M )C 4.



o4 *r-i B 0. .
- H O 0 4J .
S0 4) c.-



4.) a -H 0 O r .
O0 4V(DWo MC4 0


4 0 Ho aO )I 4 00 uM
a.)0 4 V 4J 00 A ,)4
(0 0-4 U ) 4+) 0 to


4H 0 0 0) CO C r.
CV 0 --1 -4 it ` W4J
r 1 0 Of ) 4 -W

00 H HH-t M *H 0
H ) k0 $ 0 to


Ha 4) 4 a) 0 0u


,B -.4 C >iC *He g -,
I0 *4 C O .5 m
64J (d (0 *A C q V (D I





P C *HI M t) 4) to
>40 t -r F 0 m
*U) Hto ) p r-I a) ft )



4) 44.H 1 v 4' 4


3.O I X O 'a) Va

w -H 0 4q (P -) 4 V )

VH r 0 Ni O rOl ) 0'-"
Id-I ) a) 44 0 a) 0 H H g0
04. (0 r. 40 V 0
.0 a IOVH O O


a ) to I 'l )4) p tv
3 C ) a 0 0 -.H V -Hr
V E-it l-H 4 C I M E- H + ) H

rgP
r4-HIU r4+ r. tv


49 m4 )r








72








S --------------------- .............................-....




co




















E
-\J


















ocr







i"-

III







08 OL 09 09 0 OC Oz
(oesw) AON31V71
















CJ

E
E
co0i
0 c
-J
W


'"0
I1-











c'


08 OL 09 OS Ot O0 Oz
(oesw) AON31V1








/U


60


50 -


40
S40 -






20 -


10 I I I \
0 5 10 15 20 25

anterior-posterior (mm)
Figure 12. Rate of change in mean signal RMS amplitude and
latency in Beagles.
a) Rate of change in mean signal RMS amplitude and
latency along the anterior-posterior axis in nine
Beagles.
(0) Root mean square amplitude of mean signal (in
AV). Stereotaxic coordinates along the anterior-
posterior axis are shown on the X-axis. Medial-
lateral coordinates were 8.3 mm.
(0) Latency of mean signal (in msec). Stereotaxic
coordinates along the anterior-posterior axis are
shown on the X-axis. Medial-lateral coordinates
were 8.3 mm.
b) Rate of change in signal RMS amplitude and
latency along the medial-lateral axis.
(0) Root mean square amplitude of mean signal (in
pV). Stereotaxic coordinates along the medial-
lateral axis are shown on the X-axis. Anterior-
posterior coordinates were 11.3 mm.
(0) Latency of mean signal (in msec). Stereotaxic
coordinates along the medial-lateral axis are shown
on the X-axis. Anterior-posterior coordinates were
11.3 mm.




















I I I I I I I


50 -


40 k


30 F


0 2 4 6 8 10 12 14 1


medial-lateral (mm)










Figure 12--continued
I








76

found between pre-glaucomatous Beagles and Greyhounds (p=0.48,

p=0.28 and p=0.71) respectively. In otherwords, no inter-breed

differences were found in the location of the cortical area of

central vision along the medial-lateral axis.

All three mapping techniques showed significant

differences in the location of the cortical area of central

vision along the anterior-posterior axis between normal

Beagles and Greyhounds (p=0.04, p=0.005 and p=0.04

respectively). However, none of the mapping techniques

revealed such differences between pre-glaucomatous Beagles and

Greyhounds (p=0.09, p=0.10 and p=0.20, respectively).


Comparison of the Three Mapping Techniques


Accuracy of the Three Mapping Techniques


A X2 test of observed and expected frequencies was

performed to test the hypothesis that no one mapping technique

yielded consistently more accurate results than the other two

techniques. If the hypothesis is false and one technique

yields more accurate results, the coordinates obtained by that

technique would consistently be closest to the mean

coordinates of the experimental group. However, if the

hypothesis is true and no differences exist between the

mapping techniques, each technique would be expected to prove

most (and, conversely, least) accurate in four of the 12 dogs

tested. The observed and expected frequencies of the number of

times each technique yielded the most accurate results in the








77

anterior-posterior and medial-lateral axes are given in Table

5 and Table 6, respectively.

The result of the test for the anterior-posterior axis

was X2=0.5. The result of the test for the medial-lateral axis

was X2=3.4. Based on these results it was determined (at a

confidence level of a=5%) that at neither axes was one mapping

technique superior to the other two techniques.



TABLE 5
EXPECTED AND OBSERVED FREQUENCIES
OF ACCURACY IN THE ANTERIOR-POSTERIOR AXIS

Amplitude Latency Histology Total
of signal of signal

Expected 4 4 4 12
Observed 4 3 5 12



TABLE 6
EXPECTED AND OBSERVED FREQUENCIES
OF ACCURACY IN THE MEDIAL-LATERAL AXIS

Amplitude Latency Histology Total
of signal of signal

Expected 4 4 4 12
Observed 2.5 2.5 7 12



Linkage Between the Three Mapping Techniques


It has already been stated that ANOVA and multiple

Student's t-tests failed to reveal any significant differences

in the distances between the three sets of coordinates

obtained for a given dog at any of the experimental groups.

This finding implies that no two given mapping techniques








78

yielded consistently similar results. Had there been such a

pair of mapping techniques, the distance between the

coordinates obtained by those techniques would have

consistently been the shortest of the three calculated

distances.

A X2 test of observed and expected frequencies was also

conducted to further check the degree of linkage between the

three tests. If no two given mapping techniques yielded

consistently similar results, the distance between the

coordinates obtained by those two techniques would be the

shortest (and, conversely, the longest) in four out of the 12

dogs. The observed and expected number of times in which a

distance was the shortest of the three measured distances in

a given dog are shown in Table 7.


TABLE 7
EXPECTED AND OBSERVED FREQUENCIES
OF SHORTEST DISTANCES

AL HL HA TOTAL


Expected 4 4 4 12
Observed 5 4 3 12


The result of the test was X2=0.5. Therefore, it was

determined (at a confidence level of a=5%) that no two tests

possessed an unusually high degree of linkage.












The Canine AMF


The AMF for the cortical area of central vision was

determined by calculating the area where a signal which was at

least 37% of the maximal signal could be recorded. The 37%

threshold, equivalent to a 63% drop in the maximal signal, was

chosen because it represents the distance along which

extrapolar electronic potential decays to 1/e (natural

logarithm) of its original level (141). This value, known as

the space constant, or 1, is used extensively in

neurophysiology to calculate signal amplitude at a given

distance from the generating source of the action potential

(141).

Based on this threshold value and on the RMS amplitude of

the mean signal recorded in Beagles (Figure 10), it was

determined that on the cortical surface, the canine area of

central vision is described by a circle with a radius of 12.2

mm. In other words, 468 mm2 of cortex are devoted to

representation of the central 150 of vision. Therefore, the

mean AMF for the central 150 of vision in the dog is 2.6

mm2/degree2 of visual field.














CHAPTER 4
DISCUSSION


The Cortical Area of Central Vision in the Dog


Location of the Area of Central Vision on the Surface of the
Canine Cortex


Based on the results of this study, it was determined

that the cortical area of central vision in Beagles lies 11.3

0.4 mm anterior to the IAP, and 8.3 0.1 mm lateral to the

midline (see Figures 10 and 11). In Greyhounds the cortical

area of central vision lies 15.6 0.1 mm anterior to the IAP

and 8.5 0.6 mm lateral to the midline. Anatomically, the

canine area of central vision in both breeds is located near

the medial border of the posterior occipital lobe, at the

junction of the marginal and endomarginal gyri (see Figures 13

and 14).


Glaucoma and the Location of the Area of Central Vision


No significant differences in the location of the

cortical area of central vision were found between normal and

pre-glaucomatous Beagles (p>0.54). This finding may be due to

the small sample (n=2) of glaucomatous Beagles used in the

study, but it is more likely that in mature dogs primary

glaucoma has no effect whatsoever on the location of the

80
















































Figure 13. A diagram of the canine cerebral cortex.
Top-a left dorsolateral view.
Bottom-a lateral view.
(1) endomarginal gyrus; (2) endomarginal sulcus;
(3) marginal gyrus; (4) marginal sulcus.

From: Evans HE, Christensen GC (Eds). Miller's Anatomy of the
Dog. WB Saunders Co., Philadelphia (2nd edition) 1979
(ref #142).












1 2 3 4


Figure 14.


A representation of the canine medial occipital
lobe in the dog (transverse section).
(1) endomarginal gyrus; (2) endomarginal sulcus;
(3) marginal gyrus; (4) marginal sulcus.


From: Yoshikawa T. Atlas of the Brains of Domestic Animals.
University of Tokyo Press, Tokyo, 1968 (ref #143).







83

cortical area of central vision. A careful search of the

literature failed to reveal a report of such an effect in any

species. While primary glaucoma is known to affect VEP

recordings in glaucomatous humans and in nonhuman primates

with experimentally induced IOP elevation (105-113), this is

most likely due to the effect of the elevated pressure on the

retina and the optic nerve rather than to a direct effect on

the visual cortex itself.

Numerous studies, including landmark works by Hubel and

Wiesel, have shown that visual deprivation at a young age can

permanently disrupt the neuronal organization of the visual

cortex in nonhuman primates and other species (32, 144).

However, by age 6 weeks in nonhuman primates (32), and age 8

weeks in cats (144) the visual cortex completes its

development, and visual deprivation has no effect on its

neuronal organization. While no similar data exists in dogs,

it has been shown that the canine visual cortex assumes the

morphological characteristics of an adult cortex by age 4

weeks (145), and that canine VEP's reach adult levels by age

100 days (53). The pre-glaucomatous dogs used in these

experiments were from a colony of Beagles with hereditary POAG

in which visual losses resulting from the disease occur at a

much later age (119). It is conceivable that congenital

glaucoma in dogs (and other species) would disrupt the

neuronal organization of the visual cortex. But it is

extremely hard to imagine such a disruption occurring when the








84

onset of the clinical disease is long after the cortex has

completed its physiological and morphological development, at

an age when even full deprivation has no effect in other

species (32, 144). This line of reasoning, together with the

lack of significant differences in the results of glaucomatous

and normal Beagles, led to the pooling of the results of the

two experimental Beagle groups.


Comparison of Results in Beagles and Greyhounds


No significant differences in the location of the

cortical area of central vision were found between Beagles and

Greyhounds in the medial-lateral axis (p>0.28). However

significant differences in the location of the cortical area

of central vision between the two breeds were found along the

anterior-posterior axis (p=0.04). Unfortunately, no

comparative study of the Beagle and Greyhound brain has been

conducted at neither the histological nor the macroscopic

level (Drs. C. Chrisman, F. Thompson and R. Reep, personal

communications). Therefore it cannot be determined if the

variation in the location of the cortical area of central

vision found between the two breeds is a result of differences

in brain sizes or the result of evolutionary divergence in two

canine breeds bred for different tasks.

It has been shown that the latency of BAEP's in the dog

is correlated with body weight and with cranium distance

(which reflects the cranium length and width). An increase of







85

1 cm in cranium distance can increase the canine BAEP latency

by up to 0.03 msec, while an increase of 1 kg in body weight

can increase the latency by up to 0.01 msec (146). It has been

estimated that head size, which accurately reflects brain

size, is a source of 25% of intersubject variance in the

latency of the canine BAEP (146). No data exists on the

effect of head and brain size on canine VEP's, but in humans

brain size has been shown to affect the latency of the P100

(147). These facts, combined with the similarity in the

anatomical (as opposed to the stereotaxic) location of the

cortical area of central vision in Beagles and Greyhounds tend

to support the hypothesis that differences between the two

breeds are a result of differences in head and brain size.


Histological Location of the Cortical Area of Central Vision


No significant differences were found between the

cortical areas of central of vision in Beagles (p>0.18) and in

Greyhounds (p>0.30) as determined by electrophysiology and

histology. In fact, in both breeds, there was a remarkable

degree of correlation between the electrophysiological and

histological findings. This finding is also consistent with

results of investigations of all species mapped to date,

showing the electrophysiologically-determined cortical area of

central vision to be invariably located at the junction of

Brodmann's areas 17 and 18 (3, 4, 7, 8, 11, 13, 17, 20, 26).








86

Several researchers have conducted histological studies

of the canine brain (148-151). Early maps divide the canine

occipital cortex into three cytoarchitectonically-defined

areas, which, based on Brodmann's nomenclature, have been

designated 17a, 17b and 18 (149, 150). Based on this

terminology, the cortical area of central vision lies at the

border of areas 17a and 17b. More recently, it has been

proposed that the canine occipital cortex is divided into

six cytoarchitectonically-defined areas (151). Based on this

terminology, the cortical area of central vision lies at the

border of areas M and Emr. Another study, based on the

myeloarchitecture of the region, suggests there are 15

morphologically-defined areas in the canine occipital cortex

(148). The three largest areas have been designated MP, BP and

QP (148), and the cortical area of central vision lies at the

border of areas MP and BP. It remains to be determined

whether, as in other species, the division of the canine

occipital cortex into several morphologically-defined areas

can be correlated with the existence of multiple visual maps

and higher processing centers.



Comparison to the Cat


The only other carnivore whose visual cortex has been

mapped to date is the cat (2). The stereotaxic coordinates of

the cortical area of central vision in the cat are 3 mm







87

posterior to the IAP and 5 mm lateral to the midline (see

Figure 2) (40). Based on the results of this study, the

cortical areas of central vision in the cat and the Beagle are

14.7 mm apart. In the cat and the Greyhound, the distance is

even further, 18.9 mm.

There is a much smaller difference in the anatomical

location of the cortical area of central vision between the

two species, if at all. In both carnivores, the cortical area

of central vision is located near the medial border of the

posterior occipital lobe. More specifically, it is located on

the crown of the lateral gyrus, near the junction of the

lateral and posterior lateral gyri in the cat (20). In the

dog, it is located at the junction of the marginal and

endomarginal gyri (see Figures 13 and 14). Studies that

compared the fissural pattern of the canine and feline brains

report that the marginal gyrus of the dog is homologous to the

lateral gyrus of the cat (152, 153). Indeed, at least two

textbooks refer to the marginal gyrus of the dog as the

"lateral gyrus" (118, 120), and one textbook refers to the

lateral gyrus of the cat as the "marginal gyrus" (154).

Once again, it remains to be determined whether the

difference in the location of the cortical area of central

vision in the dog and the cat is the result of difference in

brain size or the result of evolutionary changes. In other

words, has evolution in the dog consisted simply of increased

elaboration of the same parts already established in the







88

common ancestors of carnivores, or have new parts evolved?

The question has already been posed by vision researchers

(17), and obviously a definitive answer cannot be provided

without finding the common carnivore ancestor.

As in the comparison of the Beagle and the Greyhound, the

similarity in anatomical, as opposed to stereotaxic, location,

lends support to the former hypothesis. A second clue may be

found by comparing the AMF of the cat and the dog. A larger

cortical area is devoted to representation of the central

visual area in the dog brain, as compared to the cat brain. In

the mid-sized dog, 468 mm2 of cortex are devoted to

representation of the central 150 of vision, while in the cat

the respective figures are 190 mm2 and 200 (20).

It has already been established by Tusa et al. that in

the cat AMF values decrease linearly along both the horizontal

and vertical meridians (20, 39). The same assumption has been

used by Whitteridge in calculating the AMF in nonhuman

primates (35). If the same principle holds true for the dog,

then the AMF for the central 1 of vision in the dog is 3.85

mm2 of cortex/degree2 of visual field. In the cat, the peak

AMF value is 3.6 mm2 of cortex/degree2 of visual field (20).

The AMF is directly related to cone density, ganglion

cell concentration and visual resolution both in nonhuman

primates (35, 155) and in humans (156). Consequently, the

relationship of AMF to the number and concentration of

ganglion cells indicates a precise, and similar, topographical







89

arrangement of the visual pathways in nonhuman primates (35,

155) and in humans (156). The relationship between AMF and the

cone density indicates a similar arrangement of retinal

organization in these species (35, 155, 156). In humans it has

been shown that both cone and ganglion cell distributions are

related to image resolution, color perception and other

important basic visual functions (156).

As Table 8 shows, the same principles hold true in

carnivores. It would seem that cats and dogs share similar

topographical arrangement of their retina and visual pathways.

Furthermore, it seems that these two carnivores share similar

principles of organization with nonhuman primates and humans,

a finding that may help shed light on the evolution of vision

in these divergent species.


TABLE 8
RETINAL CELL COUNTS
AND VISUAL FUNCTION


Species AMF Maximal cone number of resolution
mm2/degree2 density/mm2 ganglion cells (cpd)


cat 3.6(20) 27,000(157) 116,000(158) 3.5(159)

dog 3.9a 102,000(140) 167,000(160) 4.3(102)

rhesus 5.5(155) 148,000(155) 1.OxlO6 (161) 46(155)

man 11.6(162) 260,000(163) 1.3x106 (164) 60(165)

this study













The Recorded Signal


Several hypotheses can be used to explain the changes in

signal RMS amplitude and latency as the electrode was moved

away from the cortical area of central vision (see Figure 12).

It is possible that the observed decrease in RMS amplitude is

due to "volume conduction", the decay observed in any

electrical signal as one moves away from its source. However,

such a decay would be constant (and proportional to

1/distance2) in all directions, and it is obvious by comparing

Figures 12a and 12b that the rate of decrease in RMS amplitude

was not the same along both axes. Furthermore, had the change

in signal RMS amplitude been due to volume conduction only,

one would not expect to see it accompanied by a concomitant

increase in latency. Conduction over the distances involved

should be instant.

Obviously, physiological decay in signal RMS amplitude

due to volume conduction occurs in the canine cortex. However,

it alone cannot account for the observed changes in the

recorded RMS amplitude and latency. These changes can only be

explained if the existence of other cortical visual areas is

hypothesized. As stated previously, it is more than likely,

based on findings in all other mammalian species and on the

histological divisions of the canine occipital cortex, that

the dog has more than one cortical visual area. Furthermore,








91

it has been established in both the cat (43) and in nonhuman

primates (68) that the area centralis/macula is not equally

represented in all of the cortical visual areas. In other

words, in some visual areas of these species the central

retina is hardly represented (has a low AMF value). Therefore,

stimulation of the area centralis/macula would not result in

significant biological activity in these cortical areas. It is

very likely that similar organization exists in the dog, and

as the electrode was moved into these regions less neuronal

activity was recorded. This hypothesis would also explain the

observed increase in signal latency, since visual input from

the primary visual cortex to these areas would involve one or

more synapses.

Significance of this Study


The location of the visual cortex in man has been known

for more than 70 years (1). The cat's visual cortex was mapped

50 years ago (2). In both these species, knowing the precise

location of the visual cortex and the area of central vision

has proven to be a vital piece of information for both

researchers and clinicians. This knowledge is essential for

our understanding of the evolution, organization and function

of both the visual system and the brain. Knowledge of the

accurate location of the primary visual cortex is also

valuable in diagnosis and therapy of cerebral disorders,

trauma and neoplasia, and performing effective neurosurgery




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EJU2D2JAE_L16G4Z INGEST_TIME 2012-02-20T23:04:27Z PACKAGE AA00009056_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES