University of Southern California Dissertations and Theses

Electrical stimulation of degenerate retina

(USC Thesis Other)

Electrical stimulation of degenerate retina

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
ELECTRICAL EXCITATION OF DEGENERATE RETINA

by
Lai-Hang (Leanne) Chan

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOMEDICAL ENGINEERING)

August 2009

Copyright 2009 Lai-Hang (Leanne) Chan
ii
Dedication

I would like to dedicate my dissertation work to my grandparents, my parents and
my sister, whose love and patience have strengthened me and kept me going forward to
the mark through these seemingly endless years.
I dedicate this work and give special thanks to my counselor, my advisor and my
cheerleader, Jesus Christ, who has been and still is my constant source of support and
inspiration.

iii
Acknowledgements

I owe my deepest gratitude to my advisor, Dr. James Weiland, a wonderful
mentor whose encouragement, supervision and support from the preliminary to the
concluding level enabled me to develop an understanding of the research material, as well
as broadened my horizon in the research field to imagine the impossibility and create a
better world for the disabled individuals, and also provided a role model of balancing
academia life and family life.
I am also grateful for my committee members Dr. Mark Humayun, Dr. Noberto
Grzywacz, Dr. Judith Hirsch and Dr. David D’Argenio for critically reviewing this
dissertation, and spending time and attention during busy semesters.
I would also like to thank my professors for showing me by example on how to
think, teach, and teach others to think, including Dr. Eun-Jin Lee, Dr. Armand Tanguay,
Dr. Gerald Chader, Dr. Ellis Meng, Dr. Stanley Yamashiro and Dr. Michael Khoo
I wish to thank also my colleagues, both in Health Science Campus and Main
Campus, for sharing their enthusiasm for and comments on my work, including Aditi Ray,
Alice Cho, Devyani Nanduri, Tim Nayar, Brooke Basinger, Matthew Behrend, Alan
Horsager, Neha Parikh, Vivek Pradeep, John Xie, Xiwu Cao and Junkwan Lee.
iv
I would like to acknowledge and thank the technical team in our group, Lindy
Yow, Doris Lee, Zhenhai Chen, Xiao Peng, Lina Flores, Carlos Sanchez, Fernando
Gallardo, Douglas Stevenson and Ernesto Barron, for their assistance and support.
It is also my greatest pleasure to thank my parents and my sister who gave me
moral support and words of encouragement. I am grateful for their unconditional love,
and understanding during the long years of my education. My dear friend Rocky Cheung
was by my side in these years; my thanks to him are beyond words and I am looking
forward to where God will lead us.
Thanks to God who carried me through the difficult periods of writing this
dissertation and to whom I attribute many good ideas.

v
Table of Contents

Dedication ii
Acknowledgements iii
List of Tables viii
List of Figures ix
Abstract xii
Chapter 1: Introduction
1.1 Anatomy and Physiology 1
1.1.1 Structure and Visual Processing of Retina 1
1.1.2 Central Visual Pathway 4
1.2 Principle of Neurostimulation 12
1.2.1 Electrical Activation of Neural Tissue 12
1.2.2 Cathodic and Anodic Stimulation 14
1.2.3 Monopolar and Bipolar configuration 15
1.2.4 Current-controlled Versus Voltage-controlled 16
1.2.5 Strength-Duration Relationship 17
1.2.6 Selective Stimulation of Cells 21
1.2.7 Safety of Neurostimulation 21
1.2.8 General Remarks 22
1.3 History of Neural Prostheses 23
1.3.1 Invention in Sixteeth Century 23
1.4 Outer Retinal Diseases 24
1.4.1 Aged Related Macular Degeneration and Retinitis Pigmentosa 24
1.4.2 Potential Treatments 24
1.4.3 Approach to Visual Prostheses 26
1.5 Remodeling in Retinal Degeneration 29
1.6 Animal Models of Retinal Degeneration 30
1.7 Relevant Studies of Electrical Stimulation in Degenerate Retina 31
vi
1.7.1 Human Study 31
1.7.2 Animal Study 33
1.8 Relevant Studies of Retinal Ganglion Cell Densities in Degenerate Retina 34
1.8.1 Human Study 34
1.8.2 Animal Study 35
1.9 Goals of Present Study 36
Chapter 2: Device-Tissue Interface
2.1 Dynamics Interaction at the Interface 37
2.1.1 Recording Electrodes 37
2.1.2 Stimulating Electrodes 40
2.2 Device-tissue Interaction 43
2.2.1 Equivalent Circuit Model 43
2.2.2 Cyclic Voltammetry 45
2.2.3 Impedance Spectroscopy 47
2.3 Assessment of Electrode-Retinal Proximity 50
2.3.1 Rationale 50
2.3.2 Materials and Methods 51
2.3.3 Results 55
2.3.4 Conclusions and Discussions 59
2.4 Tissue Damage at the Interface 60
Chapter 3: Electrical Excitation of Degenerate Retina
3.1 Background 62
3.2 Materials and Methods 63
3.2.1 Animals 63
3.2.2 Surgical Procedures 64
3.2.3 Insertion of Stimulating Electrode 65
3.2.4 Impedance Testing 66
3.2.5 Recording Electrode Positioning 66
3.2.6 Electrical Stimulation 68
3.2.7 Visual Stimulation 69
3.2.8 Experimental Protocol 69
3.2.9 Data Analysis 70
3.3 Results 70
3.3.1 Spatial Characteristics of Collicular Response 70
3.3.2 Characteristics of Retinal-driven Collicular Response to Light
and Electrical Stimulations 71
3.4 Conclusions and Discussions 79
vii
Chapter 4: Retinal Ganglion Cell Density of Degenerate Retina
4.1 Background 86
4.1.1 Motivation 86
4.1.2 SMI-32 Antibody 87
4.2 Materials and Methods 88
4.2.1 Tissue Preparation 88
4.2.2 Immunocytochemistry 89
4.2.3 Experiment Protocol 90
4.2.4 Data Analysis 90
4.3 Results 91
4.3.1 Distribution of SMI-32 Immuno-Reactive Cells 91
4.4 Conclusions and Discussions 99
Chapter 5: Spatial Properties of Electrical Excitation
5.1 Background 100
5.2 Materials and Methods 101
5.2.1 Animal 101
5.2.2 Visual Stimulation 102
5.2.3 Experiment Protocol 102
5.2.4 Data Analysis 104
5.3 Results 104
5.3.1 Focal Activation by Light and Electrical Stimulus Separately 104
5.3.2 Focal Activation by Light and Electrical Stimulus Simultaneously 105
5.3.3 Spatial Spread Increases with Increased Stimulus Currents 108
5.4 Conclusions and Discussions 109
Chapter 6: General Conclusions
6.1 Key Findings 111
6.2 Future Directions 112
6.2.1 Charge Densities as a Function of Small Electrodes 113
6.2.2 Investigation on Safety Limits of Short Pulses 114
6.2.3 Remodeling in Late Degenerate Stage and High Visual Pathway 114
6.3 Proposed Model 115
6.4 Implications for High Resolution Retinal Prosthesis 121
Bibliography 122
Appendix 134
viii
List of Tables
Table 2.1 Charge Density for Different Electrodes Positions 59
Table 3.1 Spearman Correlation Coefficients for Normal Retinas 76
Table 3.2 Spearman Correlation Coefficients for Degenerate Retinas 76

ix
List of Figures
Figure 1.1 Anatomy of Retina 3
Figure 1.2 Projection of Ganglion Cell Axons to High Visual Center 6
Figure 1.3 Rat Brain Atlas 9
Figure 1.4 Superior Colliculus Cells 10
Figure 1.5 Rat Retinotopic Map 11
Figure 1.6 Electrical Cable Model 14
Figure 1.7 Threshold as a Function of Electrode Orientation 16
Figure 1.8 Strength-Duration Curves 20
Figure 2.1 Device-Tissue Interface 44
Figure 2.2 Cyclic Voltammogram 46
Figure 2.3 Randles Circuit and Bode Plots 49
Figure 2.4 Block Diagram of Gamry Potentiostat 53
Figure 2.5 Impedance Measurement and Retina Histology (Group 1) 56
Figure 2.6 Impedance Measurement and Retina Histology (Group 2) 57
x
Figure 2.7 Threshold and Impedance as a Function of Distance 58
Figure 3.1 Examples of Spikes and Juxtazonal Potentials 68
Figure 3.2 Representative Spatial Maps for Normal and Degenerate Rats 72
Figure 3.3 Light-Evoked and Electrically-Elicited Responses 74
Figure 3.4 Strength-Duration Curves for Normal and Degenerate Retinas 75
Figure 3.5 Spearman Correlation between Threshold and Age 76
Figure 3.6 Strength-Duration Curves for Normal and Degenerate Retinas 77
Figure 3.7 Strength-Duration Curves for Spikes and Juxtazonal Potentials 78
Figure 3.8 Charge Density Curves for Normal and Degenerate Retinas 80
Figure 3.9 Rheobases and Chronaxies for Normal and Degenerate Retinas 80
Figure 4.1 Experiment Protocol for Vertical Section Immunocytochemistry 92
Figure 4.2 Experiment Protocol for Whole Mount Immunocytochemistry 93
Figure 4.3 SMI-32 Immuno-Reactive Ganglion Cells 96
Figure 4.4 Reduction of mean density of old degenerate retina 97
Figure 4.5 High-power view of normal and degenerate retina 98
xi
Figure 5.1 Experiment Protocol for Spatial Spread Measurement 103
Figure 5.2 Location of Collicular Responses to Light Stimuli 106
Figure 5.3 Collicular Responses to Light and Electrical Stimuli 107
Figure 5.4 Spikes of Light-Evoked and Electrically-Elicited Responses 108
Figure 5.5 Result of Spatial Spread as a Function of Currents 109
Figure 6.1 Proposed Weibull Curves of Probability of Firing as a Function of
Current Applied 118
Figure 6.2 Weibull Curves for Control, rd P500 and rd P700 groups 119
Figure 6.3 Relationship between Probability of Firing, Number of Inputs and
Current Thresholds 120
Figure A.1 Retinal Pigment Epithelium Layers (H & E Staining) 136
Figure A.2 Electron Micrographs for Melanin Granules 137

xii
Abstract
Retinitis Pigmentosa,(RP) is a blinding disease characterized by massive and
progressive reductions in the population of photoreceptor cells, result in losing night
vision, followed severely loss of peripheral vision and often total blindness. No effective
treatment is yet available for RP. Electrical stimulation of the retina through a
bioelectronic implant, replaces some of the lost function of the degenerated neurons,
allowing test subjects with experimental implants to perform simple visual tasks. Thus,
retinal implants have potential to provide an effective means of restoring vision to RP
patients.
Response thresholds and the loss of retinal neurons have been studied separately
in clinical trials and in pre-clinical studies, yet no study has simultaneously studied the
structure and function of the degenerated retina. Only a few of the threshold studies,
using animal model induced retinal degeneration, have performed histological analysis of
the same retina, and this examination has been minimal. Safety studies of electrical
stimulation of the retina are also limited. Safety is a concern for a high-resolution retinal
prosthesis, since a smaller electrode may use a high charge density to elicit a response.
This thesis investigated the response both the electrical stimulation and the
ganglion cell densities at different ages of in animal model of retinal degeneration. Acute
stimulation was carried out by electrically stimulating the retina and recording responses
xiii
from superior colliculus simultaneously. This research presents the first quantitative
measurement of electrophysiological and morphometric properties in degenerate retina. It
also represents the only known study to examine the response threshold and ganglion cell
density at multiple time-points following photoreceptor degeneration, and to note the safe
charge density with a limited pulse range using a 75-µm electrode. This thesis also
provided a means of investigating the spatial properties of electrical stimulation in a
normal animal model. Finally, this thesis proposed a model explaining the increased
threshold noted in aged degenerate retina with reduced cell number.
The ultimate goal of this work is to advance the understanding of electrical
stimulation in degenerate retina using small electrode and thus enable the transition from
low resolution to high resolution retinal prostheses. This research will hopefully lay the
groundwork for the successful development of high resolution retinal prostheses to
enable individuals suffered with RP to perform important visually-guided tasks, such as
navigation, reading, and facial recognition.

1

Chapter 1: Introduction
The work presented in this thesis contributes to the development of an epi-retinal
prosthesis, which is a neural engineering approach to partially restore vision lost due to
retinal diseases. This chapter provides necessary background for understanding the
complexity of the natural visual system as well as the reasons behind the fields of study
chosen. The following sections will focus on the anatomy and physiology of the retina
and the central visual center in human and rodent model, the principle of
neurostimulation, the retinal diseases we are targeting with a review of current treatments,
the history of neural prostheses, the relevant studies in this approach and the goals of the
work presented.
1.1 Anatomy and Physiology
1.1.1 Structure and Visual Processing of the Retina
Human Retina
Retina is an approachable part of the brain hence it is an attractive tissue for those
investigating how neurons carry and transmit information, how nerve cells receive stimuli,
and how synaptic connectivity interacts between different layers. Human retina is a
laminar structure which consists of ten anatomical layers, as shown in Figure 1.1:
pigment epithelium, photoreceptor cells (outer and inner segments of cone and rod
photoreceptors), outer limiting membrane, outer nuclear layer (cell bodies of cones and
rods), outer plexiform layer (cone and rod axons, horizontal cell dendrites, and bipolar
2

cell dendrites), inner nuclear layer (nuclei of horizontal cells, bipolar cells, and amacrine
cells), inner plexiform layer (axons of bipolar cells and amacrine cells, dendrites of
ganglion cells), ganglion cells layer, nerve fiber layer (axons from ganglion cells
traversing the retina to leave the eye at the optic disc), and internal limiting membrane
(separate the retina from the vitreous). Ambient light enters the cornea and the lens then
passes through the retina layers and reaches the visual receptors. Phototransduction takes
places in the photoreceptor cell layer by which light energy is transduced to
neurochemical energy. The resulting change in membrane potential propagates as a
graded potential in the inner nuclear layer and all-or-none action potential in the ganglion
cell layer. Intensity information in the ganglion cells is coded by rate or number of action
potentials. Visual information processed by the retina is then transmitted to higher visual
centers through the optic nerve, which is composed of ganglion cell axons
(approximately 1.1 million in number). There are approximately 120 million
photoreceptor cells within the retina. This represents an overall photoreceptor : fibre
convergence ratio of 110:1 but the ratio varies across the retina from centrally to
peripherally due to the presence of the central-to-peripheral rods and cones density
gradient in human retina. The human ganglion cell densities, within the central retina,
reach 32,000-38,000 cells/mm
2
. The total number of human ganglion cell ranges from
0.7 to 1.5 million (Curcio and Allen 1990).
3

Figure 1.1 3-D block of vertical section of retina.
Rat Retina
Ambient light reaches the visual receptors in rat in the same way as human. Both
human and rodent consist of rod majority photoreceptor cells. In human, rods outnumber
cones by a factor of ~20:1 (Curcio, Sloan et al. 1990). In mice, rods outnumber cones by
a factor of ~100:1 (Rohrer and Crouch 2006). Unlike human, rats have dichromatic
vision (blue and green cones) compared to trichromatic vision (blue, green and red cones)
in human. There are approximately 30 million photoreceptor cells and about 100,000 –
120,000 ganglion cells within the rat retina. This represents an overall photoreceptor :
fibre convergence ratio of ~ 300:1 (Mayhew and Astle 1997; Sefton A. J. 2004). This
higher ratio compared with human results in scotopic (low light) vision in rodent with
4

increase sensitivity at the expense of acuity. Ganglion cells are relatively evenly
distributed across the rat retina, with the variation from the highest to the lowest density
being 5:1 – from only 3000 to 600 cells per mm. Dendritic trees and therefore the sizes of
the receptive field centers of ganglion cells located in the area of highest density (area
centralis) are not significantly different from their counterparts located peripherally, in
the areas of lowest density (Sefton A. J. 2004).
Despite the above discrepancies between rodent and human retina, the visual
information is propagated through the ten neuronal layers to the higher visual center
through the optic nerve. Despite the fact that cone-rich fovea is only present in primate
retina, much of the circuitry of rodent retinas is virtually identical to peripheral primate
retina, making analysis of the remodeling in rodent model relative to the human condition
(Euler and Wassle 1995). Therefore, the retina and optic nerve of the rat have provided
widely used models for studies of central neural damage and degeneration due to the
availability of disease model and the cost.
1.1.2 Central Visual Pathway
Human Central Visual Pathway
The visual field in human has both binocular and monocular zones. Light from the
binocular zone strikes the retina in both eyes, whereas light from the monocular zone
strikes the retina only in the eye on the same side. The temporal and nasal hemiretinas are
defined with respect to the fovea, the region in the center of the retina with highest acuity.
Axons from the ganglion cells in the retina extend through the optic disc and, at the optic
5

chiasm, the fibers from the nasal half of each retina cross to the opposite side
(contralateral projection) of the brain while the axons from ganglion cells in the temporal
hemiretinas do not cross (ipsilateral projection).
The axons of all retinal ganglion cells project to three major subcortical targets:
the pretectum, the superior colliculus, and the lateral geniculate nucleus (Figure 1.2). The
pretectum of the midbrain controls pupillary reflexes. The superior colliculus controls
saccadic eye movements. The lateral geniculate nucleus is the main terminal for input to
the visual cortex (Kandel 1991). The retinal ganglion cells in the centrally located fovea
are more densely packed than in the periphery due to the fact that cones centralized
within the fovea region achieve high visual acuity while neurons in the lateral geniculate
nucleus are fairly evenly distributed. The ratio of the area in the lateral geniculate nucleus
to the area in the retina representing one degree of the visual field is called the
magnification factor. In another words, a small area on the retina in or near the fovea
receives more space on the LGN and visual cortex than the same area of peripheral retina.
In the mammalian visual system, there is close correspondence between retinal ganglion
cell density and the decrease of the magnification factor with eccentricity due to the fact
that the dendritic field size of the retinal ganglion cell and hence the ganglion cell
coverage factor decreases with eccentricity (Wassle, Grunert et al. 1989; Wassle and
Boycott 1991). The magnification factor of human has been proposed to be 4 mm/degree
at 2 degree eccentricity (central region) and decline to 0.5 mm/degree at 25 degree
eccentricity (peripheral region) (Cowey and Rolls 1974).
6

Figure 1.2 Axons of retinal ganglion cells project to lateral geniculate neucleus and
visual cortex. (Adopted from Kandel 1991)

Rat Central Visual Pathway
Rats have eyes on either side of their head, which allows for a large field of vision
but less binocular vision. Ipsilateral pathway occupies roughly 5-10% of axons in the
optic nerve which contributes to the binocular representation of the visual field. A
majority (>90%) of axons in the optic nerve project contralaterally.
7

The axons of rat retinal ganglion cells project to three different cortical targets:
the dorsal lateral geniculate nucleus, the pretectum, and the superior colliculus. The
dorsal lateral geniculate nucleus relays cells to various visual nuclei in the brain. The
pretectum, receives about 13% of the retinal ganglion axons, controls papillary light
reflex. The superior colliculus plays an essential role in visuomotor behavior. It is
involved in detecting moving objects in the visual field, in directing attention to and
orienting towards those objects, and in generating voluntary and involuntary eye
movements, particularly saccades (Mize 1992).
Rat Superior Colliculus
The superior colliculus (SC), where virtually all ganglion cells project in rat, is
the homologue of optic tectum of other vertebrates. SC is located is an area with
coordinates around 3.5 mm rostral and 0.5-1.5 mm lateral relative to the interaural and
midline planes, respectively (Figure 1.3). SC is segregated into seven layers. The layers,
from the surface, are zonal or stratum zonale; superficial gray or stratum griseum
superficale (SGS); optic strata or stratum opticum; intermediate gray or stratum griseum
intermediale; intermediate white or stratum album intermediale; deep gray or stratum
griseum profundum; and deep white or stratum album profundum. The SGS and upper
optic layers are innervated by retinal axons and thus process visual information (Figure
1.4). A direct retinal projection also terminates in the intermediate gray layer. Other
layers receive inputs from other sensory systems – auditory and somatosensory. A recent
8

finding showed that there are two sublaminars (SGS1 and SGS2) in the SGS layer which
perform different functional tasks (Girman and Lund 2007).
The retinotopic projection is organized topographically in SC (Fig. 1.5). Unlike
human or other mammalian species, representation of the central visual field is not
substantially magnified in comparison with the peripheral field. The magnification factor
is 0.02 mm/degree along medial-lateral (ventral-dorsal) axis and 0.04 mm/degree along
the posterior-anterior (temporal-dorsal) axis (Siminoff, Schwassmann et al. 1966). The
magnification factor in human is higher, meaning a small degree in the retina projects to a
larger area in the cortical region in human than those in rat.
9

Figure 1.3 A: Coronal section and sagittal section of rat brain, showing the locations
of the brain landmarks, sigma and bregma. B: Hand-drawn picture of the sagittal section
of an adult rat brain, showing the location of the superior colliculus (3.5 mm x 3 mm).
10

Figure 1.4 A: SGS and upper optic layers are innervated by retinal axons. ZS denotes
the zonale layer. SGS denotes the superficial gray layer. OS denotes the optic layer. B:
Examples of superior colliculus cell types. Otherwise noted, scale bar = 15 µm. (Adopted
from Edward 2002)
11

Figure 1.5 Topographical arrangement of receptive fields mapped in the left visual
field on the contralateral right superior colliculus. The number represents the electrode
position corresponding to the projection on the superior colliculus. (Adopted from
Siminoff 1966)

12

1.2 Principle of Neurostimulation
Principles of neurostimulation are discussed in this section. Activation of neural tissue,
the effect of cathodic and anodic stimulation on excitation threshold, mono- and bipolar
stimulation, an excitation model and the choice of current controlled and voltage
controlled stimulation will be described. This session is largely adopted from book
chapters on Principle of Neurostimulation and Intracellular Electrophysiological
Techniques (Holsheimer 2003).
1.2.1 Electrical Activation of Neural Tissue
Physiological activation of neural tissue is well described by Nernst Potential and
Goldman’s equation. Neural tissue can be electrically stimulated both at a cellular level
by injecting current intracellularly, and at a multicellular level by injecting current in the
extracellular space, manipulating the kinetics of the voltage-gated channel.
The primary effect of an electrical stimulation pulse on a neuron is a change of its
transmembrane voltage, being either a depolarization or a hyperpolarisation. When the
axon membrane is depolarized up to its threshold voltage, according to its electrical
membrane properties, an action potential will be generated by the excitation mechanism
first described by Hodgkin and Huxley (Hodgkin and Huxley 1952). While an action
generated under normal physiological conditions (at the axon hillock of the cell body)
propagates orthodromically (away from the soma), a stimulation-induced action potential
propagates both orthodromically and antidromically (opposite to the normal, orthodromic
direction) along the fibre. A cable network, as shown in Figure 1.6, was proposed by
13

McNeal to calculate how nodal transmembrane voltages are affected by a stimulation-
induced extracellular field.
e
V represents the nodal field potential.
a
R represents the
intra-axonal resistance.
m
R represents the nodal membrane resistance and
m
C represents
the nodal membrane capacitance. When a negative electrode (cathode) is placed near to
the nerve fibre, the node closest to the cathode will have the most negative
e
V and it will
be depolarized the most, described by the activating function AF by using Eq. (1). AF
is the driving force of the change of nodal transmembrane voltages. For node n , the
value of
n
AF is calculated as the difference of two potential differences:
( ) ( ) ( )
1 , , 1 , 1 , , , 1 ,
2
+ − + −
+ − = − − − =
n e n e n e n e n e n e n e n
V V V V V V V AF (1)
Ranck (1975) observed that the threshold stimulus of nerve fibre excitation is
smallest in the vicinity of a cathode and rises with increasing distance. This observation
was explained by the theoretical approach of the activating function. The node closest to
the cathode will be excited first when the stimulation current is sufficiently high. As the
field potential gradients on both sides of this node get steeper (the field potential gradient
near the stimulation site is the steepest), AF will rise and the stimulus needed for
excitation will be reduced. The same situation (obtaining steeper field potential gradient)
also occurs when the nerve fibre gets closer to the cathode.
14

Figure 1.6 Schematic drawings of electrical cable model (A) and myelinated nerve
fibre (B). (Adopted from Holsheimer 2003)

1.2.2 Cathodic and Anodic Stimulation
There are different stimulation schemes, apart from cathodic excitation described
above, namely anodic excitation, cathodic block and anodic block. These are well known
from experimental research (Ranck 1975). In cathodic excitation, the cathodic current
will propagate in each direction from the excitation node. The electric field pattern results
in slight hyperpolarisation of the membrane on segments (virtual anodic) just lateral to
the area of depolarization, but under typical circumstances these hyperpolarized segments
do not stop action potential propagation. Cathodic block occurs when the virtual anodic
hyperpolarisations are sufficient to compensate for the depolarization induced by the
Node of Ranvier
15

action potential at the nodes between. Hence, action potential propogation will be
blocked by these ‘virtual anodic
1
’ nodes. Anodic excitation occurs when the anodic
current is so large that the ‘virtual cathodic’ depolarization on either side of the
hyperpolarisation will generate an action potential. These action potentials will propagate
in opposite directions, as in cathodic stimulation. Finally, anodic block occurs when an
anodic current is applied to a fibre propagating an action potential. A few nodes closest to
the anode are hyperpolarized and will block the propagation when the anodic current is
large enough.
It was reported that the anodic excitation is 3–7 times their cathodic excitation
threshold (Rijkhoff, Holsheimer et al. 1994). The cathodic block threshold is more than 8
times the cathodic excitation threshold (Ranck 1975). High thresholds for both anodic
excitation and anodic block as compared to cathodic excitation have also been reported
elsewhere (Wee, Leis et al. 2000; Wee 2001). Hence, cathodic excitation is commonly
used in neurostimulation due to its relatively low excitation threshold.
1.2.3 Monopolar and Bipolar Configuration
The injection of current requires the use of two electrodes, an active electrode and
a return electrode. These can be placed in a monopolar configuration where the return
electrode is far away and the current radiates out from the active electrode, or in a bipolar
configuration where the return electrode is relatively close to the active electrode and
current is steered towards the return electrode. In monopolar configuration, the current

1
Virtual anode and virtual cathode are the side effects of cathodic and anodic stimulation
respectively.
16

injected by the active electrode is distributed more or less evenly in all directions. The
threshold current is reduced when the return electrode gets closer to the active electrode
(bipolar configuration) and when the nerve fibre axis is parallel to the active-return
electrode axis, and has been shown empirically in Figure 1.7.

Figure 1.7 More current is needed if the electrodes have a transverse orientation to
the fibres versus a longitudinal orientation. Adopted from (Ranck 1975).

1.2.4 Current-controlled Versus Voltage-controlled
In the past, clinical implantable neurostimulation systems have generally been
voltage-controlled, such as cardiac pacemakers. At present, current-controlled stimulation
is commonly used. Current-control means that the output current I is kept constant, thus
creating a rectangular current pulse (current source). The related voltage V is
automatically adapted according to the load impedance Z. Current-controlled stimulation
17

is more regularly used than voltage-controlled stimulation because of the following
reasons. First, the kinetics of charge redistribution during phase change at the interface is
better controlled. It is essential for stimulation pulse to be charge-balanced so as to
minimize net charges left at the interface which may increase the electrode potential to
the point where harmful quantities of gaseous oxygen or hydrogen are produced
(bubbling). Second, a constant electric field is ensured. Excitability of the tissue depends
on the electric field applied. The electric field applied is directly related to the injected
current. The pulse amplitude need to activate neural activities with current-controlled
pulses is not influenced by the value of Z which allows the voltage excursion to stay
within the compliance voltage of the stimulator. However, in voltage-controlled
stimulation, the voltage needed is influenced by the value if Z and may thus vary over
time. The advantage of using a voltage-controlled stimulation is that it offers high power
efficiency, at the expense of safety due to lack of control over the injected charge into the
tissue (described above) (Simpson and Ghovanloo 2007). Voltage-controlled stimulation
is good provided that a blocking capacitor is used in the circuit, at the expense of space
efficiency. Recently, a voltage-controlled stimulation for safe neural stimulation has been
proposed (Schuettler, Franke et al. 2008).
1.2.5 Strength-duration Relationship
The relationship between stimulus amplitude and duration is known as strength-
duration relationship. Strength-duration relationship measures the sensitivity of neuronal
elements to stimulation. When the duration of the stimulus duration d increases, the
stimulus amplitude I required to elicit a response becomes less. This inverse, non-linear
18

relationship is shown in Figure 1.9. The shape of a strength-duration curve is generally
characterized by two parameters, the rheobase current b (mA) and the time constant c
(ms). This curve is described by the following equation, LaPicque’s equation (Lapicque
1907)
() d c b I / 1+ = (2)
The rheobase current b, the asymptote of the strength-duration curve, is the minimum
stimulus amplitude required to elicit a response with an infinitely long stimulation
duration (which is a theoretical concept). According to this equation, c = d when I = 2b,
which defines the chronaxie value c as the pulse duration at twice the rheobase current.
Rheobase current is defined by the coupling between the tissue and the stimulating
electrode. Increasing the distance from the stimulating electrode to the tissue will
increase the rheobase current and therefore increase the necessary stimulus amplitude for
eliciting a response at all pulse durations. Chronaxie is defined by the membrane
properties as

m m m
C R ⋅ = τ (3)
with Rm being the membrane resistance and Cm being the membrane capacitance of the
target neuron. With both sides of Eq. (2) are multiplied by d, the following equation,
Weiss’ Law (Weiss 1901), is obtained
() c d b d I + = ⋅ (4)
with d I ⋅ being the threshold charge (µC) required for stimulation. This linear charge-
duration curve is also shown in Figure 1.8. This curve shows that the charged needed for
19

a threshold pulse rises when d increases. The electrical energy required U equals to the
amount of charge d I ⋅ multiplied by the electrical potential r I ⋅ , where r represents the
tissue resistance. Solving equations (2) and (4) for the electrical energy, it can be shown
in Fig. 1.9, that the most electrically efficient stimulation duration d equal to the
chronaxie c of the neuron being activated.
20

Figure 1.8 Strength-duration curve for current (I), charge-duration curve for charge
(Q) and energy-duration curve for energy (U) are plotted in (A) linear and (B)
logarithmic scales. (Adopted from Geddes 2004)
21

1.2.6 Selective Stimulation of Cells
Electrical current pulses are used to activate neurons and replace lost visual
functions due to retinal diseases. The duration of the current pulse can be manipulated to
target specific neurons. Short current pulses have been demonstrated to target retinal
ganglion cells, the neurons closest to the electrode in epiretinal approach (Greenberg
1998; Fried, Hsueh et al. 2006; Sekirnjak, Hottowy et al. 2006) while long current pulses
target bipolar cells. When a neural system is stimulated, different cells respond in
different ways. In the studies described in subsequent chapters, a range of different pulse
durations will be tested to obtain strength-duration curve which describes the relationship
between pulse amplitude and pulse duration and sets as an important tool to evaluate the
effectiveness of neural stimulation.
1.2.7 Safety of Neurostimulation
As briefly mentioned in section 1.2.4, a charge-balanced waveform will minimize
the net charge accumulation on the electrode which may increase the presence of harmful
substances on the electrode. However, having a charge-balanced waveform alone does
not guarantee a safe stimulation. Two safety limits should be taken into accounts which
are neural damage limits and electrochemical limits.
Neural damage limits refer to the amount of charge, charge per phase and charge
density, injected to the neural tissue which induces damage under several hours of
continuous stimulation; including neuronal hyperactivity such as neuronal shrinkage and
pericellular halo by using neuronal markers (McCreery, Agnew et al. 1990). Subtle
morphological changes were observed in distal region from the stimulation site in a retina
22

study under different stimulation conditions (Ray, Colodetti et al. 2009), but only damage
due to mechanical pressure was observed near the stimulation site (Colodetti, Weiland et
al. 2007). Apart from histological detectable injury, a transient effect evident as a decline
in excitability of neural tissue under continuous pulsing was observed. It was shown that
more current was required to elicit an equivalent response after several hours of
continuous pulsing. This effect could be completely reversed with rest, but typically
required hours to days to reach original sensitivity. (McCreery, Yuen et al. 1997).
Neurons beneath the stimulating electrode could be damaged by electrochemical
reactions as well when the electrode is driven such that the electrode potential exceeds a
value where irreversible reactions occur. The mechanism of damage induced by
electrochemical reaction of the electrode depends on the ability of the electrode to store
or dissipate electrical charge without exceeding the water window, which is the potential
window outside of which significant bubble formation is evident at the interface. The
most commonly used material for electrical stimulation is platinum, which can safely
supply 0.1 to 0.4 mC/cm
2
of charge (Brummer and Turner 1977; Rose and Robblee 1990).
Materials with higher limits are iridium oxide (1 to 3 mC/cm
2
) (Agnew, Yuen et al. 1986;
Beebe and Rose 1988) and titanium nitride (0.6 to 0.9 mC/cm
2
) (Weiland, Anderson et al.
2002).
1.2.8 General Remarks
Electrical pulses are vital for the functioning of living things. Signaling is
accomplished in neuronal and non-neuronal cells by manipulation of electrical potential,
which has affects ranging from the metabolism of individual cell to human consciousness
23

derived from the synchronous or asynchronous activity of the brain. Hence, the biological
mechanism responsible for electrical application may also be activated in a controlled
fashion for beneficial medical purposes.
1.3 History of Neural Prostheses
1.3.1 Invention in the Sixteenth Century
A neural implant is an electronic prosthetic device intended for people being
impaired from neuronal disorders to regain its functionality. The idea of an artificial
device to replace lost function is almost as old as recorded history. One of the first known
reports of a “prosthetic” device was a mechatronic system and described as early as 484
B.C. where an injured Persian soldier was told to cut off his foot and used a wooden
replacement for an escape. A prosthetic leg was dated approximately 300 B.C. in a tomb
in Italy. In the sixteenth century, iron legs were being created for soldiers by a French
army surgeon. In the seventeenth century, Charles LeRoy wrapped a brass wire around
the orbit of a man who was blind from cataract and the patient saw “flames passing
rapidly downwards.” In the late eighteenth century, Luigi Galvani and Allesandro Volta
conducted pioneering experiments in demonstrating that it was possible to stimulate
neuronal tissue by means of electricity. Since then, many devices have been developed to
restore neurological functions. Heart and brain pace makers, deep brain stimulators and
cochlear implants are the examples of successful medical devices. Retinal implants,
based on the same principle, targets people who develop blindness due to loss of
photoreceptors from outer retinal diseases such as retinitis pigmentosa (RP) and age-
24

related macular degeneration (AMD). Despite the fact that the retinal prosthesis has not
yet advanced to a medical product, early clinical trials (Gekeler, Messias et al. 2006;
Yanai, Weiland et al. 2007) showed promising results. Additionally, many other labs
worldwide are conducting experiments to address other aspects of retinal prostheses
(Eckhorn, Wilms et al. 2006; Pardue, Phillips et al. 2006; Shah, Montezuma et al. 2006).
1.4 Outer Retinal Diseases and Potential Treatments
1.4.1 Aged Related Macular Degeneration and Retinitis Pigmentosa
Two common outer retinal diseases are retinitis pigmentosa and age-related
macular degeneration. The worldwide incidence of RP is about 1 in 4000 for a total of
more than 1 million affected individuals (Hartong, Berson et al. 2006). It is a hereditary
disorder characterized by dysfunction and loss of photoreceptors and retinal pigment
epithelium. There are a wide variety of genes which cause RP but the rhodopsin gene
(RHO) leads to a large proportion (about 25%) of dominant RP. It results in massive and
progressive decrease in the population of rod and/or cone photoreceptor cells. AMD
affects between 20 and 25 million people worldwide (Chopdar, Chakravarthy et al. 2003),
primarily individuals over 65.
1.4.2 Potential Treatments
No generally effective treatment is yet available for Retinitis Pigmentosa and dry
AMD. Treatment is available for wet AMD but it is expensive and it needs repeated
intraocular injections. RP treatments are divided into two categories 1) treatments used to
prolong the life of the photoreceptor cell when some photoreceptors remain alive and
25

functional, 2) treatments used when photoreceptors are degenerated and their function
needs to be replaced.
For 1), there are two major possible therapies where patient yet has viable
photoreceptor cells in their retina. The first one is gene therapy. Gene therapy is the
replacement of a defective gene such that an important protein is again synthesized and
present in the cell. With this the photoreceptors function better and live longer. Gene
therapy treatment has been successful in animal models of RP. Clinical trials for gene
replacement in patients with Leber Congenital Amaurosis (LCA) have begun (Smith,
Bainbridge et al. 2009). However, this approach is limited to a particular gene defect and
the number of RP mutation is huge (Rivolta, Sharon et al. 2002) which suggests unique
therapy for each mutation. The second one is pharmaceutical therapy which is the use of
an agent that will prolong the life and function of a photoreceptor cell. A natural growth
factor, ciliary neurotrophic factor (CNTF), could delay photoreceptor cell degeneration in
an animal model (Bok 2005). The use of nutrition is the third treatment to prolong the life
of remaining photoreceptor cells (Berson, Rosner et al. 1993; Yokota, Shiojiri et al. 1997).
A recently proposed treatment regimen uses an antioxidant agent, which showed
reduction in oxidative damage in the retinal photoreceptor cells. (Sancho-Pelluz, Arango-
Gonzalez et al. 2008).
For 2), there are three possible treatments when photoreceptor cells are dead. The
first one is photoreceptor cell transplantation. Recent work in this area has shown a slight
increase in visual acuity in one patient. However, it still remains uncertain whether local
synaptic connectivity between the transplant and host has been established (Radtke,
26

Aramant et al. 2004). Stem cell transplantation has shown positive results in animal
models in RP where photoreceptors were preserved and visual function was stabilized
(Canola, Angenieux et al. 2007). However, safety issues on stem cell research are not
well documented so far. The second one is cell modification. Channelrhodopsin-2
expressed in ON bipolar cells conferred responses to light on retinal ganglion cells and
these responses led to an ability of the animal to respond to light behaviorally. However,
channelrhodpsin-2 requires stimulation with short-wavelength light at very high intensity
(Bi, Cui et al. 2006; Lagali, Balya et al. 2008). The third one is prosthesis. In recent years,
retinal prosthesis has become a promising approach for RP and AMD treatments and
several groups worldwide have conducted clinical trials. At present, there are totally four
clinical trials worldwide with different designs (DeMarco, Yarbrough et al. 2007;
Gerding, Benner et al. 2007; Yanai, Weiland et al. 2007; Besch, Sachs et al. 2008). The
device is designed to send electrical signals to bypass defective or dead photoreceptors
and stimulate remaining intact cells of the retina. The image data from the camera are
processed by a video signal-processing unit that converts the image data into stimulation
data. These data and power are then transmitted wirelessly through a telemetry circuitry
and decoded. Patterns of stimulus generated from the chip are then transmitted through a
microelectrode array to the retina and reproduce the visual image in the brain.
1.4.3 Approach to Visual Prostheses
In the development of visual prostheses, there are five different approaches, in
which three of them are retinal implant, one is optic nerve implant and the remaining one
is cortical implant.
27

Cortical implant research started in 1960s. It can potentially aid individual with
blindness with a wider variety of diseases, including dysfunction of optic nerve and
congenital blindness, since it does not require an intact eye or retina. However, given the
amount of signal processing which occurs in retina and lateral geniculate and divergence
of information to other brain areas, a visual cortex implant may require significant signal
processing. Another disadvantage of cortical prostheses is the difficulty in positioning the
electrodes in the primary visual cortex. Optic nerve implants were pioneered by Veraart
(Veraart, Wanet-Defalque et al. 2003) and is currently being developed by a group in
Shanghai JiaoTong University in China. The current implant consists of four electrodes.
However, it is relatively difficult to access the retinotopic map and hence the spatial
feature of this approach is questionable.
Retinal implants have many advantages leading to many groups investigating this
approach. It is approachable surgically and permits relatively easy access to the target
nerve cells. The four groups worldwide have performed clinical trials of experimental
retinal implants. In epiretinal implants, the electrode array is placed near to the inner
surface of the retina (by the ganglion cells). Epiretinal implants have neurostimulation
electronics, but the image acquisition is typically done via a the wearable portion of the
device. This does allow for easy upgrades to cameras and image processing without
requiring subsequent surgery. Electronics operating in the body will generate heat, but the
vitreous acts as a heat sink for thermal dissipation. One technical challenge of epiretinal
implants is to improve the methods of attachment of the device to the retina. Since
epiretinal implants stimulate at the output side of the retina, this approach may also
28

requires sophisticated image processing to account for visual algorithms processed by the
retina. Subretinal implants use photic sensors in the subretinal space (between the RPE
and retina) to detect light and produce an electrical stimulus at an electrode located close
to the light detector. Subretinal implants utilizes photodetectors which simply replace the
function of photoreceptors by producing electric current in response to ambient light. The
photocurrent produced is inadequate to stimulate the retina, so an external power source
is needed. The close proximity of this implant to the next survival retinal neurons enables
a secure placement and coupling between the electrode/sensor to the neurons. The
placement of the cable, traversing the vascularized choriocapillaris may result in
subretinal hemorrhage and cause local retinal detachment. Another big challenge of this
approach is the development of the thick glial seal in the subretinal space due to
remodeling of the retinal neurons after the death of the photoreceptors, which may
obscure the accessibility of the electrical signals to neurons. In suprachoroidal-
transretinal (STS) implants, the electrode array is placed between the choroid and the
sclera which requires less complicated surgery. Electrodes used in this approach are less
invasive to the retina and are relatively easy to remove. Higher threshold current may be
required because electrodes are further away from the target cells in this approach. The
current spreading which results from the enlarged receptive field of the electrode limits
the resolution in the STS approach.

29

1.5 Remodeling in Retinal Degeneration
Neuroplasticity often refers to the changes that occur in the organization of the CNS
as a result of experience or a process of recovery in brain injury. Retinal remodeling is a
subset of neuroplasticity when the CNS undergoes deafferentation and trauma, leading to
neuronal cell death, neuronal migration, elaboration of new neuritis, rewiring of circuits
and the evolution of glial seal.
Retinal degeneration in the mammalian retina generally progress through three
phases (Jones, Watt et al. 2003; Marc and Jones 2003; Marc, Jones et al. 2003; Jones and
Marc 2005). In phase 1, there is loss of photoreceptor cells due to gene mutation in
rhodopsin for autosomal dominant RP. In phase 2, a sequence of remodeling in the inner
nuclear layers follows as death of rods and cones progresses. Bipolar and horizontal cells
lose their presynaptic input and retract most of their dendritic tree. Muller cells, providing
support and nutrition to the retina, forms a dense fibrotic layer in the remnant subretinal
space. In phase 3, the inner nuclear layer and ganglion cell layer become depleted. Cell
translocations and neuronal loss continues and becomes evident in late phase 3.
Apart from the intense rewiring in retinal degeneration, the surviving cells appear to
be healthy and active. These surviving cells act as a communication channel for a retinal
prosthesis. Morphometric studies of enucleated eyes from patients with AMD or RP have
demonstrated partial preservation of inner retinal cells. In AMD, cell loss was most
pronounced in the outer nuclear layer, with a 76.9% reduction in the number of cells.
Ganglion cells were reduced by 30.7%. However, cell count in the inner nuclear layer
30

was not significantly different from the control group (Kim, Sadda et al. 2002; Kim,
Sadda et al. 2002). In RP, cell loss occurs in all layers. Reductions of cells are different in
the macula and the extramacular regions. In the macula, 78-88% of the inner nuclear cells
and 30-48% of ganglion cells are retained, whereas in the extramacular regions, 40% of
the inner nuclear cells and 20-30% of ganglion cells are retained (Stone, Barlow et al.
1992; Santos, Humayun et al. 1997; Humayun, Prince et al. 1999).
1.6 Animal Models of Retinal Degeneration
A study of in vivo animal model in retinal prosthesis research has recently been reviewed
(Bertschinger, Beknazar et al. 2008). Genetic mutations that more closely mimic the
human RP are found in mouse, rat, cat and dog. Rats and mice are commonly used and
their strains with RP retinal degeneration are commercially available. S334ter and P23H
rats are transgenic models of retinal degeneration developed to express a rhodopsin
mutation similar to that found in human retinitis pigmentosa patients which accounts for
more than 25% of the human RP cases. Fast degenerating rd1 mutant mouse model and
slow degenerating rd10 mouse model are natural models of recessive RP caused by
mutation of the beta-subunit of rod photoreceptor cGMP phosphodiesterase. They are
good animal models to study the threshold and charge density as well as safety concern in
a retina with degeneration. Abyssinian cat is an animal model that closely mimics human
retinitis pigmentosa (Narfstrom 1999; Menotti-Raymond, David et al. 2007). Abyssinian
cat has normal retina until two to three years, after which late-onset hereditary rod-cone
degeneration occurs, leaving the animal blind. It has limited use in the retinal prosthesis
31

research because of the cost to keep a research animal for two to three years while
degeneration occurs (Pardue, Stubbs et al. 2001; Narstrom 2007). RCD1 dog has an
early-onset autosomal recessive disease where photoreceptor cell degeneration begins by
post natal day 14 (Aguirre and Rubin 1975). The disease has rapid progression, leaving
the animal blind by one year of age, which makes it a good animal model for this
research (Guven, Weiland et al. 2005). Rabbit model has been used to study electrical
stimulation by injecting iodoacetic acid in rabbit’s retina to destroy the outer retina,
leaving the inner retina intact (Humayun, Propst et al. 1994). Recently, a transgenic
rabbit model of retinal degeneration was generated and characterized (Kondo, Sakai et al.
2009).
1.7 Relevant Studies of Electrical Stimulation in Remodeled Retina
Many studies have documented that electrical stimulation of neurons in the visual
pathway elicits responses similar to light driven responses. Most of these studes have
been performed in healthy retina. Fewer studies have looked at electrical stimulation in
remodeled retina. Studying the response of the surviving cells in the remodeled retina to
electrical stimulation is important because it directly relates to the efficacy of the chosen
approach. This section will discuss experiments conducted in degenerate retina and the
questions remain unanswered.
1.7.1 Human Study
Studies of electrical stimulation of degenerated retina started in 1960s. One of the
early experiments did not involve an implant, but used an external electrode to stimulate
32

the eye in an acute study (Potts, Inoue et al. 1968; Potts and Inoue 1969; Potts and Inoue
1970; Kato, Saito et al. 1983). It has been demonstrated that electrical stimulation
through a corneal electrode produced an electrically evoked response (EER) by recording
from scalp electrodes. Using this method, subjects with advanced photoreceptor
degeneration reported that they experienced visual sensation of streaks or dots of light
(phosphenes). Investigators also reported that there were discrepancies between the EER
and the visually evoked response (VER) obtained in some subjects with advance
photoreceptor degeneration, suggesting that the cells most sensitive to electrical
stimulation (target cells) are different from the photoreceptors, which obviously are the
most sensitive to light.
More recent clinical studies have shown that individuals suffering from RP could
detect motion and perform simple visual tasks under electrical stimulation through
implanted microelectrode array. Three RP subjects, with light or no light perception
2
,
used a chronically implanted epiretinal prostheses to perform simple visual tasks, such as
counting objects, object recognition and differentiating directions (Yanai, Weiland et al.
2007). Acute (short-term, less than 3 hours) implants have been used as well. In one
study, five RP subjects perceived visual percepts repeatedly and consistently, but the

2
Low-vision is commonly characterized through quality means, once blind people can no
longer read an eye chart, yet retain the ability to perceive light. Clinicians rate vision as
count fingers, hand motion, light perception, and no light perception (in order of visual
ability).
33

percepts matched the stimulation pattern of the microelectrode only 67% of the time
(Rizzo, Wyatt et al. 2003).
Several studies have reported lower threshold values in healthier retinas compared
to diseased retinas in human subjects. Using large electrodes (400-µm), normal-sighted
subjects had thresholds at charge densities 0.08 mC/cm
2
, whereas the threshold was 2.8
mC/cm
2
in blind patients (Rizzo, Wyatt et al. 2003; Yanai, Lakhanpal et al. 2003).
1.7.2 Animal Study
Likewise, animal studies have confirmed lower threshold values in healthier
retinas. It was found that the threshold charge densities in 8-16 week-old rd1 mouse
retina were substantially higher than that in the normal mouse retina (Suzuki, Humayun
et al. 2004; O'Hearn, Sadda et al. 2006). Retinal stimulations were performed with an
array of platinum electrode of 125-µm in diameter. Yet, these studies only investigated
the threshold charge densities at one stage of retinal degeneration. The most recent in
vitro study, involving multiple ages of rd1 mouse, reported that the thresholds of retinal
ganglion cells were consistently higher than the wild-type mouse. The investigators
proposed that the changes happening in the remodeled retina would not be the
contributing factor for the elevated thresholds (Jensen and Rizzo 2008). Yet, the
investigators agreed that the age group studied only up to P186 and the electrically
evoked responses and thresholds of RGCs will be substantially affected in much older
rd1 mice. There is yet any study in animal with RP at older age.
The contribution to the elevated EER or perceptual thresholds in human and animal
with RP is not well known. One possible cause is the remodeling of the neural retina. In
34

human with late stage RP, reduction of cells occurs in all retinal layers. Thus, it is
reasonable to speculate the remodeling of the neural retina, particularly on the cell
reduction, plays a role in the elevated electrically evoked responses or thresholds, at the
late stage RP.
1.8 Relevant Studies of Cell Densities in Remodeled Retina Due to
Retinitis Pigmentosa
The studies of survival cells in the retina with AMD and RP are few. In human, the
studies could only be done in post-mortem eyes, which limits the ability to precisely age-
match the eyes. Also, each human case will be affected by many other factors. Therefore,
transgenic animal models of AMD and RP are important alternatives to human data, since
these can be studies in greater number under controlled experimental conditions. Human
studies in RP retina showed consistent decline in all retinal cell counts throughout all the
reported morphometric analysis. However, results from ganglion cell counts studies in
the animal model with RP are inconsistent.
1.8.1 Human Study
In previous sections, it was mentioned that in morphometric analysis, there were
cell reductions in ONL and GCL in AMD retina, but the cell count in INL is similar to
that in the control group. Ganglion cells were reduced by 30.7% in retina with AMD,
whereas 30-48% of ganglion cells were retained in retina with RP in the macular region.
The ganglion cells retained in the extramacular region in retina with RP are even less, 20-
30%. The loss of ganglion cells paralleled the loss of photoreceptors. The loss of
35

ganglion cells was suggested to result from transneuronal degeneration (Stone, Barlow et
al. 1992). Proposed mechanisms for ganglion cell death include migration of RPE cells to
perivascular sites in the inner retina, decrease in their blood supply (Li, Possin et al.
1995).
1.8.2 Animal Study
Animal study of ganglion cell loss in inherited retinal dystrophy started in early
70s. The study demonstrated that in mutant mice lacking visual receptors (rd1 strain), the
retinal ganglion cells are decreased in size and number (Grafstein, Murray et al. 1972).
The investigators also showed that aside from the changes in cell number and size, the
morphology of the inner retinal layers, including the ganglion cells, is essentially normal
on an electron microscope level which is also confirmed in early studies (Karli 1952;
Caley, Johnson et al. 1972). One recent study confirmed this early finding that retinal
ganglion cells maintain dendritic morphology in rd1 mouse model of photoreceptor
degeneration (Mazzoni, Novelli et al. 2008). However, this study demonstrated survival
of retinal ganglion cells in rd1 mouse up to P270, which is conflicting with the human
studies and early animal studies about the decline in retinal ganglion cell count in late RP
retinas. Another study showed that RGCs in the adult RD retina is functionally stable in
terms of rhythmic spike activity in spite of the loss of the photoreceptor cells (Margolis,
Newkirk et al. 2008).
There are two recent studies showing the decline in retinal ganglion cell counts in
animal model with RP (Milam, Li et al. 1998; Marc and Jones 2003). The inconsistency
in the survival of retinal ganglion cells counts in animal models reveals that the
36

importance of choosing the age of the RP retina and the animal model of genetic
pathology which resembles the human late-RP condition, to obtain a reliable
investigation in terms of efficacy and safety issues in the development of a retinal
prosthesis.
1.9 Goals of Present Study
The present study focuses on examining the response threshold in different stages of
retinal degeneration by stimulating in the retina of the S334ter-line-3 rat model and
recording the bioelectric responses in the superior colliculus. This study is also designed
to examine the safe stimulation of a small platinum electrode (75-µm diameter) in a
degenerate retina, towards the effort in developing a high resolution retinal implant.
Subsequently, this study will determine the change of retinal ganglion cell density as a
function of different degenerate stages. Finally, this study will present a preliminary data
to examine the spatial properties of electrical activation, using a single electrode, which
sets as a basis for suggested future directions using a microelectrode array for assessing
spatial activation in degenerate retina.
37

Chapter 2: Device-Tissue Interface
Neural prostheses convey messages to the target tissue through a device-tissue interface.
Study of the device-tissue interface is wide, from nano-scale to milli-scale, including
kinetics between ions and electrons which determine the charge delivery and safety
performance, and biological insertion which leads to undesirable tissue response masking
the efficacy of the device. The integrity of device-tissue interface affects both stimulation
and recording schemes. In this chapter, we will describe the scope of fields that study the
device-tissue interface, and we will focus on nano-scaled effect of the interface under
electrical stimulation. Lastly, we will demonstrate the method we employ in our in vivo
experiment to maintain a stable device-tissue interface.
2.1 Dynamics Interaction at the Interface
A material, when inserted in an ionic solution (electrolyte), involves exchange of ions
and charges at the electrode-electrolyte interface. The mechanism of electric conductivity
in the body involves ions as charge carriers while electrons are the charge carriers in any
conductive element, such as metal. Thus, picking up bioelectric signals or delivering
current to the body, such as in the application of neural prostheses, involves interacting
with these ionic charge carriers (in electrolyte) and transducing ionic currents into electric
currents (in electrode). This transduction process is carried out by electrodes that consist
of electrical conductors in contact with the aqueous ionic solutions of the body. The
dynamic interaction between electrons in the electrodes and the ions in the body can
38

greatly affect the performance of the device. It plays a significant role in both the efficacy
and the safety of neuro-stimulation electrode, as well in the chronic recording electrode
application.
At the interface between an electrode and an ionic solution, either in recording or
stimulation paradigms, redox (oxidation-reduction) reactions need to occur for a charge
to be transferred across the interface. These reactions can be represented in general by the
following equations:

− +
+ → ne C C
n
(Eq. 2.1)

− −
+ → me A A
m
(Eq. 2.2)
Where n is the valence of cation material C, and m is the valence of anion material, A.
Oxidation occurs when an electron is given up and the cation goes into solution as
positively charged ions, Equation 2.1. Reduction occurs in the reverse direction, Equation
2.2. The most favourable operation of the electrodes occurs when charges move across
the interface reversibly, and do not produce toxic products that diffuse into the electrolyte
or degrade the electrode.
2.1.1 Recording electrodes
A recording microelectrode provides a means of detecting biopotentials within a
small region of the brain (Kipke, Pellinen et al. 2004). Upon immersion of the metal into
the electrolyte, thermodynamic forces bring the system to electrochemical equilibrium.
39

Depending on the metal and the electrolyte, metal molecules may oxidize to some degree
into the electrolyte according to the reaction. Non-polarizable metals are those that
undergo an oxidation process in which the metal atoms lose electrons and go off into the
solution as positively charged ions. Polarizable metals are those that do not readily
oxidize under normal physiological conditions, e.g. platinum and tungsten. In practice, no
electrode is perfectly polarizable or non-polarizable overall all voltages, but some
materials exhibit these properties over a limited range. In recording applications, the
voltage change is small, so platinum and tungsten electrodes act as polarizable in this
small voltage range. An equilibrium condition is reached between the metal and the
electrolyte when no current is allowed to flow. This results in an electrical potential, the
half-cell potential across the interface. At this equilibrium condition, a molecular layer of
water coats with metal, acting as a dielectric between the metal and the ionic solution. In
most cases, the recording sites for a microelectrode are selected to be a polarizable metal
to create a capacitive interface.
Degradation of action potential recording in chronically implanted
microelectrodes occurs when the working area of the electrode is encapsulated by
reactive tissue. Reactive tissues, include microglia, astrocytes, and extracelluar matrices,
form the encapsulation layers (Turner, Shain et al. 1999; Szarowski, Andersen et al. 2003;
Kim, Hitchcock et al. 2004). The early inflammatory response is due to insertion trauma
and a sustained response is due to micromotion and device biocompatibility. Some
intervention strategies have been proposed to limit the early response, such as
40

improvements to insertion techniques. Some groups have proposed a “rejuvenation”
method by applying repeated voltage biasing for a few seconds, the resistive tissue
impedance decreases and the extracellular unit activity resumes (Johnson, Otto et al. 2005;
Otto, Johnson et al. 2006). This method has been performed in sessions over a two-month
period and successfully demonstrated improvement of unit recording SNR.
2.1.2 Stimulating electrodes
A stimulating electrode is a neural interface where electron current flow takes
place in an abiotic conductor and subsequently converts to ionic current flow in tissue.
The conductor is typically metal or metal oxide, but recently conducting polymers and
carbon nanotubes have been implemented as stimulating electrode materials. There are
two mechanisms of charge transfer at the electrode-electrolyte interface, namely the
capacitive/non-Faradaic charge transfer and Faradaic charge transfer. When a current is
delivered to the electrodes, such as in the application of neural stimulation which forces
the electrode away from its equilibrium potential, a redistribution of charge occurs at the
interface.
In the case of capacitive charging (non-Faradaic reaction), the electrode-
electrolyte interface may be modeled as a simple electrical capacitor where only
redistribution of charge occurs. There is no transfer of electrons across the interface
provided the total amount of charge delivered is sufficiently small. This charge
redistribution results in an electric field in tissue and eventually to cell membrane
41

depolarization. Efficiency of charge transfer depends on the electrode material, which
will limit the applied voltage/current levels.
In the case of Faradaic charge transfer, charge is injected from the electrode to the
electrolyte through redox reactions (reduction and oxidation). There are two types of
Faradaic reactions, reversible and irreversible reactions. The degree of reversibility
depends on the relative rates of kinetics and mass transport. A Faradaic reaction with
very fast kinetics relative to the rate of mass transport is reversible. Reversible redox
reactions do not produce toxic products that diffuse into the body solution. A Faradaic
reaction with slow kinetics is irreversible because there is no effective storage of charge
near the electrode surface and the product has diffused away within the slow time frame
of the reaction.
For a stimulating electrode to be effective, it must deliver sufficient charge for the
cell membrane potential of the target neurons to exceed threshold for neuronal
depolarization. Electrode materials that have large capacity for safe charge delivery are
favourable candidates neural stimulation electrodes. The most commonly used material
for stimulating electrodes is platinum, which was proposed to have a limiting charge
density < 300 µC/cm
2
to ensure that no gassing occurred on the electrode surface. This
translates to 0.05 µC, of a 100 µA 0.5 ms pulse, for a 200-µm diameter electrode.
Platinum is desirable because it can transfer charge via reversible reactions. One example
of the reversible Faradaic reaction is so called “H-atom plating” at platinum.
42

H Pt H Pt − ⇔ +
+
(Eq. 2.3)
This process is proved to be safe and does not create new chemical species in the body
solution. This reaction is reversible by applying current in the opposite direction.
Capacitive electrodes are advantageous over the metal faradaic electrode. They
are more resistant to degradation in physiological solutions and are more inert to
electrical and chemical interactions with the body fluids at the interface. Capacitive
electrodes have a metal core coated with a non-conductive oxide surface layer. The oxide
insulates the conductive metal, creating a dielectric at the interface. Electrons moving
through the electrode cannot physically transverse the interface, but can attract or repel
ions in close proximity target tissue and initiate an action potential. However, the charge
storage capacities of capacitive electrodes are limited. Examples for representative
capacitor electrodes are titanium nitride (TiN) and tantalum pentoxide (Ta-Ta
2
O
5
).
Guyton & Hambrecht developed capacitor electrodes that can transfer 2.5 µC in each
pulse (Guyton and Hambrecht 1973). The charge storage for TiN was 0.87mC/cm
2
while
the charge storage for a Faradaic metal, iridium oxide (IrOx) was proposed to be 1-4
mC/cm2 (Brummer, Robblee et al. 1983; Weiland, Anderson et al. 2002). Another
attempt to increase charge storage is to increase the surface roughness, which is the
development of nanowires (carbon nanotubes) as neural probe electrodes (Whalen 2005;
Yoon, Deshpande et al. 2008).
43

In the current study, a commercial available electrode is used, composed of an
alloy of platinum and iridium electrode operating at a low voltage to inject charge in the
tissues. Platinum is a relatively soft material and thus it is alloyed with iridium to increase
the mechanical strength. In the later sections, the electrochemical analytical tools will be
discussed to measure the charge storage capacity and the impedance characteristics of
this stimulating electrode, as well as the electrical model of the interface and the method
to maintain a stable proximity between electrode and the tissue.
2.2 Device-tissue Interaction
2.2.1 Equivalent Circuit Model
The device-tissue interface can be modeled as a simple electrical capacitor, called
the double layer
dl
C . A redistribution of charge occurs in the solution during stimulation.
Suppose a negative potential (cathodic pulse) is applied to one electrode, immersed in a
body fluid solution, the metal electrode then accumulates an excess of negative charge.
Positive charge in solution will then be attracted towards the electrode and negative
charge in the solution will be repelled. Accordingly, a redistribution of ions occurs in the
bulk solution. Besides capacitive process, charge may also be injected from the electrode
to the body fluid by Faradaic process. During Faradaic process, electrons are transferred
at the interface through reduction and oxidation. This process is typically modeled as a
resistor. Hence, an equivalent electric circuit model is illustrated in Figure 2.1, based on
the occurrence of these two processes.
dl
C is the double layer capacitance representing
44

the ability of the electrode to cause charge flow in the body fluid without electron transfer,
Faradaic
Z is the Faradaic impedance, representing the Faradaic process of reduction and
oxidation where electron transfer occurs at the interface.
s
Z represents the solution
resistance that exists between two electrodes.
There are several methods of determining the characteristics of an electrochemical
interface. Parameters like charge storage capacity and load impedance can be determined.
These methods are well studied from the rich field of electrochemistry, which is the basis
of batteries, fuel cells, and corrosion.

Figure 2.1 (a) The device-tissue interface. (b) A dynamic model at the device-tissue
interface when a current is applied at the electrode. (Adopted from Merrill 2005)
45

2.2.2 Cyclic Voltammetry
Cyclic voltammetry (CV) is an experimental method to determine the nature of
the electrochemical processes, i.e. redox reactions, at the device-tissue interface. A CV
curve is obtained from applying a triangular waveform potential, as shown in Figure 2.2A,
to an electrode at a chosen scan rate from a negative potential to a positive potential and
vice versa, and measuring the resulting current. CV plots current (vertical axis) versus
potential (horizontal axis), as shown in Figure 2.2B. The measurements are carried out
using a reference electrode since this provides a stable voltage reference. There are two
important parameters that can be determined from the graph of CV. The first one is the
water window, which is defined as the potential region between the oxidation of water to
form oxygen and the reduction of water to form hydrogen. Once the potential of the
electrode exceeds these window boundaries, the injected charges go into irreversible
processes (Faradaic process), resulting in evolving H
2
and O
2
as well as releasing
chemical byproducts to the electrolyte. Therefore, it is critical to keep the potential of the
electrode within the water window during stimulation. The second parameter is the area
inside the cycle, from which the charge storage capacity can be calculated. The larger the
area per cycle, the more charge stored at the interface. By knowing the area of the cycle
and the area of the electrode, charge density of the stimulating electrode can be
determined.
Alloys of platinum with 10-30% iridium have similar charge storage capacity to
pure platinum (Robblee, Lefko et al. 1983). Charge storage capacity is greatly increased
46

Figure 2.2 A: A triangular potential waveform is the excitation signal for cyclic
voltammetry. B: A typical cyclic voltammogram with scan initiated at 0.8V. The vertical
axis in A is the horizontal axis for B. SCE denotes the saturated calomel electrode.
(Adopted from Kissinger 1970)
47

when a surface oxide is present. Iridium oxide has a charge storage capacity larger than
platinum or alloy of platinum and iridium.
2.2.3 Impedance Spectroscopy
Electrochemical impedance spectroscopy (EIS) measures the complex impedance
of an interface over a wide range of frequencies. It is used to characterize the electrical
properties of a system. In a neural stimulation perspective, the electrode impedance is
designed to be as low as possible to avoid cell damage, since high impedance would
result in a large applied electrode potential leading to undesirable electrochemical
reactions at the interface (Faradaic process). Low impedance also benefits the
microelectronic design, since low-power devices can be used. Methods of decreasing
impedance include increasing electrode surface area and increasing roughness. In a
Randles circuit, as shown in Figure 2.3A, the overall impedance consists of three
components, the electrode impedance for charge transfer, a double layer capacitor and the
tissue impedance. Electrode impedance is frequency dependent. The double layer
capacity is in parallel with the electrode impedance due to the charge transfer reaction.
This circuit can be described mathematically as follows.
e dl
e
s dl e s total
R C j
R
R C R R Z
ω +
+ = + =
1
|| (Eq. 2.4)
()
2
1
e dl
e
s total
R C
R
R Z
ω +
+ = (Eq. 2.5)
48

()
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
+ +
−
= ∠
−
e e dl s s
e dl
total
R R C R R
R C
Z
2
2
1
tan
ω
ω
(Eq. 2.6)
The low frequency part of the curve, as shown in Figure 2.3, largely represents
the capacitive double layer charging and resistive faradiac reaction, while high frequency
portion is the impedance of the tissue, since the capacitive double layer will be a short
circuit at high-frequency. A review of the characteristics of the metal-tissue interface of
stimulation electrodes can be found in (Dymond 1976).
Impedance measurements at the electrode-tissue interface have been used to
characterize biomedical applications. Reactive tissue encapsulation of chronically
implanted microelectrode probes can hinder long-term recording of extracellular action
potentials. Tissue encapsulation can be assessed by measuring the impedance
spectroscopy. By applying a voltage bias, investigators reported that the resistive tissue
impedances were reduced, resulted in higher signal-to-noise ratio (Johnson, Otto et al.
2005). An increase in impedance has been reported in chronic implant in cat cochlea in
the post-operative period. Changes in the composition of perilymph or extracellular fluid
adjacent to the electrodes as a consequence of tissue response were suggested to cause the
elevated impedance, affecting the efficiency and quality of neural stimulating electrodes
(Duan Y.Y. 2003). Another study has also demonstrated that impedance as a useful
indicator for predicting electrode-tissue contact in the in vivo pig thigh muscle model.
(Zheng, Walcott et al. 2000).

49

Figure 2.3 A: Randles circuit. Bode plots of frequency (B) and phase (C) plots.
50

2.3 Assessment of Electrode-Retinal Proximity
2.3.1 Rationale
In clinical trials of retinal prostheses, a positive correlation between patients’
thresholds and time has been reported. Patients’ thresholds tended to increase
postoperatively, consistent with the electrode array lifting off the retina. A significant
negative correlation between impedance and threshold was also shown. Patients’
thresholds tended to decrease when the impedance measured at the interface increased
(Mahadevappa, Weiland et al. 2005; de Balthasar, Patel et al. 2008). An early model
study has demonstrated that the threshold stimulation current was dependent upon the
distance of the electrode from the target tissue. Smaller distances from the target required
less electrical energy for stimulation (BeMent and Ranck 1969). A recent in situ study
showed that the impedance of the retina was greater than that of the vitreous at
frequencies of 1 kHz and above, suggesting a reliable detection of proximity of electrode
to the retina.
It is important to maintain a stable interface in our in vivo experiments such that
the change in thresholds in our experimental group is independent of the distance
between the electrode and the retina, caused by undesirable motion during surgery or by
the animal while under anesthesia, both of which may lead to the electrode lifting off the
retina.
51

Here, a method is presented to standardize the assessment of the electrode-retina
proximity, enabling a relatively stable interface while an effective current is delivered to
activate neuronal responses without mechanically damaging the retina. In the following
sections, methods of conducting histological and electrophysiological experiments are
described, together with the results of obtaining a range of impedance values which
enables a functional stimulation without damaging the retina.
2.3.2 Materials and Methods
Animals
Eight Copenhagen rats with normal retinas were used in this study. All
experimental procedures were done in accordance with the guidelines of Institutional
Animal Care and Use Committee at University of Southern California for animal care
and followed the ARVO guidelines for animal usage.
Experimental Groups
There were three experimental groups. Groups 1 and 2 were histological
experiments to examine tissue after the electrode was placed in the eye with impedance at
the retinal-electrode interface > 15 k Ω and < 10 k Ω respectively, for an hour. Group 3
collected electrophysiological result to determine the effects of impedance and distance
between electrode and retina on threshold values.

52

Electrical Stimulation
Charge-balanced, cathodic first, biphasic currents (1-100 µA) were applied to the
epiretinal surface across seven pulse durations (0.1, 0.2, 0.5, 0.8, 1.0, 1.5, 2.0 ms) through
a flat concentric bipolar stimulating Pt/Ir electrode (described above). The current pulses
were generated by the voltage-to-current converter (Model 2200, A-M Systems, WA)
which was triggered from a programmable voltage pulse produced by a
stimulus/recording system (DataWave Technologies, CO). The inter phase interval was
held constant at 100 µs. The stimulus voltage at the interface, buffered by a differential
amplifier (Model P15, Grass Technologies, RI) was monitored at each delivery of current
pulse on a computer screen to ensure the electrode integrity throughout all experiments.
Electrochemical Impedance Spectroscopy
The electrochemical impedance was measured using a commercial potentiostat
(FAS1, Gamry Instruments Inc.). The Pt/Ir inner pole electrode was connected to the
working electrode input, and two needle electrodes served as the counter and reference
electrode which were placed under the skin of the nose and tail respectively. The
reference electrode (Ag/AgCl) is used in measuring the working electrode potential. The
counter, or auxiliary, electrode is a conductor that completes the cell circuit. The current
that flows into the body fluid via the working electrode leaves the fluid via the counter
electrode. All impedances were measured at open-circuit potential using a 10 mV (rms)
AC sinusoid signal. A simplified schematic of a Gamry’s potentiostat is shown in Figure
2.4.
53

Figure 2.4 This electronic circuit block reveals the connections inside the Gamry
potentiostat, where a voltage AC signal is delivered to the electrolyte and current is
sensed and measured, hence the impedance is calculated. (Adopted from Gamry’s
Instrument Application Notes)
Electrophysiological Recording
The surgery and the recording methods were the same as the method section in
Chapter 3. Please refer to that section for details.
Tissue Preparation
All eyes were fixed and processed with the same protocol. Before the animal was
euthanized, the eyes were rapidly enucleated from the orbit. A small incision was made
with a razor blade to indicate the superior pole of the globe. The corneas were slit and the
54

lens was removed, and eyecups were immersed in 4% paraformaldehyde in 0.4 M
phosphate buffer (PB) for 2 hours at 4
o
C. Following fixation, eyecups were embedded in
Tissue-Tek OCT Compound (Tissue-Tek, Hatfield, PA) and fast frozen in liquid nitrogen.
Vertical cryostat sections of 25-µm thickness were cut perpendicular to a plane tangent to
the corneal surface from nasal to temporal order and collected onto Superfrost/Plus
microscope slides (Fisherbrand; Fisher Sientific, Pittsburgh, PA). Every 5
th
slide was
stained with hematoxylin/eosin. The unstained sections were stored at -70
o
C.
Data Analysis
Post-processing was done in Matlab. Response was defined as a deflection
exceeding 5 sigma, where 1 sigma equaled the standard deviation of the noise (signal-to-
noise ratio). Threshold was determined when the signal-to-noise ratio (SNR) was greater
than 5. Twenty sweeps were averaged. The SNR was calculated as follows.
⎟
⎟
⎠
⎞
⎜
⎜
⎝
⎛
=
noise
signal
P
P
dB SNR
10
log 10 ) ( . P
signal
was the power from 5 to 55 ms (post-stimulus) and
P
noise
was the power from 85 to 100 ms. Stimulus artifact was excluded in the calculation
of power signal. The discrete Fourier transform (DFT) of the time signal is first computed.
Average power was then calculated by Parseval’s theorem which stated that the power
calculated in the time domain is the same as the power calculated in the frequency
domain.
∑
−
=
−
=
1
0
2
N
n
kn
N
i
n k
e x X
π
i
1 ,..., 0 , − = N k .
∫
∑
= =
k k n av
dX X x P
2 2
. P
av
is the
55

average power. X
k
is the DFT of the time signal x
n
. N is the number of data points
(usually 1024).
2.3.3 Results
Group1 – Electrode Contacting the Retina
In this experimental group, the electrode was intentionally contacted to the retina
surface, as seen from the microscope. The electrode was held in the same position for an
hour. Electrochemical impedance was measured. Two experiments were performed in
this experimental group. The average impedances measured were 26.5±1.4 k Ω and
29.0±1.7 k Ω, as shown in Figure 2.5 (A) and (B). The electrochemical impedance
measured throughout an hour period changed. The change was unpredictable.
Readjustment of the electrode position was necessary to keep the impedance value
constant. For example, if the impedance increased, indicating that the distance between
the electrode and the retina decreased, the electrode was moved away from the retina and
hence the impedance value decreased accordingly. Structural changes occurred when the
retina was examined in hematoxylin/eosin-stained slides in this experimental group, as
shown in Figure 2.5 (C) and (D). A focal lesion was clearly seen in this experimental
group, suggesting the electrode was contacting the retina at impedance higher than 25 k Ω.
Group2 – No electrode contact
In this experimental group, electrode was placed very close to the retina with
electrochemical impedance maintaining below 10 k Ω for one hour. Three experiments
were performed in this experimental group. The average impedances measured were
56

6.9±0.4 k Ω, 7.2±0.5 k Ω and 7.5±0.5 k Ω. Two representative experiments were shown in
Figure 2.6 (A) and (B). Histological results showed that no mechanical damage to the
retina was observed in this experimental group, as shown in Figure 2.6 (C) and (D).

Figure 2.5 The impedance measurements for one hour are shown in A and B. C and
D show the local lesions of the electrode contact, indicated by the arrows.

Group3 – Effect of Retinal Electrode Proximity on Threshold Values
In this experimental group, the effects of the retinal electrode proximity on the
threshold values and the impedance values were determined. Three experiments were
57

Figure 2.6 The impedance measurements for one hour are shown in A and B. C and
D show no apparent damage to retina.

performed in this experimental group. All impedances were kept below 10 k Ω. The
electrode was first positioned close to the retina when the impedance value was 8 k Ω
(ref). It was then escalated, manually, in a 20-µm increment step. Twenty stimulus pulses
were delivered to the retina and the threshold values were determined (see Methods
section) at each increment. The threshold values increased monotonically as the electrode
was moved away from the retina, as shown in Figure 2.7 (A). The sum of the incremental
movements was 180-µm. The charge densities, as shown in Table 2, were all within the
electrochemically safe limit of the platinum electrode (350 µC/cm
2
). Conversely, the
58

impedances decreased monotonically as the retinal-electrode proximity increased, as
shown in Figure 2.7 (B).

Figure 2.7 A: Current threshold is positively correlated with the distance between
electrode and retina. B: In contrast, impedance is inversely correlated with the distance
between electrode and retina.

59

Table 2.1
Charge Density for Different Electrode Positions
Position Threshold Current Charge Density
(µm) (µA) (µC/cm
2
)
ref 1.33 30.17
20 1.67 37.71
40 4.00 90.50
60 4.00 90.50
80 4.00 90.50
100 4.00 90.50
120 6.67 150.83
140 5.00 113.12
160 10.00 226.24
180 15.00 339.37

2.3.4 Conclusions
A method of assessing and maintaining the retinal-electrode proximity during the
threshold measurement in vivo has been demonstrated. Impedances measured less than 8
k Ω at the retinal-electrode interface reliably obtained a visual response in the central
visual pathway without mechanically damaging the retina, under a safe stimulation of
platinum electrode. This study is in agreement with the clinical study that threshold value
is directly related to the retinal-electrode proximity and the impedance is inversely related
to the retinal-electrode proximity (Mahadevappa, Weiland et al. 2005). This study is also
in agreement with the in situ study showing the impedance value in vitreous was lower
than that on the retina (Shah, Hines et al. 2007). This study further demonstrated that a
standardized method to stimulate retina in vivo in an impedance range where no electrode
contact was found (< 8 k Ω) and a structural damage was shown (> 25 k Ω).
60

This method was employed in studies described in the following chapter, which
investigate the threshold for retinal activation in degenerate retina in vivo.
2.4 Tissue Damage at the Interface
Previous studies have been investigated various causes of retina damage. One study
reported damage after electrical stimulation of an epi-retinal type heater probe caused
immediate change of canine retina (Piyathaisere, Margalit et al. 2003). Other reports
showed that chronic stimulation by sub-retinal type electrodes caused acute change of
rabbit retina (Schwahn, Gekeler et al. 2001). Low-levels (up to 0.1 mC/cm
2
) of chronic
stimulation (6 months) of canine retina by epi-retinal type electrodes showed no damage
correlated to electrical stimulation (Guven, Weiland et al. 2005). One report
demonstrated retinal layers disorganization under high charge delivery in a supra-
choroidal approach. Histological changes, such as enlargement of choroidal vessels,
irregular arrangement of outer nuclear layer cells, hypertrophy, and expansion of the
extracellular space in the outer plexiform layer, inner plexiform layer, and nerve fiber
layer were observed (Nakauchi, Fujikado et al. 2007). In one recent study, acute
stimulation by epi-retinal type electrode caused neural damage only in the outer retina,
including a disturbance of synaptic vesicle proteins in the photoreceptor terminals and a
remodeling of horizontal and rod bipolar cells’ processes (Ray, Colodetti et al. 2009).
Study towards exploration of the histological safe injection limit, without mechanical
pressure exerted by an electrode, is important for a thorough understanding of the safe
stimulation of retinal implant development. The method employed in this chapter
61

provides a means to investigate the safe stimulation associated with neural damage in
vivo.
62

Chapter 3: Electrical Excitation of Remodeled Retina
In the previous chapters, we reviewed prior knowledge concerning neuroanatomy of the
visual pathway and the principle of neuro-stimulation, then presented an impedance-
based method of stabilizing the electrode in close proximity to the retina surface. This
serves as a basis for an investigation of the neuronal electrical excitation in aged,
degenerate, remodeled retina. In this chapter, the motivation of looking into the electrical
excitation of remodeled retina will be discussed. Followed by that, the materials and
methods employed in achieving the goal of this chapter will be described. Then, the
findings of this goal will be reported and follow by conclusions and discussions based on
the findings.
3.1 Background
Early clinical studies have shown the higher current and charge requirement in
stimulating a RP retina for a phosphene perception compared to normal healthy retina,
using large electrodes (260- to 520-µm). Reducing electrode sizes will allow more
electrodes within the same retinal area, translating into more pixels per degree of visual
angle. Simulations of prosthetic vision suggests that more electrodes within the same
retinal area may lead to a higher-resolution image and better visual task performance
(Hayes, Yin et al. 2003; Thompson, Barnett et al. 2003). Small electrode size is the trend
for a high resolution design of retinal prosthesis. However, charge densities increase as
the electrode sizes decrease. High charge injection within a small electrode area will lead
to neural damage caused by dissolution of the stimulating electrodes when safe charge
63

density limit is exceeded. Studies of charge densities using small electrode area has been
demonstrated in normal healthy retinas, which have described low threshold charge
(Jensen, Rizzo et al. 2003; Wilms, Eger et al. 2003; Sekirnjak, Hottowy et al. 2006;
Ahuja, Behrend et al. 2008). There are only a few studies investigating the current and
charge requirement in remodeled retina (O'Hearn, Sadda et al. 2006; Jensen and Rizzo
2008). There is no study on such using small electrodes. Therefore, this current study to
investigate the threshold and charge requirement in remodeled retina using small
electrode will provide important information on the efficacy and safety of a high
resolution design of a retinal prosthesis.
3.2 Materials and Methods
3.2.1 Animals
Normal Copenhagen (P90-700, n =11) and heterozygous S334ter-line-3 (P84-782,
n =25) rats were used in these studies
3
. S334ter-line-3 rats were grouped in two different
stages (young rd group, P84-310, n=11; old rd group, P490-782, n=14). Animals were
housed in covered cages and were fed with a standard rodent diet ad libitum, while kept
on a 12-h light: 12-h dark cycle in the Doheny Eye Institute animal facility. Homozygous
S334ter-line-3 rats were mated in house with Copenhagen rats (Charles Rivers, Hollister,
CA) to produce the heterozygous pigmented offspring. Homozygous breeding pairs of
S334ter-line-3 rats were produced by Xenogen Biosciences (formerly Chrysalis DNX

3
P stands for post natal, so P84 means 84 days after birth.
64

Transgenic Sciences, Princeton, NJ), developed and supplied with the support of the
National Eye Institute by the courtesy of Dr. Matthew LaVail, University of California
San Francisco. All experimental procedures were done in accordance with the guidelines
of Institutional Animal Care and Use Committee at University of Southern California for
animal care and followed the ARVO guidelines for animal usage.
3.2.2 Surgical Procedures
All surgeries were performed under general anesthesia induced by intraperitoneal
injection of ketamine (100 mg/kg; KETASET, Fort Dodge, IA) and xylazine (20 mg/kg;
X-Ject SA, Butler, Dublin, OH) and maintained by sevoflurane (1% in 100% O
2
)
throughout the entire experiment. Deeply anesthetized rats were positioned in a
stereotaxic apparatus (Kopf Instruments, CA) and the skin overlying the skull was cut
open and retracted. Gas inhalant anesthetic was administered through an anesthetic mask.
The animal’s rectal temperature and pulse were monitored during the surgical procedures.
Body temperature was monitored with a rectal thermometer and maintained at 37
o
C with
a self-regulating electric heating blanket (model 50-7053-F; Harvard Apparatus,
Holliston MA). An effective anesthetic state was ensured throughout the experiment by
monitoring the heart rate and blood pressure with the sensor clipped to the hind leg
(Heska, Loveland, CO). Increase of heart rate signaled the need for anesthetic
supplementation. After the experiments, all animals were euthanized by an intraperitoneal
injection of pentobarbital (30 mg/kg; Butler, Dublin, OH).

65

3.2.3 Insertion of Stimulating Electrode
The surgical procedures of implantation of the stimulating electrode were similar
to those reported in previous work (Colodetti, Weiland et al. 2007; Ray, Colodetti et al.
2009). The pupil was dilated with 1% tropicamide (Tropicacyl, Akorn Inc., Buffalo
Grove, IL) and 2.5% phenylephrine (AK-Dilate, Akorn Inc.). Rat’s left eyes were
proptosed under anesthesia. The fundus was viewed in focus by slightly flattening the
cornea with a glass coverslip through an operating microscope. With the eye gently
proptosed, an incision was made by a 25-gauge needle just behind the limbus. The
insertion angle of the needle was inclined at an angle of 45
o
to 60
o
with respect to the
contact surface to avoid damaging the lens which is two-third in size compared to the
globe in rodents. The stimulating electrode was the inner pole of a concentric bipolar
Pt/Ir electrode (Model CBDFG74, FHC Inc., Bouidain, ME) with a flat tip (inner pole
diameter 75-µm, surface area 4.42 x 10
-5
cm
2
). The stimulating electrode was mounted in
a 1-cc syringe for handling and attached to a single-axis linear translational stage with 10-
µm graduations (Model NT33-475, Edmund Optics, Barrington, NJ) on a magnetic based
articulating arm. The stimulating electrode was inserted through the insertion site along
the same path as the needle and positioned in the ventral temporal quadrant without
contacting the retina. The final positioned to ensure close proximity utilized impedance
feedback as described below.

66

3.2.4 Impedance Testing
The electrode-tissue contact was indicated by an electrochemical impedance
measurement (Zheng, Walcott et al. 2000; Duan Y.Y. 2003). It was noted that a low
psychophysical threshold was correlated with high electrode impedance which suggested
that an electrode is closer to the retina (Mahadevappa, Weiland et al. 2005; Shah, Hines
et al. 2007; de Balthasar, Patel et al. 2008). The placement of the electrode relative to the
retina was indirectly indicated by measuring the electrochemical impedance using a
commercial potentiostat (FAS1, Gamry Instruments Inc.). The Pt/Ir inner pole electrode
was connected to the working electrode input, and two needle electrodes served as the
counter and return electrode which were placed under the skin of the nose and tail
respectively. All impedances were measured at open-circuit potential using a 10 mV
(r.m.s.) AC sinusoid signal. The frequency chosen was 100 kHz. The impedance
measurement was monitored before each current delivery. A custom made circuit board
was built to switch the recording between the impedance measurement and the stimulus.
3.2.5 Recording Electrode Positioning
To position the recording electrodes, the skull was exposed and a craniotomy of
the right skull was made (caudal-medial corner: ~4 mm caudal and ~3 mm lateral to
lambda) with a bit driven by a drill hand piece (Dremel, Robert Bosch Tool Corporation,
Mount Prospect, IL). The overlying cortex was aspirated through craniotomy until the
surface of the superior colliculus was exposed (~ depth of 4 mm from the dura meter).
Epoxy-coated tungsten microelectrodes (10M Ω, FHC) were positioned within the
67

superficial gray or stratum griseum superficale (SGS), the upper 100 – 200 µm after SC
surface penetration. A characteristic noise, produced by “juxtazonal potentials”, first
described in experiments conducted on cats (McIlwain 1978), was heard immediately
after penetration of the pia over the SC, via speakers connecting to the recording channel.
This noise represents massive extracellular potentials from retinal axon spikes
terminating in SGS layer, which exclusively processes visual information. Retinal input
to SC is densest in this layer. The cells within this layer had no or very low spontaneous
activity (Girman and Lund 2007). Recording of spike was successful in 40% of
experiments and the trace was shown in Figure 3.1A. The juxtazonal potentials were
recorded in approximately 60% of the recording sessions, shown in Figure 3.1B. Both
spikes and juxtazonal potentials are known retina-driven SC responses. This recording
technique is similar to other work that has used the SC to investigate experimental
treatments for retinal blindness (Sagdullaev, Aramant et al. 2003; Kanda, Morimoto et al.
2004; Thomas, Seiler et al. 2004; Girman, Wang et al. 2005; DeMarco, Yarbrough et al.
2007).
The experiment required simultaneous positioning of two electrodes: one on the
retinal surface, another on the SC surface. Above we described each positioning
individually. Through trial and error, the following sequence was established to minimize
animal movement during critical times of electrode positioning: 1) craniotomy 2)
stimulating electrode insertion into vitreous cavity 3) SC exposure and recording
electrode positioning 4) stimulating electrode positioning on epiretinal surface.
68

Figure 3.1 Examples of spikes (A) and juxtazonal potentials (B) recorded from SGS
neurons.

3.2.6 Electrical Stimulation
Charge-balanced, cathodic first, biphasic currents (1-500 µA) were applied to the
epiretinal surface across seven pulse durations (0.1, 0.2, 0.5, 0.8, 1.0, 1.5, 2.0 ms) through
a flat concentric bipolar stimulating Pt/Ir electrode (described above). The current pulses
were generated by a voltage-to-current converter (Model 2200, A-M Systems, WA)
driven by a voltage pulse from a programmable analog output card (DataWave
Technologies, CO). The inter phase interval was 100 µs. Blocking capacitors were used
to minimize DC current. The electrode voltage for each current pulse was monitored by
viewing the voltage trace, buffered by a differential amplifier (Model P55, Grass
69

Technologies, RI) to ensure the electrode integrity during the stimulation experiment and
maintenance of the voltage within the compliance limits of the stimulator box.
3.2.7 Light Stimulation
A full-field strobe flash (1000 cd/m
2
) was delivered to the test eye using a
photostimulator (model PS 22 Photic stimulator; Grass, RI), positioned approximately
30-cm in front of the rat’s eye. An inter stimulus interval of 5 second was used. Light
stimulation was performed to examine the healthy state of the retinas.
3.2.8 Experimental Protocol
The threshold current was determined as followed. First, three consecutive current
levels (1 µA increments up to 50 µA and 10 µA increments up to 500 µA) were applied
to the retina to evoke a response. A response was defined as neural activity correlated
with stimulus that exceeded 5 times the baseline noise level. Electrical threshold was
defined as the amount of current required to evoke a response in 75% of 50 trials. This
threshold determination procedure was repeated across all tested pulse durations. We
measured electrical thresholds using a single pulse across seven pulse durations (100 µs
to 2 ms). All pulse waveforms were biphasic and charge balanced.

70

3.2.9 Data Analysis
Electrophysiology Recording
Data were analyzed off-line using SciWorks and Matlab. Means and stand error
(SE) values were calculated in Microsoft Excel. Error bars reported in this study are SE.
Only negative error bars were reported for clarity of graphical data presentation.
Statistical comparisons were done by performing non-parametric one way ANOVA test
(Wilcoxon and Kruskal-Wallis) in SAS 9.2. Images were processed in Adobe Photoshop
7.0. Chronaxies were calculated by fitting exponential function ()
a x
e b y
/
1 /
−
− = to the
strength duration data (Lapicque 1907; Ranck 1975), where x represents the pulse
duration and y represents the predicted threshold values of that pulse duration based on
the function. The asymptote was defined as the rheobase (coefficient b ); chronaxie was
calculated as 2 ln / a . Fit quality was assessed by calculation of R
2
and visual inspection
of the fit curve. Noticeably poor fits were not included in the calculation of chronaxie and
rheobase values.
3.3 Results
3.3.1 Spatial Characteristics of Collicular Response
Electrically evoked responses were recorded in normal rats (N=8) and old RD rats
(N=4), in a discrete region of the SC that corresponds topographically to the location of
placement of the stimulating electrode. In both normal and RD rats, a larger SC area was
71

activated by monopolar stimulation versus bipolar stimulation, as shown in Figure 3.2).
The difference between monopolar and bipolar stimulation is the placement of the return
electrode.
3.3.2 Characteristic of Retinal-driven Collicular Response to Photic and Electrical
Stimulation
The light-evoked response and electrically-elicited response (EER) were
characterized in Figures 3.3A and 3.3B respectively. Light-evoked responses were
observed in all ages (P90 – P700) of control animals. Light-evoked responses with
latency > 50-ms were typical and classified as short-latency response. Light-evoked
responses with latency > 200-ms were infrequently observed (3 out of 10) and classified
as long-latency responses. The amplitudes of the long-latency response were lower and
the response window was wider than then short-latency response. Light-evoked responses
were only observed in P84 – P185 in the transgenic animals, with the exception of one
observation in P648. EERs were observed in both control and transgenic animals at all
ages. Figure 3.3B shows both short (5-ms) and long (> 25-ms) latency EER. Short
latency responses were used for threshold, as long latency responses were observed less
than 50 % of the time. Others have reported inconsistent late responses in the EER (Chen,
Mahadevappa et al. 2006).

72

Figure 3.2 Representative spatial maps for 1 normal and 1 RD rat. The cross-hatched
area is the SC. The circled + is the low threshold location. ± show other recording
locations that were responsive at less than 200 µA. ~ shows recording areas that did not
respond. These maps show typical results for monopolar and bipolar stimulation.
Monopolar excited a larger area of SC. Normal rats had a larger response area. MP –
Monopolar, BP – Bipolar, N – Normal retina, R – Degenerate retina.

Activation Thresholds Required in Remodeled Retina
The calculation of activation threshold was defined in Materials and Methods
section. Strength-duration curves of the control and transgenic animals are shown in
Figure 3.4 with seven pulse durations varied from 100 µs to 2 ms. Threshold currents
were measured in mean ± S.E. To determine whether there was an age dependent effect
73

on the change of activation threshold, we tested whether the activation threshold
correlated with age in the control group. The activation thresholds were ranged from 5 ±
0.4 µA at 2 ms pulse duration to 25 ± 1.3 µA at 100 µs pulse duration in control group, as
shown in Figure 3.4A. We found no correlation (r = 0.14 to 0.25, P = 0.46 to 0.86) across
all tested pulse durations, as shown in Table 3.1, suggesting that aging alone did not play
any role in affecting the threshold change in normal retina. However, the activation
thresholds of the degenerate group increased progressively with age, as shown in Figure
3.4B. We found correlation of the activation threshold and age in the degenerate group (r
= 0.72 to 0.84, P < 0.0001), as shown in Table 3.2. Both results were plotted in Figure 3.5.
The thresholds between control, young rd and old rd groups were tested. There
was no significant difference (p > 0.05) in activation thresholds between the control
group and young rd group. The activation thresholds were ranged from 5.4 ± 0.7 µA at 2
ms pulse duration to 32 ± 4 µA at 100 µs pulse duration in young rd group. The
activation thresholds across all tested pulse durations in old rd group were all
significantly higher (p < 0.05) than the same age range in the control group. Thresholds at
this age group ranged from 16 ± 4 µA at 2 ms pulse duration to 97 ± 34 µA at 100 µs
pulse duration, as shown in Figure 3.6.
74

Figure 3.3 Plots of light-evoked response in P664 control animal and electrically-
elicited response in P86 transgenic animal. A: top: Trigger pulse is shown above the trace.
middle: Trace at threshold after artifact subtraction. Note that residual of the stimulus
artifact was left right after stimulus onset. bottom: Latency histogram of light-evoked
response. Bin width: 10 ms. B: top: Charged-balance biphasic stimulus pulse was shown
above the trace. middle: Trace at threshold (5 µA, 1 ms pulse duration) after artifact
subtraction. bottom: Latency histogram of early electrically-evoked response and
spontaneous responses. Bin width: 5 ms.
75

Figure 3.4 Strength duration curves of normal retinas among four different age
groups (A) and degenerate retinas among six different age groups (B).
76

Table 3.1
Spearman Correlation Coefficients for Normal Retinas, N = 11
Pulse
duration
(ms) 0.1 0.2 0.5 0.8 1.0 1.5 2.0
Rol -0.45 -0.14 0.14 0.06 0.15 0.16 0.25
p-value 0.16 0.68 0.68 0.86 0.66 0.64 0.46

Table 3.2
Spearman Correlation Coefficients for Degenerate Retinas, N = 11
Pulse
duration
(ms) 0.1 0.2 0.5 0.8 1.0 1.5 2.0
Rol 0.76499 0.72 0.83 0.84 0.80 0.79 0.78
p-value <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001

Figure 3.5 Spearman correlation test between normal and degenerate retinas. Low
coefficient indicates no correlation among the group, in normal retinas. In degenerate
groups, high coefficient (close to one) indicates a positive correlation between threshold
and age.
77

Figure 3.6 Strength duration curves after re-grouping of control, young rd and old rd
groups. Thresholds of old age group were significantly higher than control group and
young rd group for all tested pulse durations.

The threshold currents across all pulse durations were compared between the
juxtazonal potentials and spikes. The threshold currents of juxtazonal potentials and
spikes were similar in control group. In young rd group, the threshold currents of
juxtazonal potentials were higher than those of spikes. In contrast, the threshold currents
of juxtazonal potentials were smaller than those of spikes, in old rd group. Thus, there
was no consistent trend of the strength duration curves between juxtazonal potentials and
spikes, as shown in Figure 3.7, and hence they were considered to be equivalent.
78

Figure 3.7 This graph showed that there was no apparent trend of the strength
duration curves between juxtazonal potentials and spikes. Solid lines represent thresholds
of juxtazonal potential and dotted lines represent thresholds of spikes. “jux” denotes
juxtazonal potential and “su” denotes single units or spikes.

Charge Density
The charge densities of stimulating normal and transgenic retinas were calculated
for comparison with the electrochemical safety limit of the stimulating electrode. The
charge density curves were shown in Figure 3.8. Charge densities across all pulse
durations in control group were statistically similar to those in young rd group and they
ranged from 0.19 ± 0.02 mC/cm
2
at 2 ms pulse duration to 0.07 ± 0.01 mC/cm
2
at 100 µs
pulse duration. For old rd group, the charge densities required to elicit response were
79

significantly higher (p < 0.05) than the control and young rd group. There were four
charge densities, with pulse durations above 500 µs, in the old rd group exceeded the
electrochemical safety limit of platinum (0.35mC/cm
2
, for pulse duration above 600 µs).
Chronaxies and Rheobases
After fitting our activation threshold current data using the Lapacique’s equation,
we measured the chronaxie and rheobase values from the fitted curve. The average
rheobase and chronaxie were shown in Figures 3.9A and 3.9B respectively. The average
rheobase of control group was 5.3± 0.58 µA. The average rheobase of young rd and old
rd groups were 5.5 ± 0.68 µA and 16.23 ± 4.26 µA. The average rheobase of old rd group
was significantly higher than those of the control and young rd group (p < 0.05). The
average chronaxies of the control, young rd and old rd groups were 0.31 ± 0.02 ms, 0.37
± 0.02 ms and 0.36 ± 0.02 ms respectively. The chronaxies of rd groups were not
significantly different from those from control group.
3.4 Conclusions and Discussions
Our experiments demonstrate four major findings. First, the old rd retina has higher
threshold. Second, there is no threshold drift in stimulating the young rd retina but
threshold is drifted when degeneration stage reaches P500 in S334ter-line-3 rats. Third,
the same cell types (same chronaxie values), speculating retinal ganglion cell, are being
stimulated even the degeneration progresses. Fourth, the charge density stays within the
safety limit of the platinum electrode with 75-µm diameter, but in a limited pulse range.
80

Figure 3.8 Charge density curves obtained for control and degenerate retinas. The
horizontal line showed the electrochemical safety limit for platinum electrode which is
0.35mC/cm
2
.

Figure 3.9 Rheobases (A) and chronaxies (B) of control and degenerate retinas.
81

Efficacy and Safety in Stimulating Remodeled Retina
Retinal electrical stimulation has been studied in a number of animal models and
in human disease. In this study, electrophysiological responses in postsynaptic visual
target of the normal and degenerate rats were compared to evaluate the change of
electrical thresholds across an extensive age stage (P84 – P782). Electrical thresholds had
been studied extensively in healthy retinas which showed the feasibility of evoking
physiological responses. Low thresholds were measured in healthy retinas and hence the
charge densities delivered stayed within the electrochemical safety limit of the electrode.
Low charge densities using small electrodes were reported in several recent in vitro
studies of healthy retinas (Sekirnjak, Hottowy et al. 2006; Ahuja, Behrend et al. 2008).
The healthy state of the retina and the close contact between the microelectrode array and
the retina in the in vitro setup may facilitate low electrical threshold. The increase of
electrical thresholds in RP retina is present in clinical trials (Rizzo, Wyatt et al. 2003;
Mahadevappa, Weiland et al. 2005; de Balthasar, Patel et al. 2008). Likewise, animal
studies have confirmed higher threshold values in diseased retinas. It was found that the
threshold charge densities in 8-16 week-old rd1 mouse were substantially higher than that
in the normal mouse (Suzuki, Humayun et al. 2004; O'Hearn, Sadda et al. 2006) with an
array of platinum electrode of 125-µm in diameter. Yet, these studies only investigated
the threshold charge densities at one stage of retinal degeneration. The most recent in
vitro study, involving multiple ages of rd1 mouse, reported that the thresholds of retinal
ganglion cells were consistently higher than the wild-type mouse (Jensen and Rizzo
82

2008). This result is in agreement with our findings, but we studied in an extensive age
group. The investigators reported that the threshold values were stable across the tested
age group of the rd1 mice, which is conflicting with our present findings. We
demonstrated that electrical thresholds increased progressively in degenerate rat retina
(P84-P782). We found no correlation (r = 0.14 to 0.25, P = 0.46 to 0.86) across all tested
pulse durations in the control group, suggesting that aging alone did not play any role in
affecting the threshold change in normal retina.
It is important for implantable stimulators to deliver current in degenerate retina
that is efficient as well as safe. Efficient implantable stimulators successfully evoke
electrophysiological responses. Neurons beneath the stimulating electrode could be
damaged by electrochemical reactions as well when the electrode is driven such that the
electrode potential exceeds a value where an irreversible reaction occurs at the electrode-
tissue interface. The most commonly used material for electrical stimulation is platinum,
which can safely supply 0.1 to 0.4 mC/cm
2
of charge (Brummer and Turner 1977; Rose
and Robblee 1990), for pulse duration greater than 600 µs. For short duration pulses, less
than 200 µs, the safe limit for platinum electrode was proposed to be 0.15 mC/cm
2
(Rose
and Robblee 1990). High limit is recommended which enable safe stimulation of denser
and smaller microelectrodes for high resolution. Materials with higher limits are iridium
oxide (1 to 3 mC/cm
2
) (Agnew, Yuen et al. 1986; Beebe and Rose 1988) and titanium
nitride (0.6 to 0.9 mC/cm
2
) (Weiland, Anderson et al. 2002).
83

Debathalsar analyzed threshold data in human patients and concluded that
electrode size less than 200-µm in diameter would be within the safe limit of platinum
electrode and acceptable. In this study, we measured the required current for stimulating
degenerate retina using a small electrode, suggesting that electrode size as small as 75-
μm in diameter are acceptable with platinum, but in a limited pulse range (< 500 µs).
Small electrodes like those used in this study will be necessary for high-resolution
prosthetic vision. Complex stimulation patterns and temporal coding may be needed to
maximize efficacy of a high-resolution retinal implant.
Chronaxie value indicates the responsiveness of the target cell (Lapicque 1907;
Ranck 1975). We demonstrated that the chronaxie values did not change with age in rd
group, suggesting that the same cell is targeted. The target cell types, retinal ganglion
cells, are the same in normal and degenerate retina. Our chronaxies match those reported
for activation of ganglion cell in rabbit, rat, guinea pig and monkey in vitro models
(Jensen, Ziv et al. 2005; Sekirnjak, Hottowy et al. 2006). Chronaxies were measured from
0.14 to 0.43 ms using 125-µm electrode in the rabbit model and ~ 0.1 to 0.4 ms using ~ 6
and 25-µm electrode in rat, guinea pig and monkey’s retina.
Electrical Stimulation of populations of neurons vs. individual neuron
The large electrode size used in the retinal implant clinical trials most likely
stimulates several hundreds of ganglion cells. One of the strategies that the brain uses in
visual processing is population coding (Pouget, Dayan et al. 2000; Andersen 2008). How
84

the brain processes visual images depends not only on what message conveyed by each
single element conveys the message, but also on the message sent by an ensemble of
neural activity. The recruitment of a population of neurons may be necessary to provide
visual information for perception. Even though a small electrode size is used, it is
unknown to us that whether individual ganglion cell can be activated because of the poor
spatial resolution due to extension of electrical fields across large areas. The spread of the
electrical field is dependent on the placement of the return electrode. Single-cell
stimulation was recently reported to target one cell class, and all other nearby cells in the
same cell class were not activated, with electrode diameter varied between 9- and 15-µm
(Sekirnjak, Hottowy et al. 2008). The summation of electric fields when adjacent
electrodes were activating might recruit other cell class and this concept has yet to be
proved. Recent studies showed retinal ganglion cells survive and maintain normal
morphology and functionality is on single cell basis (Margolis, Newkirk et al. 2008). In
fact, it may be possibly that electrical stimulation of a single ganglion, when that cell is
tightly coupled to a small electrode, is indifferent to the degeneration state of the retina.
The perceptual impact of individual ganglion neuron is yet to be explored. It seems likely
that activation of concerted cells is required for present stimulation design to create
cortical response and hence perception.
Remodeling in Visual Pathway
The loss of retinal ganglion cell alone may not fully explain the alteration of the
response properties in this study. It is well known that remodeling is common in central
85

nervous system (CNS) pathways, triggered by pre-synaptic cell death. There is no reason
to eliminate the possible fact that the remodeling of the SC may contribute to the
increased activation threshold. One study showed that the degenerative changes, such as a
reduction in neuronal density in the lateral geniculate nucleus and superior colliculus,
continue into high order visual center after the complete loss of photoreceptors, a
reduction in the thickness of the inner plexiform layer and in the cell density of the
ganglion cell layer is observed in mice (Ward 1982). The CNS plasticity in response to
disease is unavoidable. To date, no study has investigated changes in threshold as a
function of CNS plasticity after photoreceptor loss.
86

Chapter 4: Retinal Ganglion Cell Density of Remodeled Retina
This study investigated the effect of retinal degeneration due to the loss of photoreceptors
on the retinal ganglion cell in a rat model of retinal degeneration. The goal was to
investigate possible linkage between the increased threshold measured in the previous
chapter with a reduction in retinal ganglion cell number of degenerate retina. An
introduction of relevant studies and the motive of this study are given, followed by the
method of immunocytochemistry. The results of the total cell count and the cell density
throughout the degenerate and control retina are reported.
4.1 Background
4.1.1 Motivation
Retinitis pigmentosa and aged related macular degeneration result in a progressive
decrease in the population of human retinal ganglion cells, the retinal neurons that project
to the brain via the optic nerve. For additional background information regarding to the
remodeling of degenerate retina, see Chapter 1.5. Morphological studies of degenerate
retina in animal model are not consistent because different strains and different elapsed
time were chosen. One recent study has shown survival of retinal ganglion cells up to
P280 in rd10 mouse (Mazzoni, Novelli et al. 2008) and another study has shown stable
rhythmic spiking activity is retained in rd1 mouse retina, from P124 to P210 (Margolis,
Newkirk et al. 2008), which conflicts with earlier animal studies reporting decline in
87

retinal ganglion cell count in rd1 mouse retinas, from P90 to P180 (Caley, Johnson et al.
1972; Grafstein, Murray et al. 1972) and in CBA/Ki mice from P1 to P35.
The previous chapter showed an increase of threshold values in degenerate retina which
is in agreement with clinical studies and in vitro studies. None of those studies have
performed histological analysis of the same retina used for electrophysiology, except
O’Hearn (O'Hearn, Sadda et al. 2006) which performed a coarse histology and showing
the thinning of degenerate retina. Thus, the mechanism explaining the increased threshold
consistently noted in human studies is not clear. In this chapter, we report morphemetric
properties of degenerate retina in S334ter rat model.
4.1.2 SMI-32 Antibody
There are at least 7 different types of the ganglion cell in the rat retina classified
by their soma size and dendritic field dimension (Huxlin and Goodchild 1997; Sun, Li et
al. 2002; Lin, Wang et al. 2004). SMI-32, a neurofilament antibody that labels ganglion
cells (Straznicky, Vickers et al. 1992; Coombs, van der List et al. 2006), was applied to
whole mounts of the control and the transgenic rat retinas. SMI-32 labeled ganglion cells
with large somata (~30-µm), 4-5 thick primary dendrites, and large sparsely branched
dendritic trees, and occasionally labeled melanopsin-expressing ganglion cells identified
by their small somata (~20-µm), 2-3 primary dendrites. Studies have shown SMI-32
labels retinal ganglion cells and the OPL processes in mice but almost no study has been
done in rat. To determine whether SMI-32 labels only retinal ganglion cells and processes
in the IPL in rat, we performed immunocytochemistry in vertical section. In Fig. 5A, only
88

cells in the ganglion layers were labeled. The processes in the inner plexiform layer (IPL)
and outer plexiform layer (OPL) were also labeled, indicating the presence of the
dendritic stratification of ganglion cells in the IPL and the horizontal processes in OPL.
Cells in the inner nuclear layers were not labeled with SMI-32 antibody. As a result, the
cells labeled with SMI-32 are considered retinal ganglion cells for mean density analysis.
However, it is clear that SMI-32 does not label all ganglion cells, so the density analysis
can measure relative density between normal and rd, but not absolute density of ganglion
cells.
4.2 Materials and Methods
4.2.1 Tissue preparation
Upon completion of the electrophysiology, the animals were euthanized. After
rapid enucleation, a small incision was made with a razor blade to indicate the superior
pole of the globe. The corneas were slit and the lens was removed. Eyecups were
immersed in 4% paraformaldehyde in 0.4 M phosphate buffer (PB) for 2 hours at 4
o
C.
Following fixation, eyecups were cryoprotected in 30% sucrose in 0.1 M PB. After 5
days, the eyecups were frozen in liquid nitrogen and stored at -70
o
C. For vertical sections,
the eyecups were embedded in Tissue-Tek OCT Compound (Tissue-Tek, Hatfield, PA)
and fast frozen in liquid nitrogen. Vertical cryostat sections of 25-µm thickness were cut
perpendicular to a plane tangent to the corneal surface from nasal to temporal order and
collected onto Superfrost/Plus microscope slides (Fisherbrand; Fisher Sientific,
89

Pittsburgh, PA). Sections were stored at -70
o
C until use. For retinal whole mounts, the
eyes were enucleated and the cornea and lens were removed as described above. Relaxing
cuts were made in the retinal margin to allow the retina to flatten. Following removal of
the sclera and the retinal pigmented epithelium, the neural retina along with some
overlying vitreous was rapidly removed.
4.2.2 Immunocytochemistry
For fluorescence SMI-32 immunocytochemistry, vertical frozen sections were
thawed in a 37
o
C oven for 20 min and 10 min in room temperature, then rinsed three 5
min in PBS. Non-specific labeling was attenuated with dilution of 1:10 normal donkey
serum for 1 hr in room temperature. Sections were incubated overnight at 4
o
C with SMI-
32 monoclonal antibody (1:1500). Binding of primary antibodies was detected using
fluorescent antibodies. After three 5 min rinses in PBS, sections were incubated for 1.5 h
in the secondary antibodies with AlexaFluor-488 (1:300; Invitrogen Corporation,
Carlsbad, CA ), which were raised in goat and specific for mouse immunoglobulins, in
room temperature. Sections were rinsed in 0.1 PB and coverslipped in a fade-retardant
mounting medium (Vectashield; Vector Labs or Prolong Gold Invitrogen) and examined
with the microscope.
Rat retinal whole-mounts were immunolabeled and prepared in a well container
such that they were capable of free movement within the well. Whole-mounts were rinsed
in 0.01M PBS and subsequently treated with detergent concentration of 0.1 to 1% Triton
X-100 for 40 min, rinsed three 10 min in PBS followed by incubating in a 1:10 dilution
90

for NDS for 1 h at room temperature to block nonspecific labeling. Retinas were
incubated with SMI-32 monoclonal antibody (1:1500), diluted in 0.01M PBS and applied
for 5 days at 4
o
C, and counterstained with TOPRO-3 iodide (Invitrogen). TOPRO-3
labeled nucleic acid staining (cell bodies). Since retinal degeneration condition may
affect the protein level of antibody such as SMI-32, we used TOPRO-3 as a control to
affirm that the change of cell reduction labeled by SMI-32 is due to cell death. After
washes in PBS, secondary antibody conjugated to AlexaFluor-488 (1:300; Invitrogen
Corporation, Carlsbad, CA ) were applied for 20 min at room temperature followed by
incubating at 4
o
C for 4 days covered by foil. Subsequently, retinas were then rinsed in
three 10 min in PB at room temperature, flattened onto plain microscope slides with
ganglion cell side up, coverslipped with a fade-retardant mounting medium (Vectashield;
Vector Labs or Prolong Gold Invitrogen) and examined in the fluorescent microscope.
4.2.3 Experiment Protocol
Experiment procotols for vertical section preparation and whole mount
preparation are described in the following figures.
4.2.4 Data Analysis
Imaging
Immunofluorescence images were processed in Zeiss LSM-PC software. The
brightness and contrast of the images were adjusted using Adobe Photoshop 7.0 (Adobe
91

Systems, Mountain View, CA). For presentation, all Photoshop manipulations
(brightness and contrast only) were carried out equally across sections.
Quantification
SMI-32 immuno-reactive ganglion cells in whole mount retinas were counted.
Cells were counted along both the nasal-to-temporal and dorsal-to-ventral retinal
meridians in fields of 200 µm × 200 µm per retina sampled in 1-mm steps. Densities
around the optic disc and within periphery were averaged in two opposite fields per retina
on each meridian. Total number of SMI-32 immuno-reactive ganglion cells per retina
was also counted. Cells labeled with TOPRO-3 in the ganglion cell layer were counted
and cell densities were measured in fields of 200 µm × 200 µm per retina.
4.3 Results
4.3.1 Distribution of SMI-32 Immuno-Reactive Cells
An early study has demonstrated that ganglion cells were more responsive at short
pulse durations ( ≤ 1-ms) and bipolar cells responded better to longer duration pulses ( ≥ 1-
ms) (Greenberg 1998; Fried, Hsueh et al. 2006; Sekirnjak, Hottowy et al. 2006). In
Chapter 3, the increase of activation threshold due to short pulse activation is more
prominent than with the long pulse duration, suggesting that the structural change in
ganglion cell density in old rd group may play a role in mediating the functional change
of the response. To test whether the mean density of ganglion cells was different in
control and degenerate retinas, we labeled ganglion cells throughout the retina with an
92

Vertical Section Immunocytochemistry
Take out retina and
thaw in room
temperature, 5 mins
Incubate in NDS,
1:10, 1 hour
Apply primary
antibody, SMI-32,
1:1500
Day 2
Put onto shaker,
30 mins
Rinse 3 times, 5
mins, 0.01M PBS
Apply secondary
antibody, Alexa-488,
1:300, 90 mins
Mounting
Eye enucleation and
dissection (remove
cornea and lens)
Fixation in 4%
paraformaldehyde,
2 hours, 0.4M PB
Cryoprotection in
30% sucrose,
0.1M PB
Store retina in -
70
o
C until use
Day 1
Embed eyecup in
OCT compound and
fast frozen
Rinse 3 times, 5
mins, 0.01M PBS
Rinse 3 times, 5
mins, 0.1M PB
Cryostat sectioning,
25-µm
Retinal
preparation
Immunocyto
chemistry
Immunocyto
chemistry

Figure 4.1 Experiment protocol for vertical section immunocytochemistry.
93

Whole Mount Immunocytochemistry
Take out retina and
thaw in room
temperature
Make 5 relaxing
cuts
Remove retina from
sclera and RPE and
vitreous
Obtain neural
retina
Rinse 3 times, 10
mins, 0.01M PBS
Treat with 1%
Trion, 40 mins
Rinse 3 times, 10
mins, 0.01M PBS
Incubate in NDS,
1:10, 1 hour
Apply primary
antibody, SMI-32,
1:1500
Day 5
Put onto shaker,
30 mins
Rinse 3 times, 10
mins, 0.01M PBS
Apply secondary
antibody, Alexa-488,
1:300
Put onto shaker, 20
mins, cover with foil
Day 9
Rinse 3 times, 10
mins, 0.1M PB
Mounting
Eye enucleation and
dissection (remove
cornea and lens)
Fixation in 4%
paraformaldehyde,
2 hours, 0.4M PB
Cryoprotection in
30% sucrose,
0.1M PB
Store retina in -
70
o
C until use
Day 1
Retinal
preparation
Immunocyto
chemistry
Immunocyto
chemistry
Immunocyto
chemistry

Figure 4.2 Experiment protocol for whole mount immunocytochemistry.
94

immunocytochemistry marker SMI-32. SMI-32 is a neurofilament antibody which labels
axon and dendrites. First, we performed immunocytochemistry in vertical section of P21
normal rat retina, as shown in Figure 4.3A. SMI-32 immunoreactivity was observed only
in cells in the ganglion cell layer and the processes in the IPL. In addition, nerve fiber
layer was labeled. Here we confirmed that the cells we labeled with SMI-32 were only
retinal ganglion cells. Hence, we counted the SMI-32 immuno-reactive retinal ganglion
cells in the subsequent section. Faintly labeled ganglion cells, as shown in Figure 4.3B,
have been reported in human retina in an earlier study (Straznicky, Vickers et al. 1992).
We estimated the mean density of retinal ganglion cells to observe the distribution
pattern along the nasal-to-temporal and ventral-to-dorsal meridians in whole mount
retinas, as shown in Figure 4.4A. We used five retinas from control, four retinas from
young rd group and seven retinas from old rd group. In the nasal-to-temporal meridian of
control retinas, the density of immune-reactive retinal ganglion cell density was 375 ± 42
cells/mm
2
around the optic disc and 320 ± 32 cells/mm
2
in the peripheral region.
Similarly, the density was 300 ± 49 cells/mm
2
around the optic disc and 365 ± 35
cells/mm
2
in the peripheral region in young rd group. In comparison, in old rd group, the
antibody labeled 288 ± 49 cells/mm
2
around the optic disc and 156 ± 15 cells/mm
2
in the
peripheral region. In turn, in the ventral-to-dorsal meridian, of control retinas, the SMI-32
labeled ganglion cell density was 428 ± 50 cells/mm
2
around the optic disc and 296 ± 26
cells/mm
2
in the peripheral region. Similarly, in young rd group, the density was 431 ± 49
cells/mm
2
around the optic disc and 400 ± 41 cells/mm
2
in the peripheral region. In
95

contrast, in old rd group, the antibody labeled 246 ± 34 cells/mm
2
around the optic disc
and 165 ± 19 cells/mm
2
within periphery. Consequently, in the nasal-temporal meridian,
densities in young rd retinas were not significantly different from those from control
retinas while densities in old rd retinas within the periphery were lower than the control
retinas. In the ventral-dorsal meridian, densities in the young rd retinas were not
significantly different from those from control retinas while densities in old rd retinas
were significantly lower than the control retinas. A summary graph of ganglion cell
densities was shown in Figure 4.3C, indicating that there was a great reduction of
ganglion cell densities in old rd group 241 ± 15 cells/mm
2
versus those in normal group
379 ± 16 cells/mm
2
and in young rd group 374 ± 24 cells/mm
2
. The cell densities,
counterstained with TOPRO-3, were 5975 ± 975 cells/mm
2
in P649 control retina and
3800 ± 386 cells/mm
2
in P698 rd retina, as shown in Figure 4.5. Similar TOPRO-3 results
were reported (Jakobs, Libby et al. 2005). TOPRO-3 stains all nuclei, so displaced
amacrine cells will also be stained.
We counted the total number of SMI-32 immuno-reactive ganglion cells between
control and old rd groups to calculate the percentage of cell reduction. For this counting,
we split the old rd groups in two different age groups, P500 group and P700 group. The
total number of SMI-32 immuno-reactive cells in P500 control retina was 5580 ± 1104
(n=2) and in P700 control retina was 4446 ± 639 (means ± SE; n=2). In contrast, we
observed 3381 ± 510 cells in P500 (means ± SE; n=2) and 1994 ± 238 cells in P700
96

(means ± SE; n=2) rd retinas. Hence, SMI-32 immuno-reactive cells were decreased by
39.4% in P500 and 55.2% in P700 compared to age-matched control retinas.

Figure 4.3 A: SMI labels ganglion cells and processes in the IPL (normal P21). B.
SMI-labeled ganglion cells in GCL of control retina (P522). Arrows indicated the faintly
labeled ganglion cells. C: Few SMI-labeled ganglion cells in GCL S334ter retina (P698).
N denotes normal/control retina while RD denotes degenerate retina.
97

Figure 4.4 Reduction of mean density of old degenerate retinas along both nasal-to-
temporal (A) and ventral-to-dorsal (B) meridians. C: Reduction of total number of SMI-
32 reactive ganglion cells in P500 and P700 rd retinas. D: Reduction of cell density of
SMI-32 reactive ganglion cells in P500 and P700 rd retinas. Cell density of P300 rd
retinas was similar to control group P500 and P700.

98

Figure 4.5 High-power view of normal and degenerate retina. The whole mount was
stained for the neurofilament markder SMI-32, which labels axons and a population of
ganglion cells (green). The cell nuclei are stained with TOPRO-3 (blue). A: P649 normal
retina, central. B: P649 normal retina, peripheral. C: P698 degenerate retina, central. D:
P698 degenerate retina, peripheral.
99

4.4 Conclusions and Discussions
Using immunocychemical labeling technique, the death of retinal neurons in the ganglion
cell layer has been evaluated in a rat model of retinal degeneration. SMI-32 labels almost
40% of the whole population of retinal ganglion cell in mice model. Neurofilament
staining using SMI-32 and total neuron counts provide strong evidence for ganglion cell
loss in old degenerate retinal neurons. We concluded that retinal ganglion cells were
vulnerable to degeneration in S334ter-line-3 rat model.
Combining the result obtained in this chapter and the previous chapter, in which
the occurrence of threshold increase in old rd group has been demonstrated, the phase of
degeneration and the capability to elicit electrical response are correlated. Apart from cell
deaths in ganglion cell layer, remodeling in late stage of degeneration continues. Detailed
morphological alterations in retinal neurons, such as changes in dendritic tree, are not
examined in this chapter.
100

Chapter 5: Spatial Properties of Electrical Excitation
This study investigated the spatial response of the superior colliculus induced by
electrical stimulation and photic stimulation in a normal rat model. The goal was to
compare the pattern of activity evoked by focal light and electrical stimuli. This
information may help optimize our design by reducing the effect of overlapping of
activation area, which could limit resolution.
5.1 Background
Electrical stimulation has been used to restore the loss of sensory function. Retinal
implants, to restore visual function to the blind, stimulate hundreds of retinal cells using a
multi-electrode array configuration. Current implants allow otherwise blind individuals to
locate objects and detect motion. Future implants are being designed to provide vision for
mobility and reading. To achieve high resolution vision with electrical stimulation,
spatial properties of neural excitability and current spread deserve a detailed examination,
so that adjacent electrodes can stimulate distinct populations of neurons without overlap.
The current image-processing algorithm in prosthesis work uses amplitude modulation as
the decoding method to transform brightness to electrical stimulation parameters.
However, there exists a trade-off between the stimulus intensity and the phophene size
due to the spread of excitation as the stimulus intensity increases (Nanduri, Humayun et
al. 2008). If large stimulus intensity is used to convey brightness, it may also result in a
large area of retinal activation, which weakens the degree of resolution. On the other
hand, one of the few clinical trials showed that the phosphene brightness is dependent on
101

electrical stimulus intensity and frequency (Fujikado, Morimoto et al. 2007). However,
correlation between electrical stimulation parameters and degree of brightness perception
has not been well studied. This chapter aims at investigating the effect of the stimulus
parameter (current amplitude) on the activation area in superior colliculus, a high visual
center in rodent model which possesses a well-organized retinotopic projection (see
Chapter 1 Session 1.2). First, focal stimulation of electrical and photic stimulation will be
demonstrated to show the feasibility of this experimental design. Second, the activated
area as a function of stimulus current will be measured. Finally, the limitation of the
current approach and plausible future work will also be discussed.
5.2 Materials and Methods
(For procedures of surgery, insertion of stimulation electrode, and impedance testing, see
Chapter 3.2)
5.2.1 Animal
Normal Copenhagen (P90, n = 5) rats were used in these studies. Animals were
housed in covered cages and were fed with a standard rodent diet ad libitum, while kept
on a 12-h light: 12-h dark cycle in the Doheny Eye Institute animal facility. All
experimental procedures were done in accordance with the guidelines of Institutional
Animal Care and Use Committee at University of Southern California for animal care
and followed the ARVO guidelines for animal usage.

102

5.2.2 Light Stimulation
The light stimuli was generated under Matlab, using the Psychophysics Toolbox
extensions (Brainard 1997) and presented on a monitor at 60 Hz in 1024 x 768 resolution.
The display was placed 30 cm in front of the left eye of the animal and centered on its’
left eye midline, thereby covering 45 x 40
o
of the visual field. The background was
uniformly dark gray. The light stimulus is a small spot subtending 1 to 3
o
. White square
over a black background were flashed for 500 ms with 4500 ms interval. Stimulus was
repeated 50 times for each mapping location. Mean luminance of the background and
stimulus was measured.
5.2.3 Experiment Protocol
The protocol of this study is illustrated in Fig. 5.1. The experiment was divided
into two stages, focal activation of light and electrical stimulation, and spatial properties
as stimulus current increases. In Stage I, the electrically-evoked response was first
recorded in the SC at the location with the lowest threshold, following by mapping the
corresponding visual space using a light stimulus (recording location 200- μm apart in a
5x5 grid corresponding to 25 recording locations). In Stage II, SC mapping of electrical
stimulation and light stimulation was performed with two levels of current amplitudes of
the electrical stimulus. As described in earlier chapters, impedance was used to monitor
electrode-retina proximity throughout the experiment. Threshold was defined as the
amount of current required for the excitation of collicular response above a 5:1 signal-to-
noise ratio in 10 of 10 trials.
103

Insert and
position
electrode
Pick SC
location
Is SC region with
lowest threshold ?
Move
recording
electrode
Map visual
space
Is SC region
activated ?
Change pulse
intensity
Regular control:
Electrode potential ,
electrode impedance
Move light spot
No
Yes
No
Yes
Stage I
Locate same spot of
electrical and visual
activation
Stage II
Spatial
measurement
Apply electrical
stimulation (same
parameters as in Stage I)
Apply visual stimulation
(same parameters as in
Stage I)
Is SC region
activated above
threshold ?
Move recording
electrode (200
µm apart)
Finish electrical
and visual mapping
No
Yes

Figure 5.1 Experiment protocol for spatial spread measurement of electrical and light
activation.
104

5.2.4 Data Analysis

Light evoked responses contained two components, representing responses to the
onset of light stimulus (“On” response) or to the offset of the stimulus (“Off” response).
In the present study, we focused on the Off responses, which were usually larger than the
On responses. Local field potentials (LFPs) and multi-unit activities (MUAs) were
recorded. Raster plot, post-stimulus time histogram (PSTH, bin width, 5 ms), inter-pulse
interval for the evoked responses generated were measured and plotted. The signal-to-
noise ratio of a response were defined as 5:1. For spatial properties, areas activated by
each of the two current intensity levels were compared. The spread of excitation from
light stimulus and electrical stimulus were compared.
5.3 Results
5.3.1 Focal Activation by Light Stimulus
In order to study the response to light and electrical stimulation, the size of the
stimulus needs to be comparable for analyzing a focal stimulation. In electrical
stimulation, preliminary result in Chapter 3 Section 3.1 has shown that only a small area
in SC was being activated, indicating a focal stimulation area. On the other hand, a visual
field can be mapped on the SC surface in an early study (Siminoff, Schwassmann et al.
1966). In the experimental setup, two spatially different light stimuli activated two
distinct locations on SC was shown in Fig. 5.2. Two spots in the rat’s visual field were
stimulated (S:01 and S:02) and two spots of light responses in the rat’s superior colliclus
(SC) were recorded (R:01 and R:02). SC responses were recorded in R:01 when S:01
105

was illuminated, as shown in Fig. 5.2A. No SC responses were recorded in a different
spot, R:02, when S:01 was illuminated, as shown in Fig. 5.2B. SC responses were
recorded in R:02 when the light stimulus was moved to another spot, S:02, as shown in
Fig. 5.2C.
5.3.2 Focal Activation by Light and Electrical Stimulus Simultaneously
We have shown that we could map electrical stimulus (Chapter 3 Section 3.1) and
we have repeated this for light stimulus, as shown in Fig.5.2. In this experiment, we
showed that we could successfully activate the same SC region by light and electrical
stimulation, as shown in Fig. 5.3. The large rectangular box represents the monitor. The
small box inside the rectangular box represents the light stimuli, which projects to the
temporal-ventral quadrant on the left retina. The hemispherical shape represents the right
SC. The filled crosses represent the locations of the light responses. The empty crosses
represent locations of the recording electrode without responses, as shown in the bottom
right corner in Fig. 5.3A. The light responses were recorded in the medial-rostral
quadrant in the right SC, corresponding to the retinotopic projection. The response area
subtended to an area of 800 x 1200 µm
2
in SC. Responses to electrical stimulation in the
temporal-ventral quadrant on the left retina activated the same area in right SC, as shown
in Fig. 5.3B. SC responses were only recorded in the medial-rostral region of the SC,
indicating by filled crosses, while no responses were recorded in other areas of the right
SC. Multi-unit activities, and sometimes, single-unit activities were recorded with
reasonable signal-to-noise ratio, as shown in Fig. 5.4.
106

Figure 5.2 Responses to light spot. A: There were SC responses at recording location
1 when spot 1 was illuminated. B: No SC responses at recording location 2 when spot 1
was illuminated. C: There were SC responses at recording location 2 when spot 2 was
illuminated. S – Light Stimuli. R – recording location. Scale bar = 1 second.
S:01
R:01
S:01
R:02
S:02
R:02
A B
C
A
B
C
107

Figure 5.3 A: Responses to light stimulation (recording locations are 200 μm apart).
The responses within the yellow rectangle inside the hemispherical shape are represented
by the peristimulus time histogram (PSTH) B: Responses to electrical stimulation. SC
responses were recorded in four locations (represented by empty crosses).
A
B
Light Stimuli
108

-0.5 0 0.5 1 1.5
-150
100
EER2
Time (ms)
Amplitude (μV)
-0.5 0 0.5 1 1.5
-150
100
EER
Time (ms)
Amplitude (μV)
Figure 5.4 Superimposed single-unit activities (2 ms) from light stimulation (A) and
electrical stimulation (B) on the same recording location.

5.3.3 Spatial Spread Increases with Increased Stimulus Currents
The goal of this section was to obtain spatial maps of response in the rat superior
colliculus elicited by epiretinal stimulation. The excitability and current spread properties
of the stimulated neurons were obtained by measuring the extent of excitation on the
superficial superior colliculus (SC) as a function of current levels applied to the retina
using a single electrode configuration. Three normal rats were used. The retinocollicular
projection is well organized in adult rat where the tempo-nasal axis on the retina
corresponds to the antero-posterior axis on the SC. Two different levels of 1-ms biphasic
currents were delivered through a bipolar concentric electrode, 75 um inner diameter and
300 um outer diameter. Electrical evoked responses were obtained from the contralateral
SC in 25 recording locations, 200 μm apart, in a 5x5 grid. Twice the threshold current
(2X threshold) elicited a neural excitation area represented by 17 pixels in 5x5 grid, while
1X threshold elicited a neural excitation area represented by 3 pixels in 5x5 grid. The
data suggest that there is a disproportional increase in the excitation area relative to the
applied current.
A B
N = 24
N = 84
109

Figure 5.5 Spatial Spread Increases with Increased Stimulus Currents
5.4 Conclusions and Discussions
In this chapter, the spatial characteristics of the SC response to electrical and light
stimulation have been demonstrated. Light and electrical stimulation activation on the
same focal region on the SC has also been shown. Using a single stimulating electrode,
there was a disproportional increase of the activation area in SC in response to the
applied current, suggesting that the distance between adjacent electrodes in the
microelectrode array should be 2X larger than the size of the electrode in order to reduce
overlapping of the activated region which may reduce the resolution.
This experiment was time-inefficient because single recording electrode was used.
The experiment can be improved by using a microelectrode array or intrinsic optical
imaging. Results will be obtained more efficiently with multi-array recording for spatial
mapping of the responses. Larger area can also be evaluated (more than 25 locations) for
110

the mapping results. On the other hand, one of the few clinical trials showed that the
phosphene brightness is dependent on electrical frequency (Fujikado, Morimoto et al.
2007). The model and experimental design employed in this chapter can be used to
investigate the effect of the stimulus frequency on the activation area due to electrical
stimulation.
111

Chapter 6: General Conclusions
The work presented in this thesis investigated various aspects of electrical stimulation of
the degenerated rat retina to contribute to the development of a high-resolution retinal
prosthesis. This chapter briefly summarizes the key findings of this research, a proposed
mechanism to explain the threshold data, and recommendations for future research
directions that can provide useful insights towards safe and effective high resolution
stimulation, which can allow the mobility and facial recognition in the associated
diseased community.
6.1 Key Findings
The key findings are summarized by the following points:
1) A qualitative connection between degeneration state and stimulus threshold has
been shown. Aged degenerate rat retinas (P500 – 700) require high charge
stimulation to evoke a response in the high visual pathway compared to normal
healthy retinas and young degenerate rat retinas (P100 – 300).
2) The charge density required to elicit a response in aged degenerate retinas using a
small electrode (75-µm) stays within the traditional electrochemical safety limit
of platinum electrode using short pulse duration (<0.5 ms).
3) The observation of the same chronaxie values among all control and degenerate
retinas suggests that the same cell type, i.e. retinal ganglion cell, is being
stimulated regardless of the phase of degeneration.
112

4) Retinal ganglion cell density and total cell count are reduced in aged degenerate
retina compared to age-matched normal healthy retinas and young degenerate
retinas.
5) Aged S334ter-line-3 rat model resembles the end-stage retinitis pigmentosa
condition in human in terms of retinal ganglion cell reduction and requirement of
high charge stimulation
6) Measurement of electrochemical impedance at the retinal-electrode interface at 8
kohm at 100 kHz has been standardized to assess the effectiveness of electrical
stimulation without mechanical destruction of the retina using in vivo method.
7) Using a single stimulating electrode, there was a disproportional increase of the
activation area in SC in response to the applied current, suggesting that the
distance between adjacent electrodes in the microelectrode array should be 2X
larger than the size of the electrode in order to reduce overlapping of the activated
region.
6.2 Future Directions
This work has contributed to the development of a clinical device to partially restore
vision in humans suffering from retinitis pigmentosa and aged related macular
degeneration. Electrical stimulation on aged, degenerate retina and recording in higher
visual pathway could have significant advantages over currently available methods for
assessment of safety and efficacy. The most significant contribution from this work is the
delineation of stimulus parameter range that can safely evoked responses, even from
113

severely degenerated retina. Importantly, the threshold values (in terms of charge density)
are similar to what has been previously reported in attempts to evoke response in human
RP (Humayun, Weiland et al. 2003; Chan, Ray et al. 2008). The finding that retinal
ganglion cell density correlates with threshold whereas bipolar cell density does not is
important for future research. This work provides a basis for additional areas of research
that must be addressed in order to continue moving towards a development of a safe and
effective device for a high resolution retinal prosthesis.
6.2.1 Charge Densities as a Function of Small Electrodes
This study has demonstrated the low threshold and low charge requirement of
stimulating young degenerate retinas and normal healthy retinas using a small electrode
(75-µm) with short pulses. Hence, the charge densities stay within the electrochemical
safety limit of platinum electrode of 0.35 mC/cm
2
. This study also demonstrated that
using short pulses, the charge densities required for stimulating aged degenerate retina
stay within charge density of platinum electrode. It has been shown that the thresholds
were the same for 260- and 520-µm electrodes in a clinical study, implying that the
charge density for the 260-µm electrode is higher. It is unknown to us that if the threshold
or charge requirement will remain the same using an even smaller electrode, i.e. 75-µm in
1000 electrode array (Weiland, Fink et al. 2005). By employing the animal model and the
impedance method that have been developed in this study, it can provide more
information to the safety aspect of using small electrodes in the pre-clinical study.

114

6.2.2 Investigation on Safety Limits of Short Pulses

Charge density has a positive correlation with pulse duration. It has been
proposed that platinum has a conservative, safe, and long-term stimulation limit of 0.3 -
0.35 mC/cm
2
using 0.6-ms pulse or above. Using shorter pulse duration, one study has
proposed that the charge density limit of platinum reduces to 0.05 – 0.15 mC/cm2. This
study demonstrated that the stimulation with short pulse does not exceed the stimulation
limit of 0.3 - 0.35 mC/cm
2
, but does not meet the stricter limit for any pulse duration. The
reversible reactions are incomplete during the short pulse, resulting in low charge limits
(Rose and Robblee 1990). Dissolution may occur and lead to erosion of the electrode
and/or damage to the retina. Therefore, it is worthwhile to investigate the damaging effect
to the retina using short pulse in an aged degenerate animal model in vivo in order to fully
characterize the safe limit of stimulation in a high resolution design.
6.2.3 Remodeling in Late Degenerate Stage and High Visual Pathway
The loss of retinal ganglion cell alone may not fully explain the alteration of the
response properties in this study. It is well known that remodeling is common in central
nervous system (CNS) pathways, triggered by pre-synaptic cell death. There is no reason
to eliminate the possible fact that the remodeling of the SC may contribute to the
increased activation threshold. The CNS plasticity in response to disease is unavoidable.
Since studies on the effect of the CNS plasticity to the stimulus requirement are few, such
investigations represent an opportunity for future research. Thus, it is worthwhile to study
115

the effect of the remodeling in higher visual pathway at the end-stage of RP on the
stimulation parameters.
6.3 Proposed Model
Here, two models are proposed to explain the increase of activation threshold, from
young rd group to old rd group. The assumption is that the main factor contributing to
this threshold change is the decreased number of retinal ganglion cells. At this point, this
assumption is somewhat speculative, since other changes in RGC physiology or changes
in SC may also contribute to threshold increase. However, currently we do not have as
much information on either of those two areas to propose a model, but we have, in this
study, made direct measurement of ganglion cell density.
In the first model, let
i
P be the probability of firing of a cell, where n i ,....., 3 , 2 , 1 = ,
for n cells. If
i
P is the same for all cells and follows a Weibull distribution, the
cumulative distribution function for the Weibull distribution is ()
()
k
x
e k x P
λ
λ
/
1 , ;
−
− = for
0 ≥ x , and () 0 , ; = λ k x P for 0 < x . For n cells,
n
P P P U U U ....
2 1
is equal to
∑
n
i
i
P . A
Weibull distribution function of the percent firing versus the current applied is plotted
below. As n increases, where n denotes the number of cells being activated, the curve,
as shown in Figure 6.1, will shift to the left, suggesting a decrease in applied current. It
will need lower current to obtain the same percent firing in a population of cells. In the
second model, the probability of firing is again described by the Weibull distribution.
Here, the results in Chapter 3 were incorporated in this model, in which the current
116

requirement is higher in old rd group compared to the control and young rd group, using
the same threshold definition (probability of firing, say 75% as defined in Chapter 3).
Then, a group of Weibull curves will be obtained, as shown in Figure 6.2A. It is well
known that the current applied is proportional to the potential gradient of the applied
electric field to the target tissue, meaning the activation area increases with applied
current. To examine the probability of firing as a function of degeneration stage, a
constant activation area is assumed. If the potential gradient or the current applied is
fixed as 8 µA for all groups (a vertical line is drawn in Figure 6.2B), the probability of
firing in each group can be related to the different degenerate stages, in which the number
of ganglion cells vary (translated to Figure 6.3A). As discussed in Chapter 4, if there are
more cells within the activation area, the probability of firing is higher. Hence, the
current requirement is lower. In contrast, the current requirement is higher when there are
few cells within the activation area because the probability of firing is lower, illustrated
in Figure 6.3A and 6.3B.

*Matlab script of Figure 6.1A
* x=1:0.5:20; k=5; lambda=10;
* f=1-exp(-(x./lambda).^k);
* figure; plot(x,f);
* text(10.5,0.6,['\leftarrowF = 1-exp(-(x/\lambda)^k)']);
* text(11.5,0.5,['k=5, \lambda=10']);

117

* Matlab script of Figure 6.1B
* x=1:0.5:20; k=5; lambda=7.5;
* f=1-exp(-(x./lambda).^k);
* hold on; plot(x,f);
* text(5,0.6,['\leftarrow']);
* text(2,0.7,[‘n increases’]);

Firing threshold is a function of membrane potential, how well the electrode
coupled to cell and the synaptic input to that cell. Here, the aim of proposing these
models is not to go through and fully develop a model and create a predictive tool, which
requires a more thorough experiment. Instead, a qualitative interpretation of the results
obtained has been proposed. This model proposed that a group of cells are being
stimulated. The current requirement is correlated with the number of cells presented
within the activation area, and as a function of the coupling between the electrode and the
target cells.
118

Figure 6.1 A: Proposed Weibull curves of probability of firing as a function of
current applied. B: Curve is shifted to the left when number of cells, being activated,
increases
119

Figure 6.2 A: Weibull curves of control, rd P500 and rd P700 groups. B: Under same
applied current (same activation area), the probability firing varies with different
degenerate stages, with different number of cells.
120

Figure 6.3 A: Probability of firing is plotted against the number of inputs. Number of
inputs is inversely correlated with the degenerate stage, as shown in Figure 6.2B. B:
Requirement of current threshold is plotted agains the number of inputs. Numbersof
inputs are representative numbers, not revealing the exact number of cells in each group.
121

6.4 Implications for High Resolution Retinal Prosthesis
The current study has two major implications for the development of retinal prostheses.
First, it is electrochemically safe to stimulate a severe degenerate retina using a short-
duration pulse (< 500 µs) using platinum electrode with 75-µm diameter. High current
requirement in degenerate retina may limit the use of smaller, densely packed electrode
array, which occurs in long pulse duration in the current study. Therefore, employing
short-duration pulses may lead to the use of smaller electrodes that would benefit
individuals which suffer from RP, by increasing the resolution afforded by the implant.
Second, reduction of retinal ganglion cells may play a role in the elevation of the
activation threshold, suggesting that the placement of the microelectrode array is crucial
in maintaining a low current stimulation paradigm.
The work presented in this thesis represents the most comprehensive study
relating both retina structure and function to electrical stimulation threshold. The animal
model used in this study closely resembles the end stage of autosomal dominant RP in
human, particularly in the trend of loss of retinal ganglion cells and the requirement of
the current discharge which have been reported in the clinical trials. This study suggests
that electrodes with 75-µm could evoke a response within the safe limit in a limited pulse
range. This is a very encouraging result and supports the continued development of a
high resolution retinal prosthesis for the blind.

122

Bibliography

Agnew, W. F., T. G. Yuen, et al. (1986). "Histopathologic evaluation of prolonged
intracortical electrical stimulation." Exp Neurol 92(1): 162-85.

Aguirre, G. D. and L. F. Rubin (1975). "Rod-cone dysplasia (progressive retinal atrophy)
in Irish setters." J Am Vet Med Assoc 166(2): 157-64.

Ahuja, A. K., M. R. Behrend, et al. (2008). "An in vitro model of a retinal prosthesis."
IEEE Trans Biomed Eng 55(6): 1744-53.

Andersen, R. A. (2008). How we see. Aerospace and Electronic Systems Magazine, IEEE.
23: 6.

Beebe, X. and T. L. Rose (1988). "Charge injection limits of activated iridium oxide
electrodes with 0.2 ms pulses in bicarbonate buffered saline." IEEE Trans Biomed
Eng 35(6): 494-5.

BeMent, S. L. and J. B. Ranck, Jr. (1969). "A model for electrical stimulation of central
myelinated fibers with monopolar electrodes." Exp Neurol 24(2): 171-86.

Berson, E. L., B. Rosner, et al. (1993). "A randomized trial of vitamin A and vitamin E
supplementation for retinitis pigmentosa." Arch Ophthalmol 111(6): 761-72.

Bertschinger, D. R., E. Beknazar, et al. (2008). "A review of in vivo animal studies in
retinal prosthesis research." Graefes Arch Clin Exp Ophthalmol 246(11): 1505-17.

Besch, D., H. Sachs, et al. (2008). "Extraocular surgery for implantation of an active
subretinal visual prosthesis with external connections: feasibility and outcome in
seven patients." Br J Ophthalmol 92(10): 1361-8.

Bi, A., J. Cui, et al. (2006). "Ectopic expression of a microbial-type rhodopsin restores
visual responses in mice with photoreceptor degeneration." Neuron 50(1): 23-33.

Bok, D. (2005). "Ciliary neurotrophic factor therapy for inherited retinal diseases: pros
and cons." Retina 25(8 Suppl): S27-S28.

Brainard, D. H. (1997). "The Psychophysics Toolbox." Spat Vis 10(4): 433-6.

Brummer, S. B., L. S. Robblee, et al. (1983). "Criteria for selecting electrodes for
electrical stimulation: theoretical and practical considerations." Ann N Y Acad
Sci 405: 159-71.
123

Brummer, S. B. and M. J. Turner (1977). "Electrical stimulation with Pt electrodes: II-
estimation of maximum surface redox (theoretical non-gassing) limits." IEEE
Trans Biomed Eng 24(5): 440-3.

Caley, D. W., C. Johnson, et al. (1972). "The postnatal development of the retina in the
normal and rodless CBA mouse: a light and electron microscopic study." Am J
Anat 133(2): 179-212.

Canola, K., B. Angenieux, et al. (2007). "Retinal stem cells transplanted into models of
late stages of retinitis pigmentosa preferentially adopt a glial or a retinal ganglion
cell fate." Invest Ophthalmol Vis Sci 48(1): 446-54.

Chan, L. H., A. Ray, et al. (2008). "In vivo study of response threshold in retinal
degenerate model at different degenerate stages." Conf Proc IEEE Eng Med Biol
Soc 2008: 1781-4.

Chen, S. J., M. Mahadevappa, et al. (2006). "Neural responses elicited by electrical
stimulation of the retina." Trans Am Ophthalmol Soc 104: 252-9.

Chopdar, A., U. Chakravarthy, et al. (2003). "Age related macular degeneration." BMJ
326(7387): 485-8.

Colodetti, L., J. D. Weiland, et al. (2007). "Pathology of damaging electrical stimulation
in the retina." Exp Eye Res 85(1): 23-33.

Coombs, J., D. van der List, et al. (2006). "Morphological properties of mouse retinal
ganglion cells." Neuroscience 140(1): 123-36.

Cowey, A. and E. T. Rolls (1974). "Human cortical magnification factor and its relation
to visual acuity." Exp Brain Res 21(5): 447-454.

Curcio, C. A. and K. A. Allen (1990). "Topography of ganglion cells in human retina." J
Comp Neurol 300(1): 5-25.

Curcio, C. A., K. R. Sloan, et al. (1990). "Human Photoreceptor Topography." J Comp
Neurol 292(4): 27.

de Balthasar, C., S. Patel, et al. (2008). "Factors affecting perceptual thresholds in
epiretinal prostheses." Invest Ophthalmol Vis Sci 49(6): 2303-14.
DeMarco, P. J., Jr., G. L. Yarbrough, et al. (2007). "Stimulation via a subretinally placed
prosthetic elicits central activity and induces a trophic effect on visual responses."
Invest Ophthalmol Vis Sci 48(2): 916-26.
124

Duan Y.Y., C. G. M., Cowan R.S.C. (2003). "A study of intra-cochlear electrodes and
tissue interface by electrochemical impedance methods in vivo." Biomaterials 25:
3813-3828.

Dymond, A. M. (1976). "Characteristics of the metal-tissue interface of stimulation
electrodes." IEEE Trans Biomed Eng 23(4): 274-80.

Eckhorn, R., M. Wilms, et al. (2006). "Visual resolution with retinal implants estimated
from recordings in cat visual cortex." Vision Res 46(17): 2675-90.

Euler, T. and H. Wassle (1995). "Immunocytochemical identification of cone bipolar
cells in the rat retina." J Comp Neurol 361: 8.

Fried, S. I., H. A. Hsueh, et al. (2006). "A method for generating precise temporal
patterns of retinal spiking using prosthetic stimulation." J Neurophysiol 95(2):
970-8.

Fujikado, T., T. Morimoto, et al. (2007). "Evaluation of phosphenes elicited by
extraocular stimulation in normals and by suprachoroidal-transretinal stimulation
in patients with retinitis pigmentosa." Graefes Arch Clin Exp Ophthalmol 245(10):
1411-9.

Gekeler, F., A. Messias, et al. (2006). "Phosphenes electrically evoked with DTL
electrodes: a study in patients with retinitis pigmentosa, glaucoma, and
homonymous visual field loss and normal subjects." Invest Ophthalmol Vis Sci
47(11): 4966-74.

Gerding, H., F. P. Benner, et al. (2007). "Experimental implantation of epiretinal retina
implants (EPI-RET) with an IOL-type receiver unit." J Neural Eng 4(1): S38-49.

Girman, S. V. and R. D. Lund (2007). "Most superficial sublamina of rat superior
colliculus: neuronal response properties and correlates with perceptual figure-
ground segregation." J Neurophysiol 98(1): 161-77.

Girman, S. V., S. Wang, et al. (2005). "Time course of deterioration of rod and cone
function in RCS rat and the effects of subretinal cell grafting: a light- and dark-
adaptation study." Vision Res 45(3): 343-54.

Grafstein, B., M. Murray, et al. (1972). "Protein synthesis and axonal transport in retinal
ganglion cells of mice lacking visual receptors." Brain Res 44(1): 37-48.

125

Greenberg, R. (1998). Analysis of electrical stimulation of the vertebrate retina: work
towards a retinal prosthesis. Baltimore, The Johns Hopkins University. Ph.D.

Greenberg, R. J. (1998). Analysis of Electrical Stimulation of the Vertebrate Retina-
Work Towards a Retinal Prosthesis. The Wilmer Ophthalmological Institute.
Baltimore, Johns Hopkins University. Doctoral.

Guven, D., J. D. Weiland, et al. (2005). "Long-term stimulation by active epiretinal
implants in normal and RCD1 dogs." J Neural Eng 2(1): S65-73.

Guyton, D. L. and F. T. Hambrecht (1973). "Capacitor Electrode Stimulates Nerve or
Muscle without Oxidation-Reduction Reactions." Science 181(4094): 74.

Hartong, D. T., E. L. Berson, et al. (2006). "Retinitis pigmentosa." Lancet 368(9549):
1795-809.

Hayes, J. S., V. T. Yin, et al. (2003). "Visually guided performance of simple tasks using
simulated prosthetic vision." Artif Organs 27(11): 1016-28.

Hodgkin, A. L. and A. F. Huxley (1952). "A quantitative description of membrane
current and its application to conduction and excitation in nerve." J Physiol 117(4):
500-44.

Holsheimer, J. (2003). Principles of neurostimulation. Electrical Stimulation and the
Relief of Pain. B. A. Simpson, Elsevier Science. 15.

Hu, D. N., J. D. Simon, et al. (2008). "Role of ocular melanin in ophthalmic physiology
and pathology." Photochem Photobiol 84(3): 639-44.

Humayun, M., R. Propst, et al. (1994). "Bipolar surface electrical stimulation of the
vertebrate retina." Arch Ophthalmol 112(1): 110-6.

Humayun, M. S., M. Prince, et al. (1999). "Morphometric analysis of the extramacular
retina from postmortem eyes with retinitis pigmentosa." Invest Ophthalmol Vis
Sci 40(1): 143-8.

Humayun, M. S., J. D. Weiland, et al. (2003). "Visual perception in a blind subject with a
chronic microelectronic retinal prosthesis." Vision Res 43(24): 2573-81.
Huxlin, K. R. and A. K. Goodchild (1997). "Retinal ganglion cells in the albino rat:
revised morphological classification." J Comp Neurol 385(2): 309-23.

Jakobs, T. C., R. T. Libby, et al. (2005). "Retinal ganglion cell degeneration is
topological but not cell type specific in DBA/2J mice." J Cell Biol 171(2): 313-25.
126

Jeffery, G. (1997). "The albino retina: an abnormality that provides insight into normal
retinal development." Trends Neurosci 20(4): 165-9.

Jensen, R. J. and J. F. Rizzo, 3rd (2008). "Activation of retinal ganglion cells in wild-type
and rd1 mice through electrical stimulation of the retinal neural network." Vision
Res 48(14): 1562-8.

Jensen, R. J., J. F. Rizzo, 3rd, et al. (2003). "Thresholds for activation of rabbit retinal
ganglion cells with an ultrafine, extracellular microelectrode." Invest Ophthalmol
Vis Sci 44(8): 3533-43.

Jensen, R. J., O. R. Ziv, et al. (2005). "Thresholds for activation of rabbit retinal ganglion
cells with relatively large, extracellular microelectrodes." Invest Ophthalmol Vis
Sci 46(4): 1486-96.

Johnson, M. D., K. J. Otto, et al. (2005). "Repeated voltage biasing improves unit
recordings by reducing resistive tissue impedances." IEEE Trans Neural Syst
Rehabil Eng 13(2): 160-5.

Jones, B. W. and R. E. Marc (2005). "Retinal remodeling during retinal degeneration."
Exp Eye Res 81(2): 123-37.

Jones, B. W., C. B. Watt, et al. (2003). "Retinal remodeling triggered by photoreceptor
degenerations." J Comp Neurol 464(1): 1-16.

Kanda, H., T. Morimoto, et al. (2004). "Electrophysiological studies of the feasibility of
suprachoroidal-transretinal stimulation for artificial vision in normal and RCS
rats." Invest Ophthalmol Vis Sci 45(2): 560-6.

Kandel, E. R. (1991). Principles of Neural Science. Visual Processing by the Retina,
McGraw-Hill: 507-522.

Karli, P. (1952). "[Retinae without visual cells; morphological, physiological and
physiopathological research in rodents.]." Arch Anat Histol Embryol 35(1-2): 1-
76.

Kato, S., M. Saito, et al. (1983). "Response of the visual system evoked by an alternating
current." Med Biol Eng Comput 21(1): 47-50.
Kim, S. Y., S. Sadda, et al. (2002). "Morphometric analysis of the macula in eyes with
geographic atrophy due to age-related macular degeneration." Retina 22(4): 464-
70.

127

Kim, S. Y., S. Sadda, et al. (2002). "Morphometric analysis of the macula in eyes with
disciform age-related macular degeneration." Retina 22(4): 471-7.

Kim, Y. T., R. W. Hitchcock, et al. (2004). "Chronic response of adult rat brain tissue to
implants anchored to the skull." Biomaterials 25(12): 2229-37.

Kipke, D. R., D. S. Pellinen, et al. (2004). CNS Recording Electrodes and Techniques.
Neuroprosthetics Theory and Practice. K. W. Horch and G. S. Dhillon. Singapore,
World Scientific Publishing Co. Pte. Ltd. 2: 765.

Kondo, M., T. Sakai, et al. (2009). "Generation of a Transgenic Rabbit Model of Retinal
Degeneration." Invest Ophthalmol 50: 1371.

Lagali, P. S., D. Balya, et al. (2008). "Light-activated channels targeted to ON bipolar
cells restore visual function in retinal degeneration." Nat Neurosci 11(6): 667-75.

Lapicque, L. (1907). "Recherches quantitatives sur l'excitation electrique des nerfs traites
comme un polarization." J Physiol Paris 9: 15.

Li, Z. Y., D. E. Possin, et al. (1995). "Histopathology of bone spicule pigmentation in
retinitis pigmentosa." Ophthalmology 102(5): 805-16.

Lin, B., S. W. Wang, et al. (2004). "Retinal ganglion cell type, size, and spacing can be
specified independent of homotypic dendritic contacts." Neuron 43(4): 475-85.

Mahadevappa, M., J. D. Weiland, et al. (2005). "Perceptual thresholds and electrode
impedance in three retinal prosthesis subjects." IEEE Trans Neural Syst Rehabil
Eng 13(2): 201-6.

Malik, S., D. Cohen, et al. (1986). "Light damage in the developing retina of the albino
rat: an electroretinographic study." Invest Ophthalmol Vis Sci 27(2): 164-7.

Marc, R. E. and B. W. Jones (2003). "Retinal remodeling in inherited photoreceptor
degenerations." Mol Neurobiol 28(2): 139-47.

Marc, R. E., B. W. Jones, et al. (2003). "Neural remodeling in retinal degeneration." Prog
Retin Eye Res 22(5): 607-55.
Marc, R. E., B. W. Jones, et al. (2008). "Extreme retinal remodeling triggered by light
damage: implications for age related macular degeneration." Mol Vis 14: 782-806.

Margolis, D. J., G. Newkirk, et al. (2008). "Functional stability of retinal ganglion cells
after degeneration-induced changes in synaptic input." J Neurosci 28(25): 6526-
36.
128

Mayhew, T. M. and D. Astle (1997). "Photoreceptor number and outer segment disk
membrane surface area in the retina of the rat: stereological data for whole organ
and average photoreceptor cell." J Neurocytol 26(1): 53-61.

Mazzoni, F., E. Novelli, et al. (2008). "Retinal ganglion cells survive and maintain
normal dendritic morphology in a mouse model of inherited photoreceptor
degeneration." J Neurosci 28(52): 14282-92.

McCreery, D. B., W. F. Agnew, et al. (1990). "Charge density and charge per phase as
cofactors in neural injury induced by electrical stimulation." IEEE Trans Biomed
Eng 37(10): 996-1001.

McCreery, D. B., T. G. Yuen, et al. (1997). "A characterization of the effects on neuronal
excitability due to prolonged microstimulation with chronically implanted
microelectrodes." IEEE Trans Biomed Eng 44(10): 931-9.

McIlwain, J. T. (1978). "Cat superior colliculus: extracellular potentials related to W-cell
synaptic actions." J Neurophysiol 41(5): 1343-58.

Menotti-Raymond, M., V. A. David, et al. (2007). "Mutation in CEP290 discovered for
cat model of human retinal degeneration." J Hered 98(3): 211-20.

Merrill, D. R., M. Bikson, et al. (2005). "Electrical stimulation of excitable tissue: design
of efficacious and safe protocols." J Neurosci Methods 141(2): 171-98.

Milam, A. H., Z. Y. Li, et al. (1998). "Histopathology of the human retina in retinitis
pigmentosa." Prog Retin Eye Res 17(2): 175-205.

Mize, R. R. (1992). "The organization of GABAergic neurons in the mammalian superior
colliculus." Prog Brain Res 90: 219-48.

Nakauchi, K., T. Fujikado, et al. (2007). "Threshold suprachoroidal-transretinal
stimulation current resulting in retinal damage in rabbits." J Neural Eng 4(1): S50-
7.

Nanduri, D., M. S. Humayun, et al. (2008). Retinal Prosthesis Phosphene Shape Analysis.
30th Annual International IEEE EMBS Vancouver.
Narfstrom, K. (1999). "Hereditary and congenital ocular disease in the cat." J Feline Med
Surg 1(3): 135-41.

Narstrom, K. (2007). "Subretinal implantation: a step forward to restoring dying
photoreceptors." Expert Review of Ophthalmology 2(4): 4.
129

Neveu, M. M., G. Jeffery, et al. (2003). "Age-related changes in the dynamics of human
albino visual pathways." Eur J Neurosci 18(7): 1939-49.

O'Hearn, T. M., S. R. Sadda, et al. (2006). "Electrical stimulation in normal and retinal
degeneration (rd1) isolated mouse retina." Vision Res 46(19): 3198-204.

Otto, K. J., M. D. Johnson, et al. (2006). "Voltage pulses change neural interface
properties and improve unit recordings with chronically implanted
microelectrodes." IEEE Trans Biomed Eng 53(2): 333-40.

Pardue, M. T., M. J. Phillips, et al. (2006). "Neuroprotection of photoreceptors in the
RCS rat after implantation of a subretinal implant in the superior or inferior
retina." Adv Exp Med Biol 572: 321-6.

Pardue, M. T., E. B. Stubbs, Jr., et al. (2001). "Immunohistochemical studies of the retina
following long-term implantation with subretinal microphotodiode arrays." Exp
Eye Res 73(3): 333-43.

Peters, S. and U. Schraermeyer (2001). "[Characteristics and functions of melanin in
retinal pigment epithelium]." Ophthalmologe 98(12): 1181-5.

Piyathaisere, D. V., E. Margalit, et al. (2003). "Heat effects on the retina." Ophthalmic
Surg Lasers Imaging 34(2): 114-20.

Potts, A. M. and J. Inoue (1969). "The electrically evoked response (EER) of the visual
system. II. Effect of adaptation and retinitis pigmentosa." Invest Ophthalmol 8(6):
605-12.

Potts, A. M. and J. Inoue (1970). "The electrically evoked response of the visual system
(EER). 3. Further contribution to the origin of the EER." Invest Ophthalmol 9(10):
814-9.

Potts, A. M., J. Inoue, et al. (1968). "The electrically evoked response of the visual
system (EER)." Invest Ophthalmol 7(3): 269-78.

Pouget, A., P. Dayan, et al. (2000). "Information processing with population codes." Nat
Rev Neurosci 1(2): 125-32.
Radtke, N. D., R. B. Aramant, et al. (2004). "Vision change after sheet transplant of fetal
retina with retinal pigment epithelium to a patient with retinitis pigmentosa." Arch
Ophthalmol 122(8): 1159-65.

130

Ranck, J. B., Jr. (1975). "Which elements are excited in electrical stimulation of
mammalian central nervous system: a review." Brain Res 98(3): 417-40.

Rapp, L. M. and T. P. Williams (1979). "Damage to the albino rat retina produced by low
intensity light." Photochem Photobiol 29(4): 731-3.

Ray, A., L. Colodetti, et al. (2009). "Immunocytochemical analysis of retinal neurons
under electrical stimulation." Brain Res 1255: 89-97.

Rijkhoff, N. J., J. Holsheimer, et al. (1994). "Selective stimulation of sacral nerve roots
for bladder control: a study by computer modeling." IEEE Trans Biomed Eng
41(5): 413-24.

Rivolta, C., D. Sharon, et al. (2002). "Retinitis pigmentosa and allied diseases: numerous
diseases, genes, and inheritance patterns." Hum Mol Genet 11(10): 1219-27.

Rizzo, J. F., 3rd, J. Wyatt, et al. (2003). "Methods and perceptual thresholds for short-
term electrical stimulation of human retina with microelectrode arrays." Invest
Ophthalmol Vis Sci 44(12): 5355-61.

Rizzo, J. F., 3rd, J. Wyatt, et al. (2003). "Perceptual efficacy of electrical stimulation of
human retina with a microelectrode array during short-term surgical trials." Invest
Ophthalmol Vis Sci 44(12): 5362-9.

Robblee, L. S., J. L. Lefko, et al. (1983). "Activated Ir: An electrode suitable for
reversible charge injection in saline solution." Journal of the electrochemical
society 130(3): 3.

Rohrer, B. and R. Crouch (2006). Rod and Cone Pigment Regeneration in RPE65-/- Mice.
Retinal Degenerative Diseases. J. G. Hollyfield, R. E. Anderson and M. M. Lavail,
Springer US. 572.

Rose, T. L. and L. S. Robblee (1990). "Electrical stimulation with Pt electrodes. VIII.
Electrochemically safe charge injection limits with 0.2 ms pulses." IEEE Trans
Biomed Eng 37(11): 1118-20.

Sagdullaev, B. T., R. B. Aramant, et al. (2003). "Retinal transplantation-induced recovery
of retinotectal visual function in a rodent model of retinitis pigmentosa." Invest
Ophthalmol Vis Sci 44(4): 1686-95.
Sancho-Pelluz, J., B. Arango-Gonzalez, et al. (2008). "Photoreceptor cell death
mechanisms in inherited retinal degeneration." Mol Neurobiol 38(3): 253-69.

131

Santos, A., M. S. Humayun, et al. (1997). "Preservation of the inner retina in retinitis
pigmentosa. A morphometric analysis." Arch Ophthalmol 115(4): 511-5.

Sanyal, S. and G. H. Zeilmaker (1988). "Retinal damage by constant light in chimaeric
mice: implications for the protective role of melanin." Exp Eye Res 46(5): 731-43.

Schuettler, M., M. Franke, et al. (2008). "A voltage-controlled current source with
regulated electrode bias-voltage for safe neural stimulation." J Neurosci Methods
171(2): 248-52.

Schwahn, H. N., F. Gekeler, et al. (2001). "Studies on the feasibility of a subretinal visual
prosthesis: data from Yucatan micropig and rabbit." Graefes Arch Clin Exp
Ophthalmol 239(12): 961-7.

Sefton A. J., B. D., Alan Harvey (2004). Visual System. The Rat Nervous System. P. G.,
Elsevier Academic Press.

Sekirnjak, C., P. Hottowy, et al. (2006). "Electrical stimulation of mammalian retinal
ganglion cells with multielectrode arrays." J Neurophysiol 95(6): 3311-27.

Sekirnjak, C., P. Hottowy, et al. (2008). "High-resolution electrical stimulation of primate
retina for epiretinal implant design." J Neurosci 28(17): 4446-56.

Shah, H. A., S. R. Montezuma, et al. (2006). "In vivo electrical stimulation of rabbit
retina: effect of stimulus duration and electrical field orientation." Exp Eye Res
83(2): 247-54.

Shah, S., A. Hines, et al. (2007). "Electrical properties of retinal-electrode interface." J
Neural Eng 4(1): S24-9.

Siminoff, R., H. O. Schwassmann, et al. (1966). "An electrophysiological study of the
visual projection to the superior colliculus of the rat." J Comp Neurol 127(4):
435-44.

Simpson, J. and M. Ghovanloo (2007). "An experimental study of voltage, current, and
charge controlled stimulation front-end circuitry." Conf Proc IEEE Eng Med Biol
Soc 2007: 325-328.

Smith, A. J., J. W. Bainbridge, et al. (2009). "Prospects for retinal gene replacement
therapy." Trends Genet 25(4): 156-65.

132

Stone, J. L., W. E. Barlow, et al. (1992). "Morphometric analysis of macular
photoreceptors and ganglion cells in retinas with retinitis pigmentosa." Arch
Ophthalmol 110(11): 1634-9.

Straznicky, C., J. C. Vickers, et al. (1992). "A neurofilament protein antibody selectively
labels a large ganglion cell type in the human retina." Brain Res 582(1): 123-8.

Sun, W., N. Li, et al. (2002). "Large-scale morphological survey of mouse retinal
ganglion cells." J Comp Neurol 451(2): 115-26.

Suzuki, S., M. S. Humayun, et al. (2004). "Comparison of electrical stimulation
thresholds in normal and retinal degenerated mouse retina." Jpn J Ophthalmol
48(4): 345-9.

Szarowski, D. H., M. D. Andersen, et al. (2003). "Brain responses to micro-machined
silicon devices." Brain Res 983(1-2): 23-35.

Thomas, B. B., M. J. Seiler, et al. (2004). "Superior colliculus responses to light -
preserved by transplantation in a slow degeneration rat model." Exp Eye Res
79(1): 29-39.

Thompson, R. W., Jr., G. D. Barnett, et al. (2003). "Facial recognition using simulated
prosthetic pixelized vision." Invest Ophthalmol Vis Sci 44(11): 5035-42.

Turner, J. N., W. Shain, et al. (1999). "Cerebral astrocyte response to micromachined
silicon implants." Exp Neurol 156(1): 33-49.

Veraart, C., M. C. Wanet-Defalque, et al. (2003). "Pattern recognition with the optic
nerve visual prosthesis." Artif Organs 27(11): 996-1004.

Ward, R. (1982). "[Quantitative effects of retinal degeneration in mice]." Rev Can Biol
Exp 41(2): 115-9.

Wassle, H. and B. B. Boycott (1991). "Functional architecture of the mammalian retina."
Physiol Rev 71(2): 447-80.

Wassle, H., U. Grunert, et al. (1989). "Cortical magnification factor and the ganglion cell
density of the primate retina." Nature 341(6243): 643-6.

Wee, A. S. (2001). "Anodal excitation of intact peripheral nerves in humans."
Electromyogr Clin Neurophysiol 41(2): 71-7.

133

Wee, A. S., A. A. Leis, et al. (2000). "Anodal block: can this occur during routine nerve
conduction studies?" Electromyogr Clin Neurophysiol 40(7): 387-91.

Weiland, J., W. Fink, et al. (2005). "Progress towards a high-resolution retinal
prosthesis." Conf Proc IEEE Eng Med Biol Soc 7: 7373-5.

Weiland, J. D., D. J. Anderson, et al. (2002). "In vitro electrical properties for iridium
oxide versus titanium nitride stimulating electrodes." IEEE Trans Biomed Eng
49(12 Pt 2): 1574-9.

Weiss, G. (1901). "Sur la possibilite de render comparables centre eux les appareils
servant a l'excitation electrique." Arch. Ital. Biol. 35: 413-446.

Whalen, J. J. (2005). Microelectrode arrays for neurostimulation applications: fabrication,
characterization, and in vitro retinal cell stimulation, The Johns Hopkins
Univeristy. PhD.

Wilms, M., M. Eger, et al. (2003). "Visual resolution with epi-retinal electrical
stimulation estimated from activation profiles in cat visual cortex." Vis Neurosci
20(5): 543-55.

Yanai, D., R. R. Lakhanpal, et al. (2003). "The value of preoperative tests in the selection
of blind patients for a permanent microelectronic implant." Trans Am Ophthalmol
Soc 101: 223-8; discussion 228-30.

Yanai, D., J. D. Weiland, et al. (2007). "Visual performance using a retinal prosthesis in
three subjects with retinitis pigmentosa." Am J Ophthalmol 143(5): 820-827.

Yokota, T., T. Shiojiri, et al. (1997). "Friedreich-like ataxia with retinitis pigmentosa
caused by the His101Gln mutation of the alpha-tocopherol transfer protein gene."
Ann Neurol 41(6): 826-32.

Yoon, H., D. C. Deshpande, et al. (2008). "Aligned nanowire growth using lithography-
assisted bonding of a polycarbonate template for neural probe electrodes."
Nanotechnology 19(2).

Zheng, X., G. P. Walcott, et al. (2000). "Electrode impedance: an indicator of electrode-
tissue contact and lesion dimensions during linear ablation." J Interv Card
Electrophysiol 4(4): 645-54.

134

Appendix
RPE cells in Control animals
Normal Copenhagen rats constitute the control group and heterozygous S334ter-
line-3 rats the experimental group. The original homozygous S334ter-line-3 Sprague
Dawley (SD) rats were bred with normal Copenhagen rats to produce the heterozygous
offspring. Whether the heterozygous offspring inherit the genetic background of the
albino strain or the pigmented strain has an impact on the differences we observed in the
activation threshold as well as the morphological changes in the retina. For example, it is
reported that prolonged illumination of the retina of the albino rat causes diffuse loss of
photoreceptors (Rapp and Williams 1979; Malik, Cohen et al. 1986) . This, in turn, would
exert an influence on the neural retina and leads to retinal remodeling (Marc, Jones et al.
2008). Melanin-related agents in the retinal pigment epithelium (RPE) regulate the
development of the neural retina (Jeffery 1997). A deficiency of melanin in the RPE
leads to a rod deficit and inappropriate crossing of retinal axons at the optic chiasm
(Neveu, Jeffery et al. 2003), features common in albinism. Ocular melanin plays a
protective role, shielding the neural retina from light damage (Sanyal and Zeilmaker 1988;
Peters and Schraermeyer 2001; Hu, Simon et al. 2008). The amount of light entering the
retina is reduced by the pigmentation in the iris and melanin in the RPE prevents
excessive light reflection. The existence of melanin pigment is observed in the
heterozygous offspring where the iris is pigmented. While albino rats lack melanin
135

pigment and the RPE is absent, our heterozygous offspring show melanin pigment in the
RPE as seen in vertical section (Fig. A1) and EM slides (Fig. A2) It shows that the
transgenic rats in the experimental group are pigmented and do not inherit albinism. Thus,
the retinal remodeling observed in the transgenic rats is most likely due to the mutation
transgene inherited from the homozygous parent.
We understand that the desired match of experimental animals would be those
that are heterozygous for the transgene. This would be produced from crossing
heterozygous S334ter-3 rats with Copenhagen rats to produce litters consisting of 50%
heterozygous rats and 50% wild-type rats of the same mixed genetic background.
However, at this time, we believe that the crosses we have used give us a close
approximation of the optimal situation.
136

Figure A.1 RPE layers in four hybrid RD and Copenhagen rats. Scale bar = 10 µm.
137

Fig A2 Electron micrograph of melanin granules. Most of them do not lie singly within
the cytoplasm but occur in groups enclosed within a membrane and surrounded by a
matrix. A large number of melanin granules are embedded in a less dense matrix and
surrounded by a membrane.

138

Materials and methods
Animals
One Copenhagen rat (P180, n=1) and three S334ter-line-3 (P180-600, n=3) rats
were used in this experiment. For vertical section preparation of the retina see Chapter 4
Section 2.
Microscopic examination of fixed tissues
Tissue was prepared for light microscopy and transmission electron microscopy
(LM and TEM). For LM, the cryostat retina sections, stained with H & E, were examined
in a Nikon deconvolution microscope. For TEM, in the fixative procedure, the frozen
section was washed in a 0.1M cacodylate buffer, followed by fixing in ½ Karnovsky’s
fixative. After 1 hr of fixation the tissue was washed in three changes of 0.1M cacodylate
buffer. All tissues were dehydrated in a graded ethanol series, transferred to propylene
oxide, and embedded in Epon under vacuum. BEEMs capsules were filled with Epon and
put into oven overnight. Plastic BEEMs capsules were cut off and individual sections
were obtained. Thin sections (800 to 1000 Å) were placed on copper slot grids, stained
with uranium, and examined in a JEM-2100 transmission electron microscope.

Abstract (if available)

Abstract Retinitis Pigmentosa,(RP) is a blinding disease characterized by massive and progressive reductions in the population of photoreceptor cells, result in losing night vision, followed severely loss of peripheral vision and often total blindness. No effective treatment is yet available for RP. Electrical stimulation of the retina through a bioelectronic implant, replaces some of the lost function of the degenerated neurons, allowing test subjects with experimental implants to perform simple visual tasks. Thus, retinal implants have potential to provide an effective means of restoring vision to RP patients.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Towards a high resolution retinal implant

PDF

Effect of continuous electrical stimulation on retinal structure and function

PDF

Cortical and subcortical responses to electrical stimulation of rat retina

PDF

Manipulation of RGCs response using different stimulation strategies for retinal prosthesis

PDF

Understanding the degenerate retina's response to electrical stimulation: an in vitro approach

PDF

Stimulation strategies to improve efficiency and temporal resolution of epiretinal prostheses

PDF

Investigation of the electrode-tissue interface of retinal prostheses

PDF

Improving stimulation strategies for epiretinal prostheses

PDF

Dependence of rabbit retinal synchrony on visual stimulation parameters

PDF

Electrical stimulation approaches to restoring sight and slowing down the progression of retinal blindness

PDF

Electrical stimulation of the orbicularis oculi to restore eye blink

PDF

Simulating electrical stimulation and recording in a multi-scale model of the hippocampus

PDF

Therapeutic electrical stimulation strategies for neuroregeneration and neuroprotection of retinal neurons

PDF

RPE secretome for the treatment of retinal degeneration in the RCS rat

PDF

Synaptic second wind: phosphene fading of the Argus II and glutamate secretion in the inner retina

PDF

Characterization of visual cortex function in late-blind individuals with retinitis pigmentosa and Argus II patients

PDF

Computational approaches to establish safety and efficacy assessments of electrical stimulation to peripheral nerve

PDF

Intraocular and extraocular cameras for retinal prostheses: effects of foveation by means of visual prosthesis simulation

Asset Metadata

Creator Chan, Lai Hang Leanne (author)

Core Title Electrical stimulation of degenerate retina

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Biomedical Engineering

Publication Date 07/29/2009

Defense Date 06/01/2009

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag degenerate retina,electrical stimulation,electrophysiology,ganglion cell density,immunocytochemistry,impedance sensing,OAI-PMH Harvest,threshold charge density,tissue-electrode interface

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Weiland, James D. (committee chair), Grzywacz, Norberto M. (committee member), Hirsch, Judith A. (committee member), Humayun, Mark S. (committee member)

Creator Email laichan@usc.edu,uscleanne@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2420

Unique identifier UC1494941

Identifier etd-Chan-3136 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-564374 (legacy record id),usctheses-m2420 (legacy record id)

Legacy Identifier etd-Chan-3136.pdf

Dmrecord 564374

Document Type Dissertation

Rights Chan, Lai Hang Leanne

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu