Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
PET study of retinal prosthesis functionality
(USC Thesis Other)
PET study of retinal prosthesis functionality
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
PET STUDY OF RETINAL PROSTHESIS FUNCTIONALITY
by
John Zhong Xie
A Dissertation Presented to the
FACULTY OF THE USC VITERBI SCHOOL OF ENGINEERING
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirement for the Degree
DOCTOR OF PHILOSOPHY
(BIOMEDICAL ENGINEERING)
December 2009
Copyright 2009 John Zhong Xie
ii
Acknowledgements
The acknowledgement is a difficult thing to write considering the number of
people that are responsible for this paper. However, a few contributions deserve to be
recognized individually for their effort. Foremost, I want to express my deepest
appreciation to my advisor, Dr. Mark Humayun and co-advisor, Dr. James Weiland.
Without their expertise, knowledge and support, this work could not have started, and
without their guidance and help, this work could not have been completed. My gratitude
also goes to my graduate committee members: Dr. Hossein Jadvar, Dr. Michael Khoo,
Dr. Manbir Singh and Dr. Alapakkam Sampath.
I would like to thank Lindy Yow for always being there for me throughout my
research endeavor. My great appreciation also goes to our wonderful collaborators, Dr.
Gene-Jack Wang and his team at Brookhaven National Laboratory for being instrumental
in making this project possible, Dr. Stephen Tsang and Elona Gavazi from Columbia
University for recruiting retinal degeneration subjects, and Dr. Gianluca Lazzi and Carlos
Cela for their contributions in computer modeling of transcorneal electrical stimulation.
I would like to extend my appreciation to Len Richine and David Valentine from
the Doheny Eye institute for helping me to acquire the clinical skills necessary to carry
out my experiments.
My deepest and most sincere appreciation also goes to my parents and my
girlfriend, Gloria Ling, who have always supported me and kept me focused. I also would
like to thank all my friends and colleagues for their encouragements throughout the years.
iii
Table of Contents
Acknowledgements ……………………………………………………………………..ii
List of Tables……………………………………………………………………………viii
List of Figures……………………………………………………………………………ix
Abbreviations……………………………………………………………………………xiv
Abstract…………………………………………………………………………………xvi
Chapter 1: Introduction…………………………………………………………............1
1.1 Retinal Degenerative Diseases………………………………………………1
1.2 Visual Prostheses Functional Assessment……………………………..........4
1.3 Objectives of Research……………………………………………...............5
Chapter 2: Visual Prostheses Overview…………………………………………...........8
2.1 Introduction to Visual Prostheses………………………………………......8
2.2 Retinal Prostheses…………………………………………………..............9
2.2.1 Epiretinal Visual Prosthesis……………………………………….10
2.2.2 Subretinal Visual Prosthesis…………………………….…...........12
2.3 Optic Nerve Visual Prosthesis………………………………………………14
2.4 Intracortical Visual Prosthesis……………………………………………… 15
Chapter 3: Human Visual Cortex Overview………………………………………….....19
3.1 Mapping Visual Cortex with Functional Neuroimaging…………...……….19
3.2 Positron Emission Tomography (PET)……………………………………...19
3.3 Functional Magnetic Resonance Imaging (fMRI)…………………………..20
3.3.1 Biological Source of fMRI Signal………………………………...20
3.3.2 Phase-encoded Measurements with fMRI……………….……….21
3.3.3 Flattened Cortical Representation………………………...............22
3.4 Overview of Human Visual Cortical Areas…………………………............23
3.4.1 Retinotopic Organization of Human Visual Areas V1/2/3………..24
3.4.2 Dorsal and Ventral Visual Streams………………………..............30
3.4.3 Dorsal Visual Stream……………………………………………...30
iv
3.4.4 Ventral Visual Stream…………………………………….............31
3.5 Monocular Photic Stimulation……………………………………................32
3.5.1 Findings from fMRI Studies…………………………………….32
3.5.2 Nasotemporal Asymmetry in the Human Eye……………….…..35
3.5.3 Effects of Ocular Dominance……………………………………36
3.6 Visual Cortex and Blindness……………………………………………....37
3.6.1 Cortical Effects of Early Visual Deprivation……………………37
3.6.2 Cross-modal Plasticity…………………………………………...39
3.7 Visual Cortex Remodeling in Retinal Degenerative Diseases……………..41
3.8 Plasticity in Matured Brain ……………………………...........................43
3.9 Lessons Learned from Cochlear Implant……………………………..........43
Chapter 4: Transcorneal Electrical Stimulation………………......................................46
4.1 Introduction………………………………………………………………..46
4.1.1 Electrically Evoked Phosphenes…………………………………46
4.1.2 Retinal Origin of electrically Evoked Response…………………47
4.1.3 Value of TcES in Implant Candidate Evaluation………………..48
4.2 Psychophysical Techniques………………………………………………...49
4.2.1 Concepts of Psychophysics………………………………………49
4.2.2 Visual Threshold Methods……………………………………….50
4.3 Experimental Setup for TcES………………………………………………52
4.3.1 Goal of TcES……………………………………………………..52
4.3.2 DTL-Plus Electrode……………………………………………....52
4.3.3 ERG-Jet Electrode………………………………………………..55
4.3.4 Neurostimulator…………………………………………………..55
4.3.5 Corneal Exam…………………………………………………….59
Chapter 5: Modeling of Transcorneal Electrical Stimulation………………………….60
5.1 Modeling Overview………………………………………………………..60
5.2 Modeling Eye and Corneal Electrodes…………………………………….61
5.3 DTL-Plus Modeling Results……………………………………………….67
5.3.1 DTL-Plus Potential Distribution…………………………………67
v
5.3.2 DTL-Plus Activating Function……..…………………………...70
5.4 ERG-Jet Modeling Results…………………………………………………71
5.4.1 ERG-Jet Potential Distribution……………………………………71
5.4.2 ERG-Jet Activating Function…….………..................................75
5.5 Modeling Summary……………………………………………………….75
Chapter 6: PET Imaging and Analysis Methods…………………...............................77
6.1 Overview of PET Functional Brain Imaging……………………………...77
6.1.1 Positron Emission Tomography Overview……………………...77
6.1.2 Radioisotopes……………………………………………………79
6.1.3 Image Reconstruction……………………………………………79
6.1.4 Fluorodeoxyglucose (FDG)……………………………………..80
6.1.5 Functional Brain Imaging Using FDG…………………….........82
6.1.6 Quantitative PET………………………………………………..83
6.1.7 Metabolic Rate of Glucose (MRGlc)……………………………83
6.2 Methods of Analysis………………………………………………………86
6.3 Statistical Parametric Mapping (SPM)……………………………………86
6.3.1 Overview of SPM……………………………………………….86
6.3.2 Image Pre-processing……………………………………………88
6.3.3 General Linear Model……………………………………..…….90
6.3.4 Subtraction Methodology……………………………..…….......92
6.3.5 Absolute and Relative Analyses……………………………..….92
6.3.6 Statistical Comparison………………………………………......94
6.3.7 Uncorrected P-value and Regional Hypothesis……………..….95
6.3.8 Graphical Representations…………………………………..….95
6.4 Quantitative Analysis using PMOD…………………………………..….96
Chapter 7: Preliminary Studies…………………………………………………....…..97
7.1 Epiretinal Prosthesis Implant Subject PET-H
2
O
15
Study………………97
7.2 Normal Control Light Stimulation PET Study………………………..…99
7.3 Summary……………………………..….…………………………..……..101
Chapter 8: Research Design and Methods………………………..............................102
vi
8.1 Experimental Methods……………………………………………..…….102
8.1.1 Experiment Overview………………………………………..…102
8.1.2 Baseline Study Protocol………………………………………...104
8.1.3 Light Stimulation Study Protocol…………………………...........104
8.1.4 DTL-Plus TcES Study Protocol……………….............................106
8.1.5 ERG-Jet TcES Study Protocol…………………………………...108
8.1.6 Study Flow Chart…………………………………………...........111
8.2 Experimental Techniques…………………………………………………...112
8.2.1 Screening Visit……………………………………………...........112
8.2.2 FDG PET Scanning Visits……………………………………….112
8.2.3 Randomization…………………………………………………...113
8.2.4 Arterialized Venous Catheterization…………………………….113
8.2.5 Radiotracer Synthesis……………………………………………114
8.2.6 PET Image Reconstruction………………………………………114
Chapter 9: Results and Analyses………………………………………………………115
9.1 Normal Control Subjects…………………………………………………..115
9.1.1 Subjective Sensation of Stimulation……………………….........115
9.1.2 Light Stimulation vs. Baseline……………………………..........116
9.1.3 TcES (DTL-Plus) vs. Baseline……………………………..........132
9.1.4 Light Stimulation vs. TcES (DTL-Plus)…………………………138
9.1.5 TcES (ERG-Jet) vs. Baseline…………………………………….143
9.1.6 Light Stimulation vs. TcES (ERG-Jet)…………………………..148
9.2 Retinal Degeneration Subjects...…………………………………………..153
9.2.1 Subjective Sensation of Stimulation……………………….........156
9.2.2 Light Stimulation vs. Baseline…………………………………..157
9.2.3 TcES (ERG-Jet) vs. Baseline……………………………………163
9.2.4 Light Stimulation vs. TcES (ERG-Jet)………………………….168
9.3 Normal Control and Retinal Degeneration Subject Comparison……........173
9.3.1 Phosphene Threshold Comparison……………………………...173
9.3.2 Light Stimulation Cross-Comparison……………………………175
vii
9.3.3 TcES (ERG-Jet) Cross-Comparison……………………………..181
9.3.4 Higher-level Visual Area Comparison……………………...........185
9.4 Results Summary…………………………………………………………..190
9.4.1 Normal Controls…………………………………………………190
9.4.2 Retinal Degeneration Subjects…………………………………..192
9.4.3 Normal and Retinal Degeneration Group Comparison………….193
Chapter 10: Conclusion…………………………………………………………..........196
10.1 Contributions……………………………………………………….........197
10.2 Suggestions for Future Work…………………………………………….197
Bibliography……………………………………………………………………..........199
Appendices……………………………………………………………………………...219
Appendix A. Three-Compartmental Model for FDG Metabolism…………..219
Appendix B. General Linear Model Parameter Estimation……….………...222
Appendix C. Activating Function Derivation………………………………...224
Appendix D. Human Subject Inclusion and Exclusion Criteria…………..228
viii
List of Tables
1. Comparison of various approaches to prosthetic vision…………………………….17
2. Study flow chart………………………………………………………………….111
3. Normal controls identifying data………………………………………………...115
4. Eight normal controls light stimulation vs. baseline SPM summary…………….127
5. Five normal controls light stimulation vs. baseline SPM summary……………..131
6. Five normal controls TcES (DTL-Plus) vs. baseline SPM summary……………137
7. Five normal controls light stimulation vs. TcES (DTL-Plus) SPM summary…….142
8. Five normal controls TcES (ERG-Jet) vs. baseline SPM summary………………147
9. Five normal controls light stimulation vs. TcES (ERG-Jet) SPM summary……..152
10. RD subjects identifying data………………………………………………….....155
11. RD subjects light stimulation vs. baseline SPM summary………………………162
12. RD subjects TcES (ERG-Jet) vs. baseline SPM summary………………………167
13. RD subjects light stimulation vs. TcES (ERG-Jet) SPM summary……………..172
14. Light stimulation normal controls vs. RD subjects SPM summary……………..180
15. TcES (ERG-JET) normal controls vs. RD subjects SPM summary…………….185
16. Quantitative comparison of activation in higher-level visual areas……………..189
17. Comparison of activated brain regions between normal and RD subjects………195
ix
List of Figures
1. Comparison between epiretinal and subretinal visual prostheses………………….11
2. Drawing of an intracortical artificial vision system………………………………..16
3. Stimuli used for phase-encoded measurements of retinotopic organization……….22
4. The retinotopy paradigm……………………………………………………...........26
5. Retinotopic organization of visual areas……………………………………...........27
6. Visual areas on flattened human occipital cortex …………………………………28
7. The dorsal and ventral visual pathways……………………………………………29
8. Monocular photic stimulation summary……………………………………………34
9. Schematic diagram of visual field projection to primary visual cortex……………34
10. The role of ocular dominance in visual cortex activation………………………….37
11. Tactile sensory cross-modal plasticity in early blind subjects……………………..40
12. Auditory sensory cross-modal plasticity in early blind subjects……………..........40
13. DTL-Plus electrode…………………………………………………………..........53
14. JET-ERG electrode………………………………………………………………..54
15. Digitimer DS7A neurostimulator………………………………………………………57
16. Digitimer DG2A trigger generator…………………………………………………….57
17. Neurostimulator setup for TcES……………………………………………………....58
18. Trigger and stimulus pulse trains…………………………………………………58
19. Light source for fluorescein corneal exam…………………………………………...59
20. Sagittal and Axial slices through human eye model…………………………………62
21. Coronal slices through human eye model……………………………………………63
x
List of Figures (Continued)
22. Modeling of corneal electrodes on the human eye…………………………….....64
23. Position of reference electrode on the temporal skin………………………......66
24. Stimulating pulse and its fast Fourier transform……………………………......67
25. Electric potential distribution in the eye from DTL-Plus electrode stimulation…68
26. Potential distribution and activating function maps for DTL-Plus electrode..…69
27. Electric potential distribution in the eye from ERG-Jet electrode stimulation…...72
28. Potential distribution and activating function maps for ERG-Jet electrode……..…73
29. Activating function overlays on normal fundus photograph……………………….74
30. Scheme of PET acquisition process…………..……………………….………..78
31. Chemical structure of FDG…………………………………………………..…81
32. Three-compartment FDG model…………………………………………………84
33. SPM overview………………………………………………………………………88
34. Brain activation during epiretinal prosthesis stimulation………………………..…97
35. Brain activation during light stimulation in normal sighted subjects…………..….100
36. PET experiment flow diagrams………………………………………………..…103
37. Light stimulation setup…………………………………………………………..105
38. DTL-Plus TcES study setup………………………………………………………109
39. ERG-Jet electrode placed on a subject’s right eye………………………………..109
40. Normal control subjective visual perception………………………………………117
41. Eight normal controls light stimulation activation SPM maps (p<0.05)………….121
42. Eight normal controls light stimulation activation SPM maps (p<0.01)………….122
xi
List of Figures (Continued)
43. Eight normal controls light stimulation deactivation SPM maps (p<0.05)……….123
44. Eight normal controls light stimulation activation 3D SPM projection…………124
45. Eight normal controls light stimulation deactivation 3D SPM projection………125
46. Five normal controls light stimulation activation SPM maps……………………128
47. Five normal controls light stimulation deactivation SPM maps…………………129
48. Five normal controls light stimulation activation 3D SPM projection………......130
49. Five normal controls light stimulation deactivation 3D SPM projection………..130
50. Five normal controls TcES (DTL-Plus) activation SPM maps…………………..134
51. Five normal controls TcES (DTL-Plus) deactivation SPM maps………………..135
52. Five normal controls TcES (DTL-Plus) activation 3D SPM projection…………136
53. Five normal controls TcES (DTL-Plus) deactivation 3D SPM projection………136
54. Five normal controls SPM map - Light stimulation > TcES (DTL-Plus)………..139
55. Five normal controls SPM map - TcES (DTL-Plus) > Light stimulation………..140
56. Normal control 3D SPM projection - Light stimulation > TcES (DTL-Plus)…….141
57. Normal control 3D SPM projection - TcES (DTL - Plus) > Light stimulation…..141
58. Five normal controls TcES (ERG-Jet) activation SPM maps……………………144
59. Five normal controls TcES (ERG-Jet) deactivation SPM maps…………………145
60. Five normal controls TcES (ERG-Jet) activation 3D SPM projection…………..146
61. Five normal controls TcES (ERG-Jet) deactivation 3D SPM projection………..146
62. Five normal controls SPM map – Light Stimulation > TcES (ERG-Jet)………..149
63. Five normal controls SPM map – TcES (ERG-Jet) > Light Stimulation………..150
xii
List of Figures (Continued)
64. Normal control 3D SPM projection - Light Stimulation vs. TcES (ERG-Jet)…...151
65. OCT study of RD 1’s right eye……………………………………………………154
66. Phosphene sensation reported by RD subjects during TcES (ERG-Jet)………….157
67. Five RD subjects light stimulation activation SPM maps…………………………159
68. Five RD subjects light stimulation deactivation SPM maps………………………160
69. Five RD subjects light stimulation activation 3D SPM projection………………..161
70. Five RD subjects light stimulation deactivation 3D SPM projection……………..161
71. Five RD subjects TcES (ERG-Jet) activation SPM maps…………………………164
72. Five RD subjects TcES (ERG-Jet) deactivation SPM maps……………………….165
73. Five RD subjects TcES (ERG-Jet) activation 3D SPM projection………………...166
74. Five RD subjects TcES (ERG-Jet) deactivation 3D SPM projection……………..166
75. Five RD subjects SPM map – Light Stimulation > TcES (ERG-Jet)……………..169
76. Five RD subjects SPM map – TcES (ERG-Jet) > Light Stimulation……………..170
77. RD subjects 3D SPM projection - Light Stimulation vs. TcES (ERG-Jet)……….171
78. Comparison of phosphene threshold current between normal and RD subjects…..174
79. Individual threshold current comparison between normal and RD subjects……....175
80. Light stimulation subtraction metablic image comparison………………………..177
81. Light stimulation comparison SPM – Normal controls > RD subjects……………178
82. Light stimulation comparison SPM - RD subjects > Normal controls……………179
83. Light stimulation comparison 3D SPM projection………………………………..180
xiii
List of Figures (Continued)
84. TcES (ERG-Jet) comparison SPM – Normal controls > RD subjects…………….182
85. TcES (ERG-Jet) comparison SPM - RD subjects > Normal controls………….....183
86. TcES (ERG-Jet) comparison 3D SPM projection………………………………..184
87. Higher-level visual area comparison between normal and RD subjects…………188
88. 3D SPM activation comparison between normal and RD subjects………………194
xiv
Abbreviations
AIT…………………………………………Anterior Inferior Temporal
AMD……………………………………….Age-related Macular Degeneration
APB………………………………………...2-Amino-4-Phosphonobutyric Acid
ASR………………………………………...Artificial Silicon Retina
BA………………………………………….Brodmann Area
BNL………………………………………...Brookhaven National Laboratory
BOLD………………………………………Blood Oxygen Level Dependent (fMRI)
CBC………………………………………...Complete Blood Count
CIT………………………………………….Central Inferior Temporal
CT…………………………………………...Computed Tomography
DTL…………………………………………Dawson, Trick, and Litzkow
EEPs………………………………………...Electrically Evoked Phosphenes
EER…………………………………………Electrically Evoked Response
ERG………………………………………...Electroretinography
FDG…………………………………………
18
F-fluorodeoxyglucose
fMRI………………………………………..Functional Magnetic Resonance Imaging
FOV………………………………………...Field of View
LGN………………………………………..Lateral Geniculate Nucleus
LIP…………………………………………Lateral Intraparietal
MAP………………………………………..Maximum a posteriori Projection
mCi…………………………………………Millicuries (1 Curie = 3.7 x 10
10
decays/sec)
xv
Abbreviations (continued)
MD…………………………………………Macular Degeneration
MRGlc……………………………………..Metabolic Rate of Glucose
MST………………………………………..Medial Superior Temporal
MT………………………………………….Middle Temporal
OCT………………………………………...Optical Coherence Tomography
OFL………………………………………...Optic Fiber Layer
PCP…………………………………………Phencyclidine
PET…………………………………………Positron Emission Tomography
PIT………………………………………….Posterior Inferior Temporal
RD………………………………………….Retinal Degeneration
RGC………………………………………...Retinal Ganglion Cell
RP…………………………………………...Retinitis Pigmentosa
SPM………………………………………....Statistical Parametric Mapping
STAT UCG………………………………….Immediate Urinary Chorionic Gonadotropin
STAT UDS………………………………….Immediate Urodynamic Study
TcES………………………………………...Transcorneal Electrical Stimulation
THC………………………………………....Tetrahydrocannabinol (Marijuana)
TSH……………………………………….....Thyroid Stimulating Hormone
VEEP………………………………………..Visually Electrically Evoked Potential
VEP………………………………………….Visually Evoked Potential
VIP…………………………………………..Ventral Intraparietal
xvi
Abstract
Objective measures that demonstrate activation of the retina, primary visual cortex, and
possibly higher cortical association areas are necessary to illustrate the functionality of
retinal prosthesis in sight-impaired subjects who have been implanted with such a device.
Neural remodeling that occurs in outer retinal degenerative diseases can have an impact
on the usefulness of retinal prostheses.
In this project we use quantitative positron emission tomography (PET) method
coupled with metabolic radiotracer
18
F-fluorodeoxyglucose (FDG) to evaluate brain
activation under light stimulation and transcorneal electrical stimulation (TcES)
conditions in normal sighted controls and subjects with retinal degeneration (RD). Both
light stimulation and TcES resulted in activation of primary (BA 17) and secondary
visual cortex (BA 18 and 19) in RD subjects. Transcorneal electrical sitmulation using
ERG-Jet corneal electrode resulted in similar subjective phosphene sensation in the right
peripheral visual field for both normal and RD subjects, and this led to retinotopically
matched primary visual cortex activation in both subject groups. Higher-level visual
processing areas that are involved in object vision (inferior temporal gyrus , BA 20;
fusiform gyrus, BA 37) and visual memory (parahippocampal gyrus, BA 27) are not
significantly activated in RD subjects during both light and transcorneal electrical
stimulation as they are in normal controls.
This study objectively demonstrates electrical stimulation of the retina can
activate visual cortex and lead to visual perception in retinal degeneration subjects and
establishes a basis for assessing retinal prosthesis functionality in vivo.
1
Chapter 1: Introduction
1.1 Retinal Degenerative Diseases
The two most common categories of outer retinal degenerative diseases are macular
degeneration and retinitis pigmentosa.
Macular degeneration is an umbrella term that describes the loss of vision in the
center of the visual field because of damage to the retina. Although there are macular
dystrophies that affect the young, the term macular degeneration is generally referred to
as age-related macular degeneration (AMD). In Western countries, AMD is the principal
cause of severe central vision loss among adults 65 years of age and older [59]. In the
United States, there are approximately 700,000 new AMD patients each year, 10% of
whom will become legally blind [40, 152]. The statistics for the prevalence of Macular
Degeneration show that 14.4% of patients between the age of 55 and 64 are suffering
from this condition; 19.4% of the population between 65 and 74; and 36.8% in people
over 75 years of age [118].
AMD results from abnormal aging of the retinal pigmented epithelium (RPE) in
the retina resulting in photoreceptor destruction. Common clinical findings include the
formation of yellow lesions (drusen) on the RPE and proliferation of leaky blood vessels
in the subretinal space [208]. The types and severity of AMD are defined by these clinical
findings. Persons with AMD will start to have distorted central vision and will eventually
lose most vision in the central 30° field of view. A number of treatments have shown
2
some effectiveness in slowing the progression of AMD, but currently no cure exists
[265].
There also exists a juvenile form of macular degeneration, known as Stargdart’s
disease (STGD), which is an autosomal recessive retinal disorder characterized by
juvenile-onset macular dystrophy. Mutations in a gene encoding an ATP-binding cassette
(ABC) transporter have been linked to the development of STGD.
Retinitis pigmentosa (RP) is a clinically and genetically heterogeneous group of
primary retinal degenerations in which abnormalities of the photoreceptors (rods &
cones) or the retinal pigment epithelium of the retina lead to progressive visual loss [15,
102, 225]. RP is one of the major forms of incurable blindness in the world, with a
prevalence varying from 1 in 4000 to 1 in 1000 in different regions of the world [59, 102,
152]. There are approximately 1.5 million patients living with RP worldwide [59].
Although the course and progression of the disease show considerable variation between
individuals, it is typically characterized by initial symptoms of night blindness, with onset
in adolescence or early adulthood, loss of peripheral vision and, as the disease progresses,
loss of central vision leading to complete blindness or severe visual impairment. At least
40 genes have been linked to different forms of nonsyndromic retinitis pigmentosa [102,
105]. Currently, there is no treatment or cure for RP.
Leber’s congenital amaurosis (LCA) is a variant of retinitis pigmentosa. It is an
autosomal recessive disorder caused by abnormal development of photoreceptors. LCA is
typically characterized by nystagmus and severe vision loss or blindness.
3
Although the symptoms, severity and progression vary between different outer
retinal degenerative conditions, one common finding in all of them is the loss of
photoreceptors. Postmortem evaluations of the retina primarily in subjects with RP and
AMD have shown a large number cells remain in the inner retina compared with the
outer retina [217, 235, 119]. In severe RP, only 4% of photoreceptors remained in the
macula, but 80% of inner retina and 30% of ganglion cells remained. In extramacular
regions, only 40% of inner retina remained. In AMD, 90% of the inner retina cells
remained intact morphologically compared with age-matched controls [129]. Thus, by
cell counting, the inner retina in RP and AMD appears to be less affected by the disease
compared with the photoreceptors.
Furthermore, electrical stimulation of the retina in humans with RP and AMD
results in light perception (phosphenes), supporting the idea that some neural elements
exist that can continue to be activated [120, 196, 206]. Recent studies using novel cell
tracing, however, suggest that the inner retina undergoes significant remodeling during
retinal degeneration particularly in RP patients [149, 150]. This type of remodeling may
lead to disruption of spatial cell patterning and impede rescue strategies using retinal
prosthetic implants, photoreceptor transplants or stem cell therapies, which all depend on
a viable, intact retinal circuitry following photoreceptor degeneration. In summary,
retinal dystrophies do not result in complete degeneration of the retina; inner retinal cells
and retinal ganglion cells are preserved to a large extent and remain electrically excitable.
Although on a global scale retinal degenerative diseases represent only a small
proportion of conditions causing blindness, they are common in the developed world and
4
remain incurable. In contrast, many other blinding diseases such as glaucoma and cataract
are due to lack of access to health care. These retinal degenerative dieseases, which
principally affect the photoreceptors, can benefit from a visual prosthesis in the future.
1.2 Visual Prostheses Functional Assessment
Thus far, most of the functional evaluations of the visual prostheses are done using
psychophysics. Visual prosthesis implant subjects are asked about their perception of
phosphenes while experimenters methodically vary the electrical stimulation parameters
(i.e. current amplitude and frequency, pulse-train pattern, pulse duration, etc.) that target
the various cells or axonal fibers along the visual pathway from the retina to the brain.
Although these subjective methods yield important data on optimal visual stimulation
parameters, they do not afford an objective assessment of the visual prosthesis
functionality.
Functional imaging, on the other hand, can objectively demonstrate changes in
brain activity during stimulation of a specific sensory input. By using FDG PET on a
group of retinal degenerative subjects, we want to evaluate their brain response to both
light and electrical stimulation of the retina and compare the results to a group of age-
matched normal sighted controls. This novel approach will allow us to study for the first
time the response of the visual cortex during electrical retinal stimulation and look for
changes in the pattern that are indicative of brain remodeling as a result of prolonged
severe visual loss. Findings from this study can potentially help in forming a basis for
evaluating the efficacy of retinal prosthesis in functional visual restoration. We may
5
begin to understand why some implant subjects respond better and adapt more quickly to
prosthetic vision than others.
1.3 Objectives of Research
The purpose of this research is to extend the existing knowledge of the functionality of
retinal prosthesis by looking at brain’s response to electrical retinal stimulation in
subjects with retinal degeneration (RD). The research focuses on addressing the
following three hypotheses:
• Primary and secondary visual areas in the brain are preserved and remain receptive
to both light and electrical retinal stimulation despite prolonged, severe visual
impairment resulting from retinal degenerative diseases;
• Electrical retinal stimulation can elicit similar phosphene perception between normal
and RD subjects and result in retinotopically mapped activation in the primary visual
cortex.
• Higher-level visual processing areas in the brain can become less responsive to both
light and electrical retinal stimulation following prolonged, severe visual loss.
Functional imaging studies have demonstrated light stimulus can elicit visual
cortex activation in subjects with severe outer retinal degeneration. The pattern of
activation may be different from normal sighted controls due to both visual impairment
and potential neural remodeling following visual loss. Using positron emission
tomography, it was demonstrated in one epiretinal implant subject that electrical
stimulation of the retina via the implanted electrodes can increase regional cerebral blood
6
flow in the visual cortex. We hypothesize that metabolism in the occipital cortex which
represents the primary and secondary visual areas in RD subjects can increase
significantly compared to baseline during both light and electrical retinal stimulation.
Transcorneal electrical stimulation will likely activate regions containing viable
inner retinal neurons and retinal ganglion cells by directly stimulating them to fire action
potentials. This type of stimulation bypasses degenerate photoreceptors and allows
ganglion cells to transmit phosphene sensation to the primary visual cortex. We
hypothesize that retinotopic mapping between the retinal ganglion cells and the primary
visual cortex is maintained in RD subjects, and electrical stimulation of a specific region
of the retina can elicit activation in the corresponding primary visual cortex which
receives visual projection from the same region of the retina.
Early visual areas in RD subjects who have bare light-perception vision may not
transmit enough spatial detail to activate higher-level visual processing areas of the brain
that are involved in complex visual processing, simply due to a lack of visual detail
transmitted by the retina. Furthermore, functional imaging studies have shown prolonged
blindness or severe visual loss leads to plasticity and even cross-modal utilization of
predominantly higher association visual areas by other sensory modalities such as
auditory and tactile sensations. This type of plasticity may lead to an increase in neural
processing of non-vision sensory modalities in the higher, integrative visual areas with a
compensatory loss of visual processing. We hypothesize that higher-level visual
processing regions of the brain will become less responsive to both light and electrical
retinal stimulation comparing to normal sighted controls.
7
Prior to discussing the research topics, a literature review on visual prostheses and
an overview of visual cortex will be presented.
8
Chapter 2: Visual Prostheses Overview
2.1 Introduction to Visual Prostheses
Severe forms of blindness, resulting in a loss of light perception, have a devastating
impact on the functional ability [264], physiological state [211] and psychosocial well-
being of affected patients [45]. In the developed world, the majority of end-stage
blindness is the result of pathology which damages neural tissue in the retina and optic
nerve [133]. Due to the poor capacity of neural tissue for functional repair and
regeneration [113, 160], currently little can be done to restore visual sensations to totally
blind patients.
It has been proposed by a number of investigators that a medical device that
electrically stimulates cells in the neural visual pathway might be able to bypass the
damaged tissue and restore visual sensations to blind patients [19, 118, 241, 255].
Brindley made the first attempt to implant a visual prosthesis in a human which dates
back to 1967 [19]. A series of 80 electrodes were implanted subdurally over the occipital
pole of the cortex. Each electrode was connected to a subcutaneous radio receiver that
could be activated through an external antenna. Small, precisely located phosphenes were
obtained, suggesting that a useful prosthesis could indeed be possible. However, the
cortex surface electrodes required too much power and did not provide an adequate
spatial resolution for practical use. Little is known regarding further clinical trials arising
from this research [52].
9
The visual prosthesis almost disappeared from the scientific literature until a
series of papers demonstrated that RP was characterized by a relatively selective
degeneration of rods and cones [235]. Soon thereafter, some researchers suggested the
implantation of electrodes on the retinal ganglion cells [117, 204]. This marked the
beginning of the epiretinal approach. Other groups alternatively suggested a subretinal
approach where degenerated rods and cones would be replaced by artificial photocells
[30, 274].
Since then a number of different visual prostheses have been developed [156]
which electrically stimulate the visual pathway at the level of the retina [151], the optic
nerve [257] or the visual cortex [52]. Concurrently, work has being carried out to
improve implantable electrodes. Miniature multi-contact intracortical electrodes are built
by the application of methods from the semiconductor industry [99, 123, 240]. Spiral cuff
electrodes are also being developed for the stimulation of optic nerves [172, 255].
2.2 Retinal Prostheses
During the early 1970s it became clear that blind humans can perceive electrically-
elicited phosphenes in response to ocular stimulation, with a contact lens as a stimulating
electrode [196, 197]. These electrically elicited responses indicated the presence of at
least some functioning inner retinal cells despite retinal degenerative and dystrophic
diseases resulting in a loss of most photoreceptor cells. Because a number of blinding
retinal diseases are due predominantly to outer retinal degeneration (specifically targeting
the photoreceptors) [120, 217, 235], the idea of stimulating the remaining inner retinal
10
cells came about. Early experiments showed that inner retinal layers can be electrically
stimulated and elicit an electrical-evoked response (EER) [118, 120, 131, 247].
Two of the most common outer retinal degenerative diseases are RP and AMD.
Postmortem morphometric analysis of the retina of RP patients revealed that many more
inner nuclear cells (bipolar cells and others [78.4%]) are retained compared to outer
nuclear layer (photoreceptors [4.9%]) and ganglion cell layer (29.7%) [119, 217]. Similar
results were obtained from AMD patients [129].
The retinal prosthesis has actually developed in two different directions. One is
called the epiretinal approach, in which the device is implanted into the vitreous cavity
and attached to the inner retinal surface. The second approach uses a potential space
between the neurosensory retinal and the retinal pigmented epithelium into which the
device is implanted (see Figure 1).
2.2.1 Epiretinal Visual Prosthesis
In the epiretinal approach, microelectrode arrays are controlled by electronic circuitry
connected via a radio frequency telemetry link to an image acquisition and processing
unit [143]. The image information is processed and transformed into electrical current
that is delivered wirelessly to the stimulating electrodes. The implant is located on the
vitreal side of the retina, contacting the retinal layer containing the retinal ganglion cells
(RGCs) and their axons. Electrical stimulation generated by the implant is supposed to
activate primarily the RGCs and possibly bipolar cells. Feasibility studies were conducted
on isolated retina [94], in cats [218, 268], and dogs [147]. Humayun et al were the
11
original inventors of the active epiretinal prosthesis in collaboration with the company
Second Sight Medical Products, which was founded after ten years of research [251].
Their first generation implant (ARGUS I) had 16 electrodes and was implanted in six
blind subjects between 2002 and 2004. Five subjects still use the device in their homes
today. These subjects, who were all completely blind prior to implantation, can now
perform a surprising array of tasks using the device. In 2007, the company announced
that it has received FDA approval to begin a trial of its second generation, sixty-electrode
implant, in the United States.
Figure 1: Comparison between epiretinal and subretinal visual prostheses. An object (in this
case a face) is projected by the cornea and lens onto the retina in an upside-down manner and is
transformed into an electrical image by the photoreceptor cells (rods and cones) of the outer
retina. With a subretinal implant, the rods and cones are replaced by a silicon plate carrying
thousands of light-sensitive microphotodiodes, each equipped with a stimulation electrode. Light
from the image directly modulates the microphotodiodes, and the electrodes inject tiny currents
into the remaining neural cells (horizontal cells, bipolar cells, amacrine cells, and ganglion cells)
of the retinal inner layer. In contrast, the epiretinal implant has no light-sensitive areas but
receives electrical signals from a distant camera and processing unit outside of the body.
Electrodes in the epiretinal implant (small black knobs) then directly stimulate the axons of the
inner-layer ganglion cells that form the optic nerve (adapted from Zrenner [276]).
12
The main advantages of an epiretinal visual prosthesis are less invasive
implantation requirements and easier spatial mapping onto the retinal ganglion cells
(RGC) [177]. Because of the high ocular rotational speed (more than 400 degrees/sec
[3]), one of the concerns has to do with adequate attachment of the implant to the retina.
Currently, retinal tacks are used to attach the multielectrode array to the retinal surface,
and it appears that the electrode array remained firmly affixed to the retina for up to 1
year of follow-up with no significant clinical or histological side effects [147]. Some
ARGUS I subjects have the implant in the eye for fives now without retacking, while in
one case the tack did become loose. Concerns have also been raised regarding the
preferential activation of passing RGC fibers in certain stimulation conditions and result
in aberrant perceptions [62, 106, 118, 120, 205]. Axonal stimulation remains an open
question yet to be fully characterized and addressed.
2.2.2 Subretinal Visual Prosthesis
Several groups have been developing subretinal prostheses [28, 275]. The subretinal
space is obviously a good location to stimulate the bipolar cells because of its proximity
to the inner nuclear layer (see Figure 1).
The artificial silicon retina (ASR), co-invented by Alan and Vincent Chow, is a
subretinal prosthetic device, measuring 2 mm in diameter. It is comprised of five
thousand microphotodiodes and stimulating electrodes, each measuring 20 × 20 µm in
dimension [28]. The density of the subunits is approximately 1100 subunits/mm
2
. These
subunits are designed to convert the light energy from images into electrical impulses to
13
stimulate the remaining functional cells of the retina in patients with AMD and RP. The
ASR is powered solely by incident light and does not require the use of external wires or
batteries. By converting light into electrochemical signals, the microscopic solar cells on
the chip mimic the function of the photoreceptors to stimulate the remaining viable cells
in the retina. Feasibility studies conducted on isolated retina [234, 275] as well as
biocompatibility and long-term side-effect studies performed on animal models [29, 95,
183, 222, 275] showed encouraging results. The subretinal device has been implanted in a
number of patients with RP. After 6-18 months, no significant side effects were noticed.
All the implants were electrically functional, but no objective benefits of the implant
itself have been demonstrated [31].
Several biocompatibility issues of the subretinal approach have been studied.
These include limitation of the implant power generation by size requirements [188], and
creation of a mechanical barrier between the retina and the choroids, which provides
nourishments to the outer layers of the retina [275].
Modern-day photovoltaic cells are extremely inefficient compared to the natural
photoreceptors of the eye. In order to activate neurons, solar cells need to provide
electrical energy many orders of magnitude higher than the amount of current produced
under typical retinal illumination. Zrenner’s team demonstrated, in a separate study of
subretinal implants, that currents generated by the microphotodiodes would be much too
low to activate the retinal elements. An additional energy source is thus compulsory.
Moreover, subretinal implants may impede nutrient diffusion to the outer retina [276],
and render the device harmful to the long-term survival of the retinal cells. The
14
mechanical barrier between the retina and the choroids might be responsible to patches of
fibrosis and RPE changes that were found after chronic subretinal implantation [276].
2.3 Optic Nerve Visual Prosthesis
A number of investigators have also tried to stimulate the optic nerve [224, 255]. The
optic nerve is a compact bundle of ganglion cell axons that transmit the final neural
output from the retina to the brain. The rationale for using a nerve cuff electrode is
completely distinct from all other approaches where an increased number of electrode
contacts would theoretically increase the resolution. In contrast, in the optic nerve cuff
approach, a few contacts are considered to be capable of generating a large number of
potential field shapes within the nerve. These fields will selectively activate different
parts of the nerve and thereby allow a large number of distinctly located phosphenes to be
generated. However, the phosphenes obtained by optic nerve stimulation remain small
[255]. This discrepancy between the projection of the distribution of activated optic nerve
axons and the size and localization of the perceived phosphenes is related to the poorly
known information encoding in the optic nerve. Despite this discrepancy, findings have
confirmed the classically described retinotopic arrangement of the axons in the pre-
chiasmatic optic nerve. It is believed that this retinotopic arrangement, the small size of
the perceived phosphenes and the possibility to control their position by adjusting the
stimulation parameters [48] make it possible to convey image information [49], even
without resorting to more complex selective stimulation schemes. Optic nerve spiral cuff
electrodes have been recently implanted in human subjects. Electrical stimuli applied to
15
the optic nerve produced localized, often colored, phosphenes, and phosphene sensations
can be reinduced 118 days after surgery. Changing the pulse duration, pulse amplitude, or
pulse repetition rate could vary phosphene brightness [255]. After training, the patient
could recognize different shapes, line orientations, and even letters [256]. Results
obtained in the field of pattern recognition [257] and object recognition and grasping
[136] are also promising.
The high density of the axons (1.2 million within a 2 mm-diameter cylindrical
structure) could make it difficult to achieve focal stimulation and detailed perception. In
addition, any surgical approach to the optic nerve requires dissection of the dura and can
have harmful adverse effects, including central nervous system infection and disturbance
of blood flow to the optic nerve. Similar to the retinal prosthesis approach, optic nerve
stimulation requires intact retinal ganglion cells and is limited to outer retinal pathologies
[151].
2.4 Intracortical Visual Prosthesis
Implicitly, each approach to the visual prosthesis aims at the production of a complete set
of small phosphenes that would neatly tile the visual field as pixels of an image. The
cortical surface electrode approach is not projected to be clinically useful due to large
power consumption and limited spatial resolution. Investigators have turn to look at
intracortical electrodes with contacts implanted in the immediate vicinity of the cells of
the visual cortex. In human trial, 38 microelectrodes were implanted in the right visual
cortex, near the occipital pole, for a period of 4 months. Thresholds were acceptable
16
[219]. Two-point resolution was about five times closer than had typically been achieved
with surface stimulation [19]. Later, 13 mm
2
arrays of 100 closely spaced
microelectrodes as well as an intracortical insertion method have been developed by
Normann’s group [177]. Studies done with this type of implant in cats conclude that
production of individual phosphenes would require an interelectrode spacing of 0.5-1.2
mm [178]. Biocompatibility remains a major point of concern. To date, no human
implantation of an intracortical device has yet been performed. An artist rendition of a
human intracortical visual prosthesis system is shown is Figure 2.
Figure 2: Drawing of an intracortical artificial vision system. The system consists of a
miniature video camera built into the nosepiece of the glasses, signal-processing electronics
(carried in a pocket), and arrays of microelectrodes implanted in the primary visual cortex. Inset
shows the magnified view of an example of an intracortical Utah Electrode Array (adapted from
Normann et al. [179]).
17
Visual Prosthesis
Approach
Advantages
Disadvantages
Epiretinal
● less invasive implantation
● easier spatial mapping
● more natural visual processing
● possible retinal ganglion cell axon
stimulation
● difficult to achieve consistent electrode
to retina proximity
● useful only in outer retinal pathologies
Subretinal
● direct stimulation of bipolar cells
● easier spatial mapping
● electrode-retina proximity less of an
issue
● insufficient energy for stimulation from
passive microphotodiode array devices
● limited subretinal space place restriction
on the size of the electrode array
● useful only in outer retinal pathologies
● surgical approach across choroid is
potentially difficult
Optic nerve
● few electrodes can produce different
phosphene patterns
● difficult to achieve focal stimulation and
visual perception
● invasive implant procedure
● viable for retinal pathologies only
Intracortical
● not limited to retinal pathologies
● skull provide rugged housing for
implant
● non-conformal retinotopic mapping
● multiple visual representation in cortex
● pattern stimulation does not produce
pattern perception
● convoluted cortical anatomy makes
implantation difficult
● invasive implant procedure with
potentially life-threatening complications
Table 1: Comparison of various approaches to prosthetic vision.
There are advantages and disadvantages that are associated with the cortical
stimulation approach (see Table 1). The skull will protect both the electronics and the
electrode array, and an intracortical visual prosthesis will bypass all diseased neurons
18
distal to the primary visual cortex. By doing so, it has the advantage to be applicable in
rehabilitating blindness of many causes other than RP and AMD alone. However, spatial
organization is more complex at the cortical level, and two adjacent cortical loci do not
necessarily map out to two adjacent areas in space, so that patterned electrical stimulation
may not produce patterned perception [151]. Moreover, every small area of the cortex,
even at the level of primary visual cortex, is highly specialized for color, motion, eye
preference, and other parameters of visual stimuli. Thus, it is unlikely to get simple
perceptions even when stimulating few hundreds of neurons in the case of intracortical
microstimulation. The convoluted cortical surface also makes it difficult for implantation,
and surgical complications can be devastating to human subjects [151].
19
Chapter 3: Human Visual Cortex Overview
3.1 Mapping Visual Cortex with Functional Neuroimaging
Functional Neuroimaging is used to acquire images of physiologic function in the brain.
Such images can be acquired by imaging the decay of radio-isotopes bound to molecules
with known biological properties, as in positron emission tomography (PET), or by
measuring differences in local magnetic fields generated by electron spin as in functional
magnetic resonance imaging (fMRI). Both approaches have been applied to the study of
human visual cortex because of their noninvasive nature and their ability to capture the
inner workings of the visual system in alert human subjects. Using visual simulation
paradigms of varying complexity, investigators are beginning to decode the various parts
of the visual cortex that are involved in vision.
3.2 Positron Emission Tomography (PET)
One of the pioneering PET experiments carried out to study the human visual cortex is
done by Michael Phelps [189]. In his experiment, Phelps used
14
C-deoxyglucose as the
radiotracer. He observed activation of the primary visual cortex during simple white light
visual stimulation. Stimulation with more complex visual scenes activated not only the
primary, but also secondary (association) visual areas. Since then, a number of studies
have been carried out using H
2
O
15
and
18
F-fluorodeoxyglucose (FDG) radiotracers to
measure cerebral blood flow and glucose metabolic rate, respectively, in the visual cortex
20
of human subjects following exposure to different visual stimuli. A detailed overview of
PET is given in Chapter 5.
3.3 Functional Magnetic Resonance Imaging (fMRI)
Since the first measurements of fMRI signals from human cortex were reported [134,
180], much progress has been made in both measurements and analysis of visually driven
activity. Prior to fMRI, neuroimaging primarily tested hypotheses about the localization
of function by asking whether two stimuli (or tasks) caused statistically different signals
within the brain. The signal-to-noise ratio (SNR) of fMRI is large enough to measure the
size of differences, not just their presence or absence. For this reason, hypotheses about
neural computations, beyond localization, can be framed and tested.
3.3.1 Biological Source of fMRI signal
The fMRI signal is an indirect measure of neural activity by measuring changes in the
local blood oxygenation level [170]. In a control condition, arteries supply nearly 100%
oxygenated blood and roughly 40% of the oxygen is consumed locally, so that the blood
returning in the veins comprises 60% oxygenated and 40% deoxygenated blood. When a
stimulus causes significant neural activity, an additional supply of oxygenated blood is
delivered, but not all of the incremental supply is metabolized. Because the fraction of
extracted oxygen is reduced compared with the control state, the proportion of
oxygenated to deoxygenated blood increases slightly. This change in the ratio of
oxygenated to deoxygenated blood alters the local magnetic field and can be detected in
21
the magnetic resonance signal. This signaling mechanism is known as the blood oxygen
level dependent (BOLD) signal.
It is important to address the question whether the BOLD signals arise only from
relatively large blood vessels, or whether they arise also from the fine vascular mesh
called parenchyma. Identifying the vascular source of the BOLD signals is important
because oxygenation levels in large veins reflect activity pooled over large brain regions.
To answer this question, measurements that exploited the retinotopic organization of
visual cortex were carried out [17, 64, 135]. Together the important findings demonstrate
the fine vascular parenchyma of the visual cortex as the point of origin of the BOLD
signal.
3.3.2 Phase-encoded Measurements with fMRI
Figure 3A shows a contracting ring stimulus. When it is seen through a small aperture in
the visual field, such as the receptive field of a neuron, the visual field alternates between
the flickering checkerboard and the neutral gray field. The alternation of the contrast
pattern and uniform field causes a waxing and waning of the neural response. If the
stimulus moves at a constant velocity from periphery to fovea, the alternation frequency
of the response will be the same for all points in the visual field. However, the temporal
phase of the response will differ. Neurons with receptive fields in the periphery will
respond earlier than neurons with receptive fields near the fovea. Hence, the phase of the
response defines the receptive field position along the dimension of eccentricity.
Retinotopic mapping methods that rely on this phase signal to map the visual field
22
responses are called phase-encoded. Figure 3B shows another stimulus example, a
checkerboard wedge that rotates counterclockwise as indicated by the arrows. In this
case, the phase-encoded dimension of interest is angular rotation.
Figure 3: Stimuli used for phase-encoded measurements of retinotopic organization. The
contracting ring (A) and rotating wedge (B) are shown at different time points with the direction
of motion of the stimuli indicated by the arrow. These stimuli cause a traveling wave of neural
activity to sweep across the cortex in either the eccentricity dimension (A) or the angular
dimension (B). (Courtesy of Wandell et al. [260])
3.3.3 Flattened Cortical Representation
The calcarine sulcus, like most large brain structures, does not stay within a plane over a
large distance. To obtain a more complete representation of the activity in the calcarine
sulcus, measurements can be integrated from several different images. One way to do this
is to take advantage of the fact that neocortical gray matter is much like a folded sheet.
Segment the gray matter in a series of image planes and then computational algorithms
23
can be used to unfold the cortical sheet, representing it in a single image known as
flattened cortical representation [24, 41, 56, 242].
3.4 Overview of Human Visual Cortical Areas
Recent advances in neuroimaging methods have allowed us to measure in intact, alert
human brain cortical responses that are associated with visual motion, color, and pattern
perception. In individual human brain, we can now identify the positions of several
retinotopically organized visual areas; measure retinotopic organization within these
areas; identify the location of a motion-sensitive region in individual brains; and measure
responses associated with contrast and color [260].
Over the past 40 years, our understanding of the visual cortex has been
revolutionized by the identification and analysis of a set of roughly 30 visual areas in
macaque monkeys [67, 254, 271]. These visual areas have been defined using a variety of
criteria, including (a) anatomical connections, (b) stimulus selectivity of single-unit
responses, (c) architectural properties (e.g. immunohistochemistry), and (d ) retinotopic
organization. Although the function of most of these areas is unclear, studies of the visual
cortex in lower primates [50, 250, 262, 271] and clinical correlation with cerebral lesions
in patients [115, 253, 203] along with electrophysiologic studies [13, 69], postmortem
histologic examinations [22, 33], and functional imaging studies [37, 63, 223, 246, 272]
have identified cortical areas that are important in human vision.
Using functional brain imaging, four successive polar representations of visual
field quadrants were shown to exist in the occipital cortex through phase-encoded retinal
24
stimulation [51, 223]. Subsequent studies demonstrated the involvement of several
cortical areas along the ventral visual stream (occipital–temporal network) in various
aspects of visual form analysis and color constancy [9, 127, 148]. Concurrent studies
mapped the dorsal visual stream (occipital–parietal–frontal networks) involved in
complex visual functions such as saccade control [107, 186], visual attention [7, 12, 36],
spatial visual processing [103, 191, 249], and motion perception [182, 237].
3.4.1 Retinotopic Organization of Human Visual Areas V1/2/3
Within each hemisphere, the primary visual cortex (V1 or Brodmann Area 17) occupies a
roughly 4- by 8-cm area located at the posterior pole of the brain in the occipital lobe. A
large fraction of area V1 falls in the calcarine sulcus (Figure 4). Studies of patients with
localized cortical damage showed that the receptive fields of neurons within area V1 are
retinotopically organized [111, 112, 114]. From posterior to anterior cortex, the visual
field representation shifts from the center (fovea) to the periphery. This dimension of
retinotopic organization is commonly referred to as eccentricity. Because of the
retinotopic organization of primary visual cortex (V1) and several nearby association
visual areas (V2 and V3), it is possible to create stimuli that control the location of the
neural activity. Figure 3A represents a flickering ring that contracts slowly from the
periphery to the center of the visual field. When the ring reaches the center of the viewing
aperture, it is replaced by a new ring at the edge of the display. This contracting ring
stimulus creates a traveling wave of neural activity that spreads from anterior to posterior
cortex, along the eccentricity dimension. The activity spans several adjacent visual areas,
25
including V2 and V3, which share a similar retinotopic organization with respect to
eccentricity. Figure 3B represents a flickering wedge that rotates slowly about fixation.
This rotating wedge stimulus creates a traveling wave of neural activity that spreads from
the lower to the upper lip of the calcarine sulcus. Using phase-encoded fMRI technique
and the flickering checkerboard patterns shown in Figure 4, Dougherty et al.
demonstrated this retinotopic paradigm [54].
The angular dimension of the retinotopic organization is represented on a path
that traverses from the lower to the upper lip of the calcarine sulcus. Along this path, the
visual field representation shifts from the upper vertical meridian through the horizontal
meridian to the lower vertical meridian (Figure 5B). Because of the different angular
representations in V1 and V2 (Figure 5B), the traveling wave reverses direction at the
border. Similarly, although not shown, visual area V3 surrounds V2 and the traveling
wave reverses direction again at the V2/V3 border. Hence, measurements using the
rotating wedge stimulus (Figure 3B) reveal the boundaries between retinotopically
organized visual areas.
By measuring the locations of the traveling wave reversals using fMRI, the visual
area boundaries can be located [51, 64, 223]. Figure 6 shows the angular dimension of
retinotopic organization on a flattened cortical representation. The colors represent the
angular position in the visual field of the rotating wedge that activated the region. The
visual area boundaries, derived from phase-encoded fMRI measurements, are placed at
locations where the traveling wave reversed direction.
26
Figure 4. The Retinotopy paradigm. Two stimuli are used to measure the retinotopic maps in
cortex. Expanding ring stimuli map eccentricity and rotating wedge stimuli map polar angle. The
phase of the best-fitting sinusoid for each voxel indicates the position in the visual field that
produces the maximal activation for that voxel. Thus, these pseudo-color phase maps are used to
visualize the retinotopic maps. Data are shown for the left hemisphere of one human subject.
(Adapted from Dougherty et al. [54])
Area V1, in calcarine cortex of each hemisphere, represents an entire visual
hemifield. The dorsal and ventral boundaries of V1 are separated by roughly 4 cm and
fall on the cortical surface at the upper and lower lips of the calcarine. V1 has a very
well-defined retinotopy mapping of the spatial information in vision. For example, in
humans the upper bank of the calcarine sulcus responds strongly to the lower half of
visual field (below the center), and the lower bank of the calcarine to the upper half of
visual field. In human and animals with a fovea in the retina, a large portion of the V1 is
27
Figure 5: Retinotopic organization of visual areas. (A) Icons indicating the right visual field
positions of the central fixation and periphery (black dot with white surround; white dot with
black surround. respectively) and the horizontal and vertical meridia (dashed lines) are shown.
(B) The approximate positions of areas V1 (hatch marks) and V2 (dotted area) in a sketch of the
medial surface of the left occipital lobe are shown. The visual field icons are superimposed on the
sketch to indicate the retinotopic organization within areas V1 and V2. The positions of the
fundus of the calcarine sulcus (Ca) and parieto-occipital sulcus (PO) are also shown. [Courtesy of
Wandell et al, 1999] (C) Midsagittal view of the right cerebral hemisphere; the occipital lobe
above the calcarine sulcus (blue line) is overlaid in pink, and the portion of the occipital lobe
below the calcarine sulcus is colored overlaid yellow. Red line indicates the parieto-occipital
sulcus. (Adapted from Haines [97])
28
Figure 6: Visual areas on flattened human occipital cortex. The maps are generated from
fMRI data. The color at each position on the cortical map shows the visual stimulus angular
position (see middle image) that maximally evoked activity at that position. The boundaries
between the visual areas are shown (dashed white lines), and their positions are determined from
reversal in the direction of traveling wave caused by a rotating wedge. Visual area labels are
superimposed in black on the maps. The yellow dashed line in the left image that bisects V1
represents the calcarine fissure. (Courtesy of Stanford Vision and Imaging Science and
Technology [233])
mapped to a small central portion of the visual field, a phenomenon known as cortical
magnification [64, 114, 223]. Area V1 is bordered by two cortical regions, each roughly 1
cm, which form area V2. Dorsal V2 (V2d) represents the lower quarter of the visual field
and ventral V2 (V2v) represents the upper visual field. V2 is the second major area in the
visual cortex. It receives strong feedforward connections from V1 and sends strong
connections to V3, V4, and V5. It also sends strong feedback connections to V1. The
strip V2v (V2d) is bounded by another distinct cortical strip called V3v (V3d). Like their
neighbors, V3d and V3v each represent one quarter of the visual field, and each spans
29
roughly 1 cm of width and 6–8 cm of length. An additional retinotopically organized
area, V3A, is adjacent to V3d [246]. Area V3A represents the entire hemifield.
Figure 7: The dorsal and ventral visual pathways. Separate pathways to the temporal and
parietal cortices course through the extrastriate cortex beginning in V2. The connections shown in
the figure are based on established anatomical connections, but only selected connections are
shown and many cortical areas are omitted. Note the cross connections between the two pathways
in several cortical areas. The parietal pathway receives input from the M (magnocellular)
pathway, but only the temporal pathway receives input from both the M and P (parvocellular)
pathways. (Abbreviations: AIT = anterior inferior temporal area; CIT = central inferior temporal
area; LGN = lateral geniculate nucleus; LIP = lateral intraparietal area; Magno = magnocellular
layers of the lateral geniculate nucleus; MST = medial superior temporal area; MT = middle
temporal area; Parvo = parvocellular layers of the lateral geniculate nucleus; PIT = posterior
inferior temporal area; VIP = ventral intraparietal area.) (Based on Kandel, Schwartz and Jessell
[126])
30
3.4.2 Dorsal and Ventral Visual Streams
The signals from the two retinas are communicated to the primary visual cortex (area V1)
via the lateral geniculate nucleus (a subdivision of the thalamus). After the signals are
processed in V1 they are communicated via multiple pathways to other visually
responsive exstrastriate cortical areas. Two important visual pathways have been studied
extensively, and they are named based on their anatomical distribution as the “dorsal
visual stream”, which stretches from the occipital cortex forward into the parietal lobe,
and the “ventral visual stream”, which passes into the inferior temporal cortex (Figure 7).
3.4.3 Dorsal Visual Stream
The dorsal stream begins with V1, goes through visual area V2, then to the dorsomedial
area and visual area MT (also known as V5) and to the inferior parietal lobule. The dorsal
stream, sometimes called the “Where pathway”, is associated with motion, representation
of object locations, and control of the eyes and arms, especially when visual information
is used to guide saccades or reaching [89]. It is interconnected with the parallel ventral
stream (the “What pathway”) which runs downward from V1 into the temporal lobe.
The dorsal stream commences with purely visual functions in the occipital lobe
before gradually transferring to spatial awareness at its termination in the parietal lobe.
The posterior parietal cortex is essential for the perception and interpretation of spatial
relationships, and the learning of tasks involving coordination of the body in space [11].
It contains individual functioning lobules, one area of which contains neurons that
produce enhanced activation when attention is moved onto the stimulus (LIP-lateral
31
intraparietal area), and another section where visual and somatosensory information are
integrated (VIP-ventral intraparietal area). In the human brain an area devoted to motion
has been identified at the junction of the parietal, temporal, and occipital cortices. This
area is labeled MT (middle temporal area or V5), and almost all of the cells in the MT
region are directionally selective and the activity of only a small fraction of these cells is
substantially altered by the shape or the color of the moving stimulus. MT has a retintopic
map of the contralateral visual field, but the receptive fields of cells within this map are
about 10 times wider than those cells in the striate cortex [20, 32, 57, 153]. A cortical
area adjacent to MT, the medial superior temporal area (MST), also has neurons that are
responsive to visual motion and these neurons appears to process a type of global motion
in the visual field, which is important for a person’s own movements through an
environment [58, 92, 248].
3.4.4 Ventral Visual Stream
The ventral stream begins with V1, goes through visual area V2, then though visual area
V4, and to the inferior temporal lobe. From caudal to rostral, the ventral visual stream
consists of visual areas V1, V2, V4, and the areas of the inferior temporal lobe: PIT
(posterior inferior temporal), CIT (central inferior temporal), and AIT (anterior inferior
temporal) areas (see Figure 7). The ventral stream, sometimes called the “What
Pathway”, is associated with form recognition and object representation. It also has
strong connections to the medial temporal lobe (which stores long-term memory), the
32
limbic system (which controls emotions), and the dorsal visual stream (which deals with
object locations and motion).
Visual area V4 receives strong feedforward input from V2 and sends strong
connections to the posterior inferior temporal area (PIT). It also receives direct inputs
from V1, especially for central visual space. V4 shows strong attentional modulation
[166]. It is now agreed upon that V4 is responsive to both color and form [110, 132, 145,
154, 272].
3.5 Monocular Photic Stimulation
3.5.1 Findings from fMRI Studies
A few functional MRI studies have looked at cortical activation during monocular full-
field photic stimulation of the human eye [158, 159, 209, 245]. The findings from these
studies are consistent and can be summarized by Figure 8.
Visual stimulation was done using two LED matrices that were fitted into
modified goggles worn over the subject’s eyes. The contraption emitted 8 Hz flashing
red-light which gave full-field stimulation of each eye. The investigators had control over
which LED matrix was turned on; therefore, each eye could be stimulated separately or
together. These studies demonstrated asymmetric activation patterns in the visual cortex
of normal humans during monocular photic stimulation (Figure 8). The contralateral
hemisphere was activated more strongly and to a greater spatial extent than the ipsilateral
hemisphere when either eye was stimulated. The right anterior visual cortex was
33
activated only during left eye stimulation, and the left anterior visual cortex was activated
only during right eye stimulation. The area in the visual cortex containing the difference
in monocular stimulation lied on the calcarine fissure; however, it was not confined to the
area corresponding to the monocular temporal cresent [26, 114, 137, 159].
Miki and Toosy et al. [158, 245] attributed this finding to nasotemporal
asymmetries (see Section 3.5.2). In part, this could be explained by the representation of
the monocular crescent of the temporal hemifield of either eye (Figure 9), which exists
only in the crossed projection. In addition, within the binocular field, there is a biased
crossed projection of nasal retinal ganglion cells which may drive the contralateral ocular
dominance columns in V1. Using retrograde tracers, Fukuda et al. [83] found that near
the split between nasal and temporal hemiretina (perifoveal region), large ganglion cells
(α cells) that reside in the temporal hemiretina project to the contralateral lateral
geniculate nucleus (LGN); whereas, medium size ganglion cells (β cells) project to the
ipsilateral LGN. ‘α cells' have a fast temporal response and are tuned to temporal changes
and bland spatial content, whereas β cells are tuned to spatial details with a much slower
temporal response. Since the light stimulus used for the full-field photic stimulation was a
flashing red light (8 Hz) with little spatial detail, it is expected to predominantly activate
α cells in the temporal hemiretina which crossed the optic chiasm to reach the
contralateral visual cortex. This would also result in a more significant activation of the
left visual cortex with right eye stimulation. Fukuda attributed this uneven overlap of
retinal ganglion cell projection to possible mechanism in binocular stereopsis [83].
Rombouts et al. [209] studied this asymmetric visual cortex activation from the
34
Figure 8: Monocular photic stimulation summary. The results from group analysis (n = 8)
displayed in three orthogonal slices. The SPM{Z} map is overlaid on SPM96 T1 template,
showing the area responded to the left eye more than the right eye (1) and the area responded to
the right eye more than the left eye (2). Left side of the brain is on the upper side of the transverse
images and the left side of the coronal images. The threshold was set at p<0.001 (uncorrected).
(Adapted from Figures 1 & 2, Miki et al. [159])
Figure 9: Schematic diagram of visual field projection to primary visual cortex. The
temporal crescents, labeled in the diagram, are regions of the visual field that lie immediately
outside of the binocular visual field of both eyes which project to the contralateral visual cortex.
(Courtesy of http://thebrain.mcgill.ca/)
35
perspective of ocular dominance (see Section 3.5.3). He found more extensive activation
of the contralateral visual cortex during monocular photic stimulation of the dominant
eye.
3.5.2 Nasotemporal Asymmetry in the Human Eye
Nasotemporal asymmetry in the distribution of photoreceptors [38] and retinal ganglion
cells [39] and in visual perception [65, 93] has been shown in humans.
The topography of the density of photoreceptors and retinal ganglion cells in
humans shows a nasotemporal asymmetry [38, 39]. Because the cone density is higher in
nasal compared to temporal retina at equivalent eccentricities and the nasal retina is
larger, there is 17-63% more cones in the nasal retina. Similarly, the entire nasal
hemiretina has 20-50% more rods than the entire temporal hemiretina. There is 35-66%
more ganglion cells in the nasal retina than in temporal retina [39]. Even in central retina
from 0.4 to 2.0 mm eccentricites, higher ganglion cell densities are found in the nasal
retina [39]. The normal visual field of an eye is not symmetrical and is wider in the
temporal visual field (about 100 degrees) than in the nasal visual field (about 60 degrees).
Also, the retinal sensitivity as measured by psychophysical experiments is asymmetric for
the nasal and temporal retina (nasal > temporal) in the peripheral portion (especially
beyond 20 degrees eccentricity) [65, 93].
36
3.5.3 Effects of Ocular Dominance
Ocular dominance, sometimes also called eye dominance, is the tendency to prefer visual
input from one eye to the other. Approximately two-thirds of the population is right-eye
dominant [25, 61, 202]. Porac and Coren [193] found that 65% of the subjects studied in
a group of 160 showed right eye dominance, 32% left eye dominance and 3% no
consistent dominance. In those with anisometropic myopia (different amounts of
nearsightedness between the two eyes), the dominant eye has been found to be the one
with more myopia. The laterality of eye and limb do not appear to be generally correlated
[193]. Sex differences emerge indicating more consistent eye and limb preferences as
well as stronger eye dominance in male subjects than female subjects [193].
To examine the functional basis of ocular dominance, Rombouts et al. used fMRI
to study visual cortex activation during photic stimulation with the modified LED
goggles (flashing red light, 8 Hz) each worn in front of one eye [209]. In each subject, the
left and right eye was stimulated separately and together, in a randomly alternating order.
They found the size of the activated area was bigger upon monocular stimulation of the
dominant eye than the non-dominant eye (Figure 10). In addition, activated areas in the
visual cortex of subjects with a dominant right eye was more extensive in the
contralateral (left) hemisphere with activation extending from the occipital pole to more
anterior striate cortex.
37
Figure 10: The role of ocular dominance in visual cortex activation. Left image, the black line
marks the location of the fMRI measurement, which is parallel to the calcarine sulcus; middle
image, shows activated areas in the visual cortex of a subject with a dominant right eye,
superimposed on the MRI anatomical image. The areas of activation are shaded in three different
gray levels (as indicated in the legend on the bottom) to represent the three different conditions of
photic stimulation. The right image shows areas of visual cortex activation in a subject with a
dominant left eye; same shading scheme for the three stimulation conditions as in the middle
image is applied here. (Adapted from Rombouts et al. [209])
3.6 Visual Cortex and Blindness
3.6.1 Cortical Effects of Early Visual Deprivation
The eye captures light but the brain is where vision is experienced. The capacity of retinal
prostheses to restore some vision in blind individuals has been demonstrated at the retinal
level, but it is unknown how the brain will be receptive to the new neural message after a
38
prolonged period of visual deprivation resulting from early-onset retinal degenerative
diseases.
Cortical effects of early visual deprivation are extensively studied [42, 108].
Visual deprivation in animals shortly after birth leads to dramatic reduction of visually
responsive neurons within cortical visual areas [201]. The timing and type of deprivation
affects the character and severity of alteration of cortical function [42, 108, 267].
Binocular eyelid suture, which produces modest light attenuation but severe form
deprivation, produces greater abnormalities in cortical physiology than an equivalent
period of dark rearing [171]. The standard model is that early visual experience during a
critical period of neuronal plasticity defines the response properties of cortical visual
neurons and that after this period these properties become relatively immutable [108].
It is unknown, however, whether patients with severe visual loss from birth have
any receptive cortical substrates for restored retinal input. Research suggests early-blind
patients could show markedly abnormal anatomy in the post-retinal visual pathways. A
diffusion tensor imaging study of early-blind patients demonstrated atrophic or absent
optic nerves and geniculocortical tracts [227], while a voxel-based morphometry analysis
revealed atrophy of cortical gray matter in early visual areas [176].
A few case reports have examined recovery of vision in adulthood following
relatively late treatment of ophthalmic diseases, typically lens or corneal opacities. In one
particularly well-studied case, correction of anterior segment disease in adulthood
resulted in limited recovery of vision and markedly reduced cortical responses to visual
stimuli, particularly in extrastriate areas as demonstrated with fMRI [70]. Indeed,
39
persistent deficits in integrative visual functions are seen even when bilateral cataracts are
treated in childhood [155], although there is recent intriguing evidence that recovery of
some higher-level visual function is possible with eye treatment during adulthood [229].
3.6.2 Cross-modal Plasticity
Cortex deprived of stimulation from its primary modality could become instead
responsive to alternative sensory modalities [10] as the brain reorganizes itself to exploit
other sensory inputs at its disposal. This is known as cross-modal plasticity where the
loss of input from one sensory modality leads to reorganization in the brain
representation so that other senses can now use the brain area that was originally devoted
for the processing of the lost sensory input. In fact, the loss of light has been associated
with superior non-visual perception in the blind. This well-described phenomenon may
underlie enhanced tactile and auditory skills in blind individuals. Despite their lack of
vision, blind individuals manifest remarkable abilities in many aspects of life. On the
perceptual level, these abilities are often observed in tasks such as auditory spatial
localization [142, 207, 258], pitch discrimination [90], and somatosensory-based Braille
reading [212]. It is generally postulated that such abilities rest partly on plastic changes
within the brain, involving cross-modal reorganization [140, 142, 263].
Studies in functional imaging also demonstrated a modulated occipital-cortex
activation in blind individuals during tactile [46, 198, 213, 216]
and auditory tasks [6,
259], suggesting that the visual cortex deprived of its normal inputs has adopted a new
role in information processing. Furthermore, the hypothesis proposing a new functional
40
Figure 11: Tactile sensory cross-modal plasticity in early blind subjects. (A) Statistical
parametric maps of activation within the group of blind subjects reading Braille with their right
hand compared with rest period. Task-related increase in magnetic resonance signal is
superimposed on three orthogonal sections of three-dimensional T1-weighted standard brain.
Statistical corrected threshold is P < 0.05. Results show activation of visual cortex accompanied
by activation of sensorimotor areas. (B) Statistical parametric maps of activation within the group
of normal sighted subjects “reading” Braille with their right hand compared with rest period.
Statistical corrected threshold is P < 0.05. Results show expected activation of sensorimotor areas
and no activation of visual cortex. (Courtesy of Gizewski et al. [86])
Figure 12: Auditory sensory cross-modal plasticity in early blind subjects. Monaural sound
localization in PET-H
2
O
15
experiments performed in the three groups of subjects: (1) SIG –
sighted normal; (2) EBNP – early blind with normal sound-localization performance; (3) EBSP –
early blind with superior sound-localization performance. Cerebral blood flow (CBF) increases
are plotted in pseudo color overlay on the MRI anatomic brain slices. Activations of the right
striate and extrastriate cortices are observed in EBSP but not in the two other groups for the
contrast. (Courtesy of Gougoux et al. [91])
41
role for visually deafferented cortex has received support from studies demonstrating an
increase in Braille reading errors in blind but not sighted individuals following
transcranial magnetic stimulation of the occipital cortex [34], the loss of the ability to
read Braille in a patient who suffered bilateral occipital damage due to an ischemic stroke
[100], and from numerous neuroimaging studies showing strong correlations between
occipital activity and superior behavioral abilities [23, 86, 91, 214]. The fMRI images in
Figure 11 demonstrate the concept of tactile sensory cross-modal plasticity in early blind
individuals [86]. Contrasting with normal sighted controls, there is a clear activation of
visual cortex during Braille reading in early-blind subjects. Results from an auditory
localization experiment done on early-blind subjects (Figure 12) demonstrate cross-
modal utilization of occipital cortex in blind subjects who have superior performance in
sound localization.
3.7 Visual Cortex Remodeling in Retinal Degenerative Diseases
Mammalian retinal degenerations initiated by gene defects in rods, cones or the retinal
pigmented epithelium often trigger loss of the sensory retina, effectively leaving the
neural retina deafferented. In animal models, neural retina responds to this challenge by
remodeling, first by subtle changes in neuronal structure and later by large-scale
reorganization [124, 149, 236]. Similar synaptic reorganization at the retinal level is also
seen in humans with retinitis pigmentosa [66]. It is possible that this type of retinal
remodeling can alter normal light transmission pathway in the retina and result in
42
different activation patterns of ganglion cell groups between light and electric
stimulation.
Using fMRI, Baker et al [8] studied reorganization of visual processing in
macular degeneration (MD). In the two MD subjects, one had cone-rod dystrophy, while
the other had a juvenile form of MD. They found activation of foveal cortex by peripheral
visual stimuli in both subjects. Such activation was not observed with visual stimuli
presented to the same retinal position in sighted controls. Poggel et al. [192] investigated
the influence of different patterns of visual field loss on the topography of brain
activation. Using fMRI on patients with age-related macular degeneration (AMD) and
retinitis pigmentosa (RP), they observed blindness in the central visual field in AMD
patients led to silencing of projection areas of the scotoma; whereas in RP, activation
spreaded into cortical areas representing the blind periphery. Both studies suggest a
reorganization of visual cortex activity in response to gradual retinal degeneration.
In a psychophysical case study [4], a patient who developed retinitis pigmentosa
in childhood and became blind at the age of 40, started to experience visual sensation
during tactile stimuli on the hand. The study demonstrates somatosensory cross-modal
plasticity as a result of visual deafferentation.
A better understanding of the remodeling capacity of the brain in patients with
retinal degenerative diseases will lend itself to answering some of the fundamental
questions regarding the viability and functionality of retinal prostheses at the cortical
level.
43
3.8 Plasticity in Matured Brain
The mature visual system of primates and other mammals is capable of extensive
reorganization as the roles of inputs and pathways are altered by visual experience,
sensory loss, or cortical lesions. Although this plasticity declines with age [14], adult
visual cortex can still respond to experience with plastic changes as shown by the effects
of perceptual learning [220] and retinal lesions [55]. This plastic change in the brain
probably will allow blind subjects to extract greater information from touch and hearing,
thus improving the quality of life and enhancing the integration of the blind in the social
and working environment of a sighted society. The understanding of these neuroplastic
processes will provide the scientific foundation for improved rehabilitation and teaching
strategies for the blind. In addition, the modulation of such plasticity will be crucial in
developing and projecting the success of novel, visual neuroprosthetic strategies. It is
hoped that this neural plasticity will allow quick re-association of the “phosphenes
world” with the physical visual world.
3.9 Lessons Learned from Cochlear Implant
In terms of restoring lost sensory function, cochlear implant research has arguably been
the most successful and has translated into a viable therapeutic option for some
profoundly deaf patients [144, 157, 185]. Simply explained, a wire electrode bundle is
surgically inserted into the inner ear thus serving as an electronic replacement for
damaged hair cells. A microphone and speech processor picks up and decodes sounds
44
from the environment and the coded signals in turn drive the electrodes that stimulate the
nerve fibers of the cochlear, creating the sensation of hearing. In conjunction with intense
rehabilitation programs that are tailored to the individual, deaf patients can learn to
comprehend and in some cases, even acquire speech. By analogy, immediately after the
microelectrodes are implanted on the retina, the expected phosphenes are likely to
engender a poor perception of any object. However after some time and adequate
training, the reorganization and cognitive abilities of the highly adaptive neuronal
networks of the brain will result in improved performance and more concordant
perceptions. Furthermore, the patterns of stimulation are likely to change over time due to
factors such as the progression of disease, changes in electrode stimulation thresholds and
the effects of learning and plasticity, thus creating new associations between the patterns
of electrical stimulation and the visual patterns perceived.
Modern imaging techniques for measuring brain activity in humans have provided
evidence for plasticity of the central auditory pathway following a profound hearing loss
[16, 85, 101, 109, 122, 175, 181, 184]. Collectively, these studies reported low levels of
auditory cortical activity among profoundly deaf subjects – the longer the duration of
deafness, the lower the level of activity recorded.
Evidence from research investigating neuroplasticity in the deaf subjects before
and after cochlear implantation bears striking parallels with the case seen in visual
prostheses development. For example, similar to activation of the occipital cortex by
auditory stimuli in the blind, the auditory cortex (Brodmann’s areas 41, 42 and 22) is
activated in deaf subjects in response to visual stimuli [71], as well as in the observation
45
of sign-language [174, 187, 215]. Thus, as in the case of the blind, the removal of one
sensory modality leads to neural reorganization of the remaining ones. Visual-auditory
plasticity in the deaf has also been found in patients with cochlear implants. Functional
imaging studies indicate that after implantation of a cochlear device, primary auditory
cortex is activated by the sound of spoken words in deaf patients who had lost hearing
before the development of language. Interestingly, the levels of baseline activity before
implantation were positively correlated with the extent of language improvement
following the operation. It appears that if metabolism in the auditory cortex is restored by
cross-modal plasticity changes before implantation, the auditory cortex no longer
responds to signals from a cochlear implant and patients do not show improvement in
language capabilities [141]. These results have important implications and may even have
parallel implications with respect to neuroplastic changes following blindness.
46
Chapter 4: Transcorneal Electrical Stimulation
4.1 Introduction
Since the first demonstration that an electrical current passing through the cornea can
evoke a visual cortical response [195], a number of studies have been conducted to tap
into the potential of transcorneal electrical stimulation (TcES) as a diagnostic tool for
evaluating retinal function and therapeutic strategy for rescuing photoreceptors and
retinal ganglion cells.
Several groups [82, 163, 196] have used TcES to estimate the residual function of
retinal ganglion cells by the threshold current to evoke phosphenes in RP patients. It was
demonstrated that threshold current to evoke phosphene in RP patients was significantly
higher than in normal subjects [84, 168].
In addition to eliciting phosphene sensation, various studies have shown the
neurotropic effects of transcorneal electrical stimulation such as rescuing of retinal
ganglion cells [167], improving visual functions after optic nerve injury [81, 165], and
promoting the survival of photoreceptors in animal models of retinal degeneration [169].
Possible mechanism may involve the up-regulation of neurotrophic factors such as
insulin-like growth factor (IGF-1) [167].
4.1.1 Electrically Evoked Phosphenes
Phosphenes are defined as luminous visual sensation caused by excitation of the visual
47
system with sources other than light. They can be elicited by mechanical force, eye
movements, chemical agents, radiation outside the visible spectrum, magnetic fields, as
well as electrical currents passing through the eye [266]. Recently, the use of electrically
evoked phosphenes (EEPs) has increased to the examination of eyes that are to undergo
implantation of a visual prosthesis [47, 270, 118, 168].
4.1.2 Retinal Origin of Electrically Evoked Response
Potts et al. [195] first demonstrated that subjective phosphenes can be elicited by
applying a current through the eye, thereby coining the term “electrically evoked
response of the visual system (EER).” Dorfman et al [53] named this evoked potential
“visually electrically evoked potential” (VEEP).
The EER is the evoked visual potential elicited by transcorneal electrical
stimulation of the eyeball. The latency of the EER is shorter than that of the visually
evoked potential (VEP) of the same subjective brightness, which suggests that at least
one step in the retinal transmission chain has been bypassed [195]. Various investigators
have designed experiments to identify the site of retinal stimulation for the EER with
human subjects [161, 195, 196, 238] and animals [128, 197]. In rats with complete loss of
photoreceptor outer segments and nuclei, it was discovered that the VEP and the
electroretinogram (ERG) were completely absent, while the EER was normal [197]. In
patients with a dysfunctional rod or cone visual pathway, the EER was nearly normal
[162], although in patients with central retinal artery occlusion or optic nerve disease, the
EER was reduced [163, 164]. These findings indicate that the site activated by electrical
48
stimulation in the retina is more proximal than the photoreceptor. Shimazu et al. [226]
investigated the response properties of retinal ganglion cells to transcorneal electrical
stimuli. After administering APB (a glutamate agonist that eliminates the light responses
of depolarizing ON-bipolar cells, but has no significant effect on any other retinal cells),
they found the responses of retinal ganglion cells to transcorneal electrical stimulation
were significantly reduced. This suggests that the ON-bipolar cells and their related
synaptic sites are involved in transcorneal electrical stimulation. In addition, they also
found oscillatory discharges of retinal ganglion cells evoked by transcorneal electrical
stimulation. They attributed the phenomenon to interactions between center-surround
receptive fields which are constituted at the bipolar cell level.
4.1.3 Value of TcES in Implant Candidate Evaluation
Accurate methods to identify blind subjects with retinal cells that are sensitive to low-
level electrical current will improve the effectiveness of a microelectronic retinal implant
designed to restore useful vision to patients with outer retinal degenerations.
Conventional means of assessing vision are nearly impossible in eyes with bare light
perception or no light perception. Since light perception requires some intact
photoreceptors. In severe cases of retinal degeneration, where most photoreceptors are
lost, light perception is no longer possible even with the brightest light flash; we can still
evaluate the function of the remaining inner retinal layers with electrically evoked
response (EER). Yanai et al. [270] have demonstrated the value of EER in the selection
of blind patients for a permanent microelectronic retinal implant. They have shown that
49
RP patients with lower vision correlated well with higher EER threshold current for
eliciting phosphene sensation. Lower EER thresholds also correlated strongly with lower
intraocular stimulation current thresholds. This close correlation allows for predictions of
the required stimulus current for each implant patient. A device with lower current
requirements can use a large number of small electrodes, since the stimulus current
requirement sets a minimum size for electrodes. Thus, a reliable estimate of these
requirements can identify better candidates for long-term implants with a higher density
of electrodes and potentially greater resolution [121].
4.2 Psychophysical Techniques
Psychophysical procedures can provide a measure of visual function with both light and
electrical stimuli. Unlike electrophysiological responses (i.e. ERG), psychophysical
responses represent the properties of the entire visual pathway, from photoreceptors to
cortex. Furthermore, psychophysical measurements are subject to the potential influence
of cognitive factors such as attention. Nevertheless, psychophysical procedures,
particularly in combination with electrophysiological techniques, can provide important
insights into the visual system [106].
4.2.1 Concepts of Psychophysics
The general aim of psychophysical methods is to relate sensory states to the physical
properties of stimuli (i.e. visual or electrical stimuli) in a quantitative manner. The
50
physical properties can be easily obtained through the appropriate instrumentation, such
as photometers. Information about sensory states is less readily available. Observers
typically communicate information about sensory states through a verbal or motor
response.
Psychophysical methods and procedures are useful in determining threshold,
including visual field analysis. For a perfect observer, threshold is the point where the
stimulus can just be detected. Humans are not perfect observers, and often thresholds are
defined in probabilistic terms (for example, half the stimuli presented would be detected
and half would not). So under certain psychophysical techniques, threshold can be
considered the point where 50% of the stimuli are detected. Threshold variability most
likely depends on neural noise.
4.2.2 Visual Threshold Methods
Visual perceptual threshold measurement can be achieved through several methods. They
include 1) method of adjustment; 2) method of limits; and 3) method of constant stimuli
[106].
The “method of adjustment” involves asking the subject to either increase the
stimulus intensity from non-seeing until the stimulus can just be seen or to decrease the
stimulus intensity until the stimulus has just disappeared.
The “method of limits” involves presenting a stimulus well above threshold and
decreasing the stimulus intensity in small steps until the subject can no longer detect the
stimulus. This is called descending limits. Ascending limits is when a stimulus is first
51
presented well below threshold, then the stimulus intensity is increased to reach
threshold. Threshold is considered the average of the threshold points estimated by
several ascending and descending limits.
Both methods above suffer from errors of habituation and anticipation, but are
useful to obtain an estimate of threshold that can be investigated with more complex
techniques. The error of habituation occurs when subjects develop a habit of responding
to a stimulus. For example, in ascending limits, the subject may respond to seeing the
stimulus three steps past the threshold every time, thus giving a false threshold point. The
error of anticipation occurs when subjects prematurely report seeing the stimulus before
the threshold has been reached.
A third method, the “method of constant stimuli”, involves repeated presentation
of a number of stimuli of the same intensity. These stimuli are randomly presented in a
train, and the threshold value is defined as the value where 50% of the stimuli are
detected correctly.
We combined both “method of limits” and “method of constant stimuli” for
threshold determination in the experiment. First, reference electrical stimulus intensity is
presented. Subject is asked whether or not phosphene is perceptible. The current level is
adjusted up or down according to the “method of limits” to quickly determine a rough
threshold. This threshold is then used in the second step to generate a number of random
short stimulus trains. The subject is asked to count the number of pulses as the stimulus
train is presented (according to method of constant stimuli). The number of correct
answers less than and greater than 50% are judged as subthreshold and suprathreshold,
52
respectively. Fine adjustment of current amplitude is made and the new threshold is then
evaluated. Threshold is determined as the stimulus parameter in which the subject
answers correctly 50% of the times [106].
4.3 Experimental Setup for TcES
4.3.1 Goal of TcES
The goal is to use positron emission tomography to evaluate visual cortical activation in
normal sighted controls and patients with retinal degeneration (RD) following
transcorneal electrical stimulation (TcES). This novel study will serve as a precursor for
evaluating the functionality of retinal prosthesis in implant subjects. We want to study
visual cortex activation during electrical retinal stimulation and compare the pattern to
visually evoked cortical activation using light stimulus. In addition, cross-comparison
between normal control and RD group allows us to investigate potential changes in brain
activation pattern as a result of visual impairment from retinal degenerative diseases.
4.3.2 DTL-Plus Electrode
DTL fiber electrodes (Dawson, Trick, and Litzkow electrode) is a very low-mass, silver-
impregnated microfiber corneal electrode for clinical electroretinography (ERG) [43].
The DTL electrode is based on an extremely low-mass conductive thread that makes
contact with the tear film of the eye and is electrically coupled to an insulated wire. The
individual fibers of the nylon thread are approximately 50µm in diameter and are
53
impregnated with metallic silver. The thread is draped in the lower fornix, touching the
inferior limbus of the eye (Figure 13).
The advantages of the DTL electrode lie in the area of subject comfort, optical
quality, and reduced electrode impedance. The DTL electrode is well tolerated by
children and adults alike, and it is not blinked out of the eye. Recently, the DTL electrode
has been modified, and it is now commercially available (DTL-Plus, Diagnosys LLC,
Littleton, Massachusetts) as illustrated in Figure 13. The DTL-Plus electrode is composed
of 7-cm-long, low-mass spun nylon fibers impregnated with metallic silver. Small sponge
pads at each end with adhesive backing are secured to the nasal and tempral canthi,
securing the electrode and positioning the microconductive thread at the limbus. In
addition to recording ERG, the DTL electrode can also be used to deliver current to the
retina via transcorneal electrical stimulation [84] by connecting it to a neurostimulator.
Figure 13: DTL-Plus electrode. (a) Photograph of DTL-Plus electrode (courtesy of Diagnosys
LLC) (b) DTL-Plus electrode placement on the right eye; the two white adhesive pads are
attached to the medial and lateral canthi of the eye, and the fiber electrode is draped over the
lower eyelid and touching the inferior limbus.
54
Before electrode placement, two drops of topical anesthesia Tetracaine HCl 0.5%
(TetraVisc, OCuSOFT, Inc) is instilled in the patient’s right eye for rapid, sustained
anesthesia. One minute later, subject is asked to close his/her right eye while the smaller
adhesive pad of the electrode is affixed onto the side of the nose adjacent the medial
canthus. Subject is then asked to look up while the electrode fibers are draped over the
eye just above the lower eye lid. Subject is then asked to blink a once to seat the fiber
just on top of the lower lid in contact with the eye. The larger adhesive pad is then affixed
approximately 10 mm from the lateral canthus. The electrode lead is then connected to
the extension lead of the output cable of the stimulator.
Figure 14: ERG-Jet electrode. (a) Photograph of ERG-Jet electrode (courtesy of Fabrinal SA);
(b) position of the ERG-Jet electrode on the right eye.
55
4.3.3 ERG-Jet Electrode
ERG-Jet (Fabrinal SA, Switzerland) is a disposable mono-polar hard contact lens
electrode that has been FDA approved for human use. The electrode contains a thin
circular gold foil overlaid on top of the concave surface of the poly(methylmethacrylate)
contact lens (Figure 14a). The circular geometry of the gold contact makes this electrode
ideal for delivering small electrical current to more extensive regions of the retina,
potentially allowing more extensive activation of the visual cortex.
The size of the ERG-Jet electrode is suitable for most eyes and tolerated well with
proper topical anesthesia. Shortly after anesthetizing the cornea with 2 drops of
Tetracaine HCl 0.5% topical ocular anesthetic in the subject’s right eye, the concave cup
of the ERG-Jet electrode, filled with Lidocaine Ophthalmic gel 3.5% (Akten, Akorn,
Inc), is placed on the subject’s right eye as shown in Figure 14b. To maintain the ERG-
Jet electrode position on the cornea, an adhesive tape is used to fix the lead wire (wire
that delivers electrical current to the ERG-Jet) to the cheek just inferior to the lower
eyelid.
4.3.4 Neurostimulator
An FDA approved clinical grade DS7A series neurostimulator (Digitimer, LTD) is used
for transcorneal electrical stimulation (Figure 15). The stimulator is connected to either a
DTL-Plus microfiber electrode or an ERG-Jet contact lens electrode via a D185-HB4
output cable (Digitimer, LTD). Stimuli pulse trains are generated using DG2A Trigger
generator (Figure 16) which connects to the neurostimulator for continuous, repetitive
56
electrical stimulation (Figure 17). A second return electrode (silver silver/chloride, pre-
gelled, adhesive-backed electrode, Integra NeuroSupplies), is affixed on the temporal
skin of the right side of the subject’s face, and it is connected to the reference output of
the neurostimulator via a button snap connector cable (24” Button Snap, Integra
NeuroSupplies).
The Digitimer stimulator model DS7A provides constant current pulses of brief
duration for percutaneous nerve stimulation. The output current is continuously variable
over the range between 0 and 100 milliamps, from a source voltage continuously variable
from 100 Volts to 400 Volts, to meet the human requirements for both normal and
pathological cases. The pulse duration can be varied from 50 microseconds to 2
milliseconds in six steps and a specially designed isolated output stage maintains a square
current pulse shape while minimizing stimulus artifacts.
The DG2A is a compact, battery powered trigger generator for repetitive
stimulation. Various modes allow output pulses to be produced singularly, continuously
or in a burst, with the burst/train duration and pulse frequency determined by the front
panel controls. The DG2A is connected to the DS7A neurostimulator to generate a
continuous train of pulses during the TcES experiment.
Figure 18 demonstrates a sample trace of trigger pulse train from DG2A trigger
generator along with the output stimulus pulse train from DS7A neurostimulator. For
every output trigger pulse from the trigger generator, the neurostimulator outputs a
constant current pulse of 2 msec pulse duration. The output from the neurostimulator
57
alternates between cathodic and anodic pulse at 2 Hz repetition rate. The same pulse train
is delivered to the corneal electrodes (DTL-Plus or ERG-Jet) continuously during TcES.
Figure 15: Digitimer DS7A neurostimulator.
Figure 16: Digitimer DG2A trigger generator.
58
Figure 17: Neurostimulator setup for TcES. Pulse generator (DG2A) sends a continuous train
of trigger pulses to the DS7A neurostimulator, which in turn outputs a stimulus train with user-
defined current amplitude and pulse width to the corneal electrodes.
Figure 18: Trigger and stimulus pulse trains. Above, trigger pulse train at 2 Hz; the amplitude
of each trigger pulse is 5 V. Bottom, stimulus pulse train measured as potential drop across a 1
kΩ resistor with the neurostimulator current amplitude set to 1 mA. The data were acquired using
Tektronix TDS 5034B oscilloscope.
59
4.3.5 Corneal Exam
Corneal examination is done before and after the TcES experiment using Fluorescein
strips (Ful-Glo, Fluorescein sodium ophthalmic strips, USP, 0.6 mg; manufactured by
Akorn). Mag-Lite flash light is used as the light source, and it is fitted with #47 Blue
Tricolor Gel filter (B&H Photo-Video-Pro-Audio) to mimic emission spectrum of the
cobalt blue filter (Figure 19). To minimize any stinging sensation from the fluorescein, 1-
2 drops of Tetracaine topical anesthetic is instill in the right eye just prior to the use of
fluorescein strips. Surface corneal images are captured using Olympus digital camera
(Olympus, CAMedia, model C-5050 Zoom, Olympus Optical Co., LTD). Digital images
are recorded at 5 megapixel resolution and enlarged to examine for any corneal abrasion.
Figure 19: Light source for fluorescein corneal exam.
60
Chapter 5: Modeling of Transcorneal Electrical Stimulation
5.1 Modeling Overview
The modeling work is in collaboration with Dr. Gianluca Lazzi and his graduate student
Carlos Cela from the Bioelectromagnetics Research Laboratory at North Carolina State
University. The goal is to simulate the flow of current from the corneal electrodes (DTL-
Plus & ERG-Jet) through the ocular tissue to the retina and to determine the regions of
the retina most likely to become activated as a result of current pulse injection. Results
from the modeling work can be compared to both the subjective description of phosphene
reported by subjects during transcorneal electrical stimulation and the brain activation
maps derived from functional PET imaging. It can serve as an objective method to
explain the pattern of brain activation seen during TcES.
A brief summary of the modeling method and procedures are provided in this
overview; more details can be found in the literature [60] and in Carlos Cela’s doctoral
thesis work. The modeling method is a 3-dimensional expansion of the 2-dimensional
multiresolution admittance method for low-frequency electromagnetic interactions [60].
This simulation method is suitable for quasi-static electromagnetic radiation problems
that arise in the field of bioelectromagnetics. It is attractive for biomedical simulations
because of its relative simplicity, which is retained even when applied to nontrivial
problems involving spatially varying electromagnetic sources, or oddly shaped
heterogeneous objects, such as the human eye along with its surrounding soft and hard
tissues.
61
Admittance method involves the following steps. First, the physical model
(human eye) is discretized into a 3-dimensional mesh of simple cells; to decrease the
computational burden of the method, a multiresolution (nonuniform) discretization is
used [60]. Then an admittance network is constructed using lumped admittance derived
from the material (tissue) properties of each cell. Admittance is essentially the dual of the
impedance method [5]. The electromagnetic stimulus present in the physical model is
realized as currents injected into nodes of the network. The circuit network is then
expressed as an admittance matrix, and the excitation is stated as a current vector. A
linear system is then formed and solved numerically for current. Finally, the
correspondence between admittance in the circuit network and cells in the discretized
model is used to transform the branch currents into values of voltages with magnitude
and direction in the physical model. All the other parameters (e.g. current density, electric
fields, etc) can be calculated during post-processing.
5.2 Modeling Eye and Corneal Electrodes
The human eye model is based on the Visible Human Project male 1 mm dataset,
interpolated to 0.25 mm, and manually modified to include finer structures of the eye.
Figure 20 and 21 show slices of the model anatomy from the three principal (sagittal,
axial, and coronal) orthogonal axes. The different colors in the slices represent different
tissue types. To create an accurate model of transcorneal electrical stimulation, a large
model dataset is chosen which includes not only the eye but also the neighboring soft and
62
hard tissues. This dataset is then discretized using the method described in section 5.1 to
form 3-dimensional mesh of multidimensional cells as input to the admittance matrix.
Figure 20: Sagittal and Axial slices through human eye model. Top image shows the
orientation of the two orthogonal slices through the eye along the midlines; bottom left, shows the
sagittal slice of the model anatomy; bottom right, shows the axial slice of the model anatomy. The
various colors in the anatomy represent the different tissues. A: anterior; N: nasal; P: posterior; T:
temporal.
Figure 22 shows the geometry for both DTL-Plus and ERG-Jet electrodes used in
the model. DTL-Plus electrode is a fiber electrode comprised of multiple strands of nylon
fiber impregnated with medical grade silver. The approximate diameter of the fiber
bundle is 200µm. To represent this electrode in the model, the smallest cell dimension is
used (0.25 mm = 250 µm). We assume approximately 1 cm length of the fiber is in direct
63
contact with the bulbar conjunctiva immediately below the lower limbus of the cornea.
Thus a linear curved array of cells (one-cell wide × 40 cells long) is used to represent the
fiber electrode on the eye, which has an electrical conductivity of 63.01×10
6
S·m
-1
.
Figure 21: Coronal slices through human eye model. Upper left, orientation of the coronal
slice through the eye; lower left, the locations of the mid and posterior coronal slices through the
eye; upper right, mid coronal slice; lower right, posterior coronal slice. The various colors in the
anatomy represent the different tissues and materials as listed in middle. N: nasal; T: temporal.
The ERG-Jet is a hard contact lens electrode. As shown in Figure 22, it is
comprised of a gold ring directly in contact with the cornea. The width of the ring is
approximately 1 mm and the diameter is 1 cm. To represent the ERG-Jet electrode in the
64
model, we simply created an annular array of cells (4 cells wide, 1 cm in diameter)
centered on the cornea and has an electrical conductivity of 45.20×10
6
S·m
-1
.
Figure 22: Modeling of corneal electrodes on the human eye. Upper left, position of DTL-Plus
electrode on the right eye; lower left, position of ERG-Jet electrode on the right eye; upper right,
geometry of the DTL fiber electrode used in the model; lower right, geometry of ERG-Jet contact
lens electrode used in the model. Thickness refers to the width of the silver fiber bundle in the
DTL-Plus model or the width of gold contact ring in the ERG-Jet model. Return electrodes are
attached to the skin on temporal side of the eye; the green lighting bolts in the two models (upper,
lower right) denote intervening tissues between the return electrode (fixed on the temporal skin
surface) and the eye, which are omitted from the diagram for simplicity.
65
The return electrode is an adhesive backed, pre-gelled, silver-silver/chloride
electrode with an impedance of approximate 800 Ω over the frequency range of the
stimulus pulse used in TcES. It is affixed to the skin on the temporal side of the eye as
shown in Figure 23. In the model, the return electrode is 1 cm in diameter and is
positioned 1.5 cm above and 2.5 cm posterior to the lateral canthus of the right eye. The
return electrode is modeled as the drain for current return during TcES.
To validate that the admittance method is admissible for modeling transcorneal
electrical stimulation, we performed the fast Fourier transform on a single current pulse
(Figure 24). The current pulse is 2 msec in duration with amplitude of 0.5 mA. Fourier
analysis shows that most of the frequency components (> 95%) lie below 2500 Hz. The
pulse qualifies for low-frequency electromagnetic wave under the modeling method.
Following the construction of the eye model along with the corneal electrodes, the
3-dimensional physical dataset is discretized and transformed into the admittance matrix
for simulation. The subsequent sections present the results of the simulation for both the
DTL-Plus and ERG-Jet electrodes. Simulation results are compared to the subjective
description of phosphene seen by human subjects, as well as the brain activation maps
derived from functional PET imaging following transcorneal electrical stimulation.
66
Figure 23: Position of return electrode on the temporal skin. (Adapted from the human face
profile in http://artofdon.com)
67
Figure 24: Stimulating pulse and its fast Fourier transform.
5.3 DTL-Plus Modeling Results
5.3.1 DTL-Plus Potential Distribution
The static modeling of DTL-Plus electrode showed an area of highest electric potential
located on the inferior, nasal aspect of the retina in the right eye. This observation is
demonstrated in both the sagittal and mid coronal views of the potential distribution
(Figure 25). The arrow in the mid coronal slice points to the location of maximal
68
potential on the retina. Because of the high conductivity of vitreous humor, the potential
drop across the eye from the anterior to the posterior is very small.
Figure 25: Electric potential distribution in the eye from DTL-Plus electrode stimulation.
The potential profiles match the anatomic slices of model anatomy shown in Figures 20 and 21;
position of DTL-Plus electrode is indicated in the sagittal plot (upper left); A-anterior; N-nasal; P-
posterior; T-temporal; scale bar indicates potential value in volt.
69
Figure 26: Potential distribution and activating function maps for DTL-Plus electrode.
Cylindrical projection map of retinal surface potential (top right) shows the potential at every
point on the human retina; vertical axis represents radially directed meridians as indicated by
white-arrow on a series of fundus images (top left) with the corresponding radial directions (a – e)
labled in the potential map; horizontal axis indicates position along each radial line from posterior
pole (back of eye) to near the ora serrata (front of eye); scalebar indicates potential value in volt.
Cylindrical projection map of activating function (bottom) shows the normalized activating
function value at every point on the retina; the map axes are identical to those of surface potential
projection map above. OD – oculus dexter or right eye; OFL – optic fiber layer.
70
5.3.2 DTL-Plus Activating Function
The activating function, a measure of neuronal activation from an external current source
(Appendix C), can be used to predict the region of the retina preferentially stimulated
during transcorneal electrical stimulation [199, 200]. It is calculated from the potential
vector along the direction of the retinal ganglion cell axons (aka optic fiber layer, OFL).
Before deriving the activating function, local averaging of the retinal surface potentials
was first conducted to ease the effect of cubic meshing in the potential distribution of the
model. This way, no information was discarded, but local anomalies were averaged out
with adjacent voxels and thereby reducing error contribution. Vertical stripe pattern in the
activating function maps was caused by a combination of the meshing algorithm and the
geometrical transformation to extract the retinal voltage map. To facilitate cross-
comparison between the two electrode models, the activating function map is made
dimensionless by normalizing it to the maximum activating function value in the map.
A region of highest positive activating function is found in the corresponding
inferior, nasal location of the peripheral retina where the potential is also highest (Figure
26). The retinal surface potential map (Figure 26, top) is a cylindrical projection of the 3-
dimensional retinal surface potential onto a 2-dimensional surface map (the same way a
3-dimensional earth surface is projected onto a 2-dimensional atlas in which meridians
are mapped to equally spaced vertical lines and circles of latitude (parallels) are mapped
to horizontal lines). According to the retinal surface potential map, highest potential is
located in the inferior, nasal aspect of the peripheral retina, and the most positive values
of activating function are also found in the corresponding region (Figure 26, bottom;
71
centered at abscicca - 85, ordinate - 20 ). An activation function hotspot near the superior
peripheral retina (abscissa - 72, ordinate – 62) is an artifact resulting from superior rectus
muscle tissue touching the choroid in the model and leading to abnormal conductance.
To better visualize their approximate locations on top of the retina, regions
containing highest positive activating function values (≥ 0.9 on normalized scale) in the
inferior, nasal retina are overlaid on a normal fundus photograph fused to a semi-
transparent retinal ganglion cell axon tracing (Figure 29, top). The positive loci are
clustered along a few meridians posterior to the ora serrata which indicate the general
area of activation. The inferior, nasal retina receives visual input from the upper right
visual field; consequently, the modeling result predicts phosphene sensation in the upper
right visual field of the right eye for normal-sighted controls. And based on the linear
activating function pattern, the sensation of light may have a streaking distribution along
the upper right visual field.
5.4 ERG-Jet Modeling Results
5.4.1 ERG-Jet Potential Distribution
The static modeling of ERG-Jet electrode demonstrates more of a circularly symmetric
potential distribution in the right eye, as shown in both the sagittal and mid coronal views
of the potential distribution (Figure 27). The highest potential is found off-centered
towards the inferior, nasal aspect of the retina. Comparing the potential distribution maps,
72
ERG-Jet model has a broader potential crest spanning the inferior, nasal and a small
portion of the superior peripheral retina compared to that of DTL-Plus (Figures 26 & 28).
Figure 27: Electric potential distribution in the eye from ERG-Jet electrode stimulation.
The potential profiles match the anatomic slices of model anatomy shown in Figures 20 and 21;
A-anterior; N-nasal; P-posterior; T-temporal; scale bar indicates potential value in volt.
73
Figure 28: Potential distribution and activating function maps for ERG-Jet electrode.
Cylindrical projection map of retinal surface potential (top right) shows the potential at every
point on the human retina; vertical axis represents radially directed meridians as indicated by
white-arrow on a series of fundus images (top left) with the corresponding radial directions (a – e)
labled in the potential map; horizontal axis indicates position along each radial line from posterior
pole (back of eye) to near the ora serrata (front of eye); scalebar indicates potential value in volt.
Cylindrical projection map of activating function (bottom) shows the normalized activating
function value at every point on the retina; the map axes are identical to those of surface potential
projection map above. OD – oculus dexter or right eye; OFL – optic fiber layer.
74
Figure 29: Activating function overlays on normal fundus photograph. The color fundus
photo is fused with a semi-transparent retinal ganglion cell axon tracing (adapted from Oyster
CW (1999) The Human Eye: Structure and Function. Sinauer Associates, Sunderland, MA) using
Adobe Photoshop. Piecewise activating function map containing the highest activating function
values ( ≥ 0.9 on normalized activating function) are overlaid on the fused fundus image; (top)
DTL-Plus activating function overlay; (bottom) ERG-Jet activating function overlay.
75
5.4.2 ERG-Jet Activating Function
Activating function map for ERG-Jet model (Figure 28, bottom) shows a less well
defined and more diffuse pattern of maximal values scattered in the superior and nasal
edge along the peripheral retina,. The same hotspot near the superior peripheral retina
(Figure 28, bottom; abscissa - 72, ordinate – 62) is an artifact resulting from abnormal
conductance at the site where the superior rectus muscle tissue touches the choroid in the
model anatomy. To facilitate visualization, regions containing highest positive activating
function values (≥ 0.9 on normalized scale) are overlaid on a normal fundus photograph
fused with a semi-transparent retinal ganglion cell axon tracing (Figure 29, bottom).
Based on the shape of the activating function overlay, preferential nerve fiber stimulation
is likely to occur in a semicircular arc on the nasal hemiretina extending between the
superior and inferior retinal poles. This prediction translates to phosphene sensation
encompassing most of the right peripheral visual field in normal controls.
5.5 Modeling Summary
Results from admittance modeling predict preferential activation of the inferior, nasal
peripheral retina in the right eye during TcES using DTL-Plus electrode, and more
extensive peripheral, nasal hemiretina activation with the ERG-Jet. Higher electrical
conductivity of silver and smaller surface area of the fiber electrode result in slightly
higher potential distribution in the DTL-Plus model compared to ERG-Jet; and this may
76
translate to slightly lower threshold phosphene current needed to achieve phosphene
percept using DTL-Plus than ERG-Jet electrode in normal controls.
77
Chapter 6: PET Imaging and Analysis Methods
6.1 Overview of PET Functional Brain Imaging
6.1.1 Positron Emission Tomography Overview
Positron emission tomography (PET) is a nuclear medicine imaging technique which
produces a three-dimensional image of the functional processes in the body [232, 243].
The system detects pairs of high-energy photons emitted by a positron-emitting
radioisotope, which is attached onto a metabolically active molecule and introduced into
the body. When the radioisotope decays, a proton is broken down into two particles: (1) a
neutron, which remains within the nucleus, and (2) a positron, an unstable particle that
travels away from the nucleus a short distance before colliding with an electron. This
collision leads to their mutual annihilation and the emission of two gamma rays (511
keV) at 180° from one another. The site where the positron is annihilated is the site
detected by the scanner. The distance between the site of annihilation and the emitting
nucleus, which can be several millimeters, limits the spatial resolution of PET.
PET neuroimaging scanners contain a ring of gamma ray detectors (scintillation
crystals coupled to photomultiplier tubes) encircling the subject’s head. The two gamma
rays emitted by the annihilation of a positron and electron ultimately reach pairs of
coincidence detectors (Figure 30) that will record an event when two simultaneous
detections are made. If two photons are detected within a short (~10 ns) timing window
(the coincidence timing window), an event is recorded along the line connecting the two
78
detectors (sometimes referred to as a line-of-response). By summing many such events
results in quantities that approximate line integrals through the radioisotope distribution.
The validity of this approximation depends, of course, on the number of counts collected.
For two dimensional imaging, these line integrals form a discrete approximation of the
Radon transform [44] of a cross-section of the radioisotope concentration, and can be
inverted to form an image of the radioisotope distribution.
Figure 30: Scheme of PET acquisition process. (Courtesy of Langner [138])
79
6.1.2 Radioisotopes
Radionuclides used in PET scanning are typically isotopes with short half lives such as
11
C (~20 min),
13
N (~10 min),
15
O (~2 min), and
18
F (~110 min). These radionuclides are
incorporated either into compounds normally used by the body such as glucose, water or
ammonia, or into molecules that bind to receptors or other sites of drug action. Such
labeled compounds are known as radiotracers. Some tracers distribute in tissues by
partially following the metabolic pathways of their natural analogs; others bind with
specificity in the tissues containing the particular receptor proteins for which they have
affinity. It is possible to use PET technology to trace the biologic pathway of any
compound in living humans (and other species), provided the compound can be
radiolabeled with a PET isotope.
6.1.3 Image Reconstruction
The raw data collected by a PET scanner are a list of “coincidence events” representing
near-simultaneous detection of annihilation photons by a pair of detectors. Each
coincidence event represents a line in space connecting the two detectors along which the
positron emission occurred.
Coincidence events can be grouped into projection images, called sinograms. The
sinograms are sorted by the angle of each view and tilt (3D images). The sinogram
images are analogous to the projections captured by computed tomography (CT) scanner,
and can be reconstructed in a similar way. However, the statistics of the data is much
80
worse than those obtained through transmission tomography. A normal PET data set has
millions of counts for the whole acquisition, while CT can reach a few billion counts. As
such, PET data suffer from scatter and random events much more dramatically than do
CT data.
Filtered back projection is used to reconstruct images from the projections. This
algorithm has the advantage of being simple while having a low requirement for
computing resources. However, shot noise in the raw data can streak across the image.
Iterative expectation-maximization algorithms have the advantage of better noise profile
and resistance to the streak artifacts, but the disadvantage is higher computer resource
requirements.
As different lines-of-response must traverse different thicknesses of tissue, the
photons are attenuated differentially. The result is that structures deep in the body are
reconstructed as having falsely low tracer uptake. Attenuation correction needs to be
made during image reconstruction.
6.1.4 Fluorodeoxyglucose (FDG)
Fluorodeoxyglucose is a glucose analog. Its full IUPAC chemical name is 2-fluoro-2-
deoxy-D-glucose, commonly abbreviated as FDG or
18
FDG. FDG is most commonly
used in positron emission tomography (PET). The fluorine in the FDG molecule is
chosen to be the positron-emitting radioactive isotope fluorine-18, with a half-life of
109.8 minutes (Figure 31).
81
Figure 31: Chemical structure of FDG. (Courtesy of Wikipedia)
Tatsuo Ido [252] at the Brookhaven National Laboratory was the first to describe the
synthesis of FDG in the 1970s. The compound was first administered to two normal human
volunteers by Abass Alavi [1] in August, 1976 at the University of Pennsylvania. Images
obtained with an ordinary (non-PET) nuclear scanner demonstrated the concentration of FDG
in the brain.
FDG, as a glucose analog, is taken up by glucose-using cells in the brain, kidney
and cancer cells. FDG is prevented from further metabolism following the first
phosphorylation reaction because of the fluorine at the 2’-position in the FDG molecule.
A 2’-oxygen in glucose is needed for further glycolysis, therefore the FDG-6-phosphate
formed does not undergo glycolysis before radioactive decay and is trapped inside the
cell. As a result, the distribution of FDG is a good reflection of the distribution of glucose
uptake and phosphoryation by cells in the body.
To maximize binding of FDG to cell-surface receptors and minimize competition
with normal glucose in the body, patients must have suitably low blood sugar level and
are required to fast for at least 6 hours prior to the PET study. After FDG injection,
physical activity is kept to a minimum, in order to minimize uptake of radioactive sugar
82
in the muscles. The PET scan is composed of a series of one or more scans which may
take from 20 minutes to as long as one hour.
To avoid the destruction of organic molecules like deoxyglucose due to high
energy particle bombardment conditions in the medical cyclotron, the radioactive
18
F
must be made first as fluoride in the cyclotron. This usually is done by proton
bombardment of
18
O-enriched water, causing a proton-neutron reaction (neutron
knockout) in the
18
O to produce
18
F as labeled hydrofluoric acid, HF. The quickly-
decaying
18
F is then collected and immediately attached to the deoxyglucose in an
automated series of chemical reactions performed in a radioisotope chemistry preparation
chamber [27]. Following this, the labeled FDG compound is rapidly shipped to points of
use by the fastest possible mode to maximize radioactivity during application. Recently,
on-site cyclotron with integral shielding and portable chemistry stations for making FDG
makes it possible to perform routine metabolic functional imaging studies in even remote
hospitals.
6.1.5 Functional Brain Imaging Using FDG
For imaging glucose metabolism in the brain, the isotope, FDG is used. FDG accumulates
within the cell because it is phosphorylated by hexokinase but is not metabolized further,
and the amount accumulated reflects the rate of glucose metabolism. Louise Sokoloff
[231] first showed that local glucose consumption, measured by radioactive
deoxyglucose accumulation, is a reliable index of local neuronal activity. Most of the
energy derived from glucose is used to reestablish ionic gradients across the membranes
83
of neurons that have fired (through the Na
+
-K
+
ATPase). It is important to know that
energy is consumed by the activity of both excitatory and inhibitory synapses, and this
should not be confused with activation and deactivation of cerebral energy consumption.
6.1.6 Quantitative PET
If suitably calibrated, PET images can yield quantitative estimates of the concentration of
the radiopharmaceutical at specific locations within the body. The kinetics of the
radiopharmaceutical can be modeled as a linear dynamic system with the arterial
concentration of radioisotope in the blood as the input and the PET measurement as the
output. The state variables are the concentrations in different tissue compartments, where
the compartments can be vascular space, interstitial space between cells, and intracellular
space. The exchange rates between the compartments are parameters of the model.
Acquiring sequential images after radiopharmaceutical injection yields a time-course of
the output of the model, which can then be used to estimate the model’s parameters.
These estimated parameters can be used to calculate physiological quantities of interest,
such as glucose metabolism, in subsequent studies. Thus, PET can be adapted for precise
quantitative measurements of specific physiological quantities.
6.1.7 Metabolic Rate of Glucose (MRGlc)
We have used Sokoloff deoxyglucose method [230] to calculate the metabolic rate of
glucose (MRGlc) for all the PET functional images. Glucose supplies approximately 95%
84
to 99% of the brain’s energy under normal physiological state, and the rate of glucose
utilization is an excellent indicator of energy-requiring functions of the brain.
FDG is similar to glucose in several respects. Like glucose, it is transported from
the blood to the brain by a carrier-mediated diffusion mechanism. Hexokinase catalyzes
the phosphorylation of glucose to glucose-6-PO
4
and FDG to FDG-6-PO
4
. In both the
transport and phosphorylation steps, FDG is a competitive substrate with glucose. FDG-
6-PO
4
, however, is not a significant substrate for further metabolism. In the brain, it is not
converted into glycogen to any significant extent and is not further metabolized in the
glycolytic pathway. FDG-6-PO
4
also does not diffuse across cell membranes and is
therefore metabolically trapped in tissues, which is convenient both from an imaging and
modeling viewpoint.
Figure 32: Three-compartment FDG model. The model contains four first-order rate constants
describing transport between the compartments. C
p
, C
t
, and C
m
are the concentrations of glucose
in plasma, tissue, and metabolized glucose (glucose-6-PO
4
) in tissue, respectively. C
p
*, C
t
*, and
C
m
* are the corresponding concentrations for FDG.
85
FDG kinetics can be modeled with three compartments as shown in Figure 32. By
applying Michaelis-Menten equation of enzyme kinetics and using steady-state
compartmental modeling, we can produce an equation that gives us MRGlc as a function
of steady-state plasma glucose concentration, the lumped constant (LC in 6.1), which is
the ratio of net extraction of FDG to that of glucose, and the three constants ( k
1
*, k
2
* and
k
3
*). In practice, the Sokoloff operational equation (6.1) of the deoxyglucose model will
be used for the calculation.
T T
t k k T k k
T
t k k T k k
dt e e dt LC
dt e e k
MRGlc
0 0
) ( ) (
0
) ( ) ( *
1
*
3
*
2
*
3
*
2
*
3
*
2
*
3
*
2
) ( (
p
*
p
p
*
p
*
p
*
i
C
t C
C
t) C
t C T C
(6.1)
Where C
i
*(T) is the total
18
F tissue concentration (both FDG and FDG-6-PO
4
) at time T,
and C
p
*(t) is the plasma concentration of FDG (taken from serial blood samples), and C
p
is the plasma concentration of glucose. These values are measured during the experiment,
and the three rate constants and LC are obtained from population estimates [116, 190]. A
complete derivation of the three-compartment FDG metabolic model can be found in
Appendix A.
86
6.2 Methods of Analysis
Reconstructed PET images containing values of photon-counts per voxel are first
converted to metabolic images that have units of metabolic rate of glucose (MRGlc),
µmol of glucose consumed per minute per gram of tissue. Analysis is then conducted
using Statistical Parametric Mapping software (Wellcome Department of Imaging
Neuroscience, University College London, UK) to look for changes in brain activity that
is statistically significant. Quantitative imaging software PMOD (PMOD Technologies
Ltd., Zurich, Switzerland) is used to complement statistical analysis to acquire useful
quantitative data for in-depth study and comparison.
6.3 Statistical Parametric Mapping
6.3.1 Overview of SPM
Statistical parametric mapping (abbreviated SPM) is a statistical method for examining
differences in brain activity recorded during functional neuroimaging experiments using
either positron emission tomography (PET) or functional magnetic resonance imaging
(fMRI) techniques. MATLAB based software has been created by Wellcome Department
of Imaging Neuroscience, University College London to carry out SPM analysis.
SPM entails the construction of spatially extended statistical processes to test
hypotheses about regionally specific effects [75]. Statistical parametric maps are images
with voxel values that are, under the null hypothesis, distributed according to a known
87
probability density function (Student’s T or F distribution). In SPM, one analyzes each
and every voxel using a standard (univariate) statistical test. The resulting statistical
parameters are assembled into an image – SPM map.
Functional brain imaging experiments are commonly carried out to examine brain
activity linked to specific psychological process or processes. An experimental approach
to this problem might involve asking the question “which areas of the brain are
significantly more active when a person is doing task A compared to task B?” Although
each task will be designed to be identical, except for the aspect of behavior under
investigation, the brain is still likely to show changes in activity between tasks due to
factors other than task difference (as the brain is involved with coordinating a whole
range of parallel functions unrelated to the experimental task). Furthermore, the signal
may contain noise from the imaging process itself. To accommodate these random
effects, and to highlight the areas of activity linked specifically to the process under
investigation, statistics are used to look for the most significant difference above and
beyond background brain activity. This involves a multi-stage process to prepare the data,
and to subsequently analyze it using a statistical method known as the general linear
model. A schematic overview of the standard SPM analysis is shown in Figure 33.
88
Figure 33: SPM overview. This schematic depicts the transformations that start with an imaging
data sequence and end with a statistical parametric map (SPM). Voxel-based analyses require the
data to be in the same anatomical space: This is accomplished by realigning the data (and
removing movement-related signal components that persist after realignment). After realignment
the images are subject to nonlinear warping so that they match a template that already conforms
to a standard anatomical space. After smoothing, the general linear model is employed to (i)
estimate the parameters of the model and (ii) derive the appropriate univariate test statistic at
every voxel. The test statistics that ensue (usually T- or F-statistics) constitute the SPM. The final
stage is to make statistical inference on the basis of the SPM and Random Field theory. (Adapted
from Frackowiak [73])
6.3.2 Image Pre-processing
The analysis of neuroimaging data starts with a series of spatial transformations. These
transformations aim to reduce unwanted variance components in the voxel time-series
that are induced by movement or shape differences among a series of scans. Voxel-based
89
analyses assume that the data from a particular voxel all derive from the same part of the
brain.
The first step is to realign the functional images to ‘undo’ the effects of subject
movement during the scanning session. Despite restraints on head movement, cooperative
subjects still show displacements of up to several millimeters. Realignment involves (1)
estimating the 6 parameters of an affine “rigid-body” transformation to minimize the sum
of squared differences between each successive scan and a reference scan [77] and (2)
applying the transformation by re-sampling the data using tri-linear, sinc or spline
interpolation.
After realignment, the mean image of the series is used to estimate some warping
parameters that map it into the PET template that conforms to the ICBM 152 standard
anatomical space. The space is based on the Talairach system [239], but does not make
assumptions about brain symmetry, and also includes the cerebellum. The stereotactic
space is based on 152 normal MRI scans that have been matched to the MNI305
(Montréal Neurological Institute 305 Atlas) using a 9-parameter affine transform. The
International consortium for Brain Mapping (ICBM) adopted this as their standard
anatomical space, and it is used in SPM 5. The PET template used for normalization is
the average of PET images from 12 normal subjects spatially normalized to ICBM. These
images were acquired on a Siemens ECAT HR+ using oxygen-15 labeled water. They
were registered to the T1-weighted MR images, and spatially transformed using the same
transformation and smoothed using an 8 mm FWHM Gaussian kernel.
90
The final step of image preprocessing is spatial smoothing. According to the
central limit theorem, smoothing the data will render the errors more normal in their
distribution and ensure the validity of inferences based on parametric tests. In addition, in
the context of inter-subject averaging it is often necessary to smooth the images (e.g. 8
mm in fMRI; 16 mm in PET) to project the data onto a spatial scale where homologies in
functional anatomy are expressed among subject.
6.3.3 General Linear Model
The general linear model can be written succinctly using matrix notation “Y = Xβ + ɛ”
that expresses the observed response variable Y in terms of a linear combination of
explanatory variable X plus an error term, ɛ [78]. Thus with J number of observations,
and L number of effect parameters, the model can be written as a set of simultaneous
equations in (6.2):
This has an equivalent matrix form (6.3):
(6.2)
91
The matrix X that contains the explanatory variables is called the design matrix. Each
column of the design matrix corresponds to some effect one has built into the experiment
or that may confound the results. The response variable Y contains the measured data for
each observation. In SPM, the response variable represents all the values from an
individual voxel for each of the scans in the analysis. ‘ ɛ’ represents the residual error. ‘ β’
contains the effect parameters that we want to estimate from the set of linear equations.
Once an experiment has been completed, we have a number of observations
(voxel values contained in Y). Because the number of effect parameters (β) is typically
chosen to be less than the number of observations (Y), the general linear model cannot be
exactly solved. In order to estimate the parameters that best fit the data required, the least
square method is used. The least square method provides a set of optimal effect
parameters that minimizes the sum-of-square of residual errors. If (X
T
X) is invertible,
which it is if the design matrix X is of full rank, then the least square estimates can be
calculated as in (6.4):
(6.4)
(6.3)
92
Appendix B contains a detailed mathematical derivation for the General Linear
Model parameter estimation.
6.3.4 Subtraction Methodology
Conventional neuroimaging experiments often use the subtraction methodology, which
involves the following logical steps. The experimenter defines a pair of conditions that
are believed to differ with respect to a single cognitive or perceptual process. Statistical
comparisons of the activity are made between the two conditions, and the neural basis for
the putative cognitive process is assigned to the region or regions with significant
differences in activity. Conventional subtraction method is useful for identifying
increases or decreases in brain activity between two conditions. It is most appealing when
used with PET measurements because the relatively low signal-to-noise ratio produces
fewer statistically reliable differences. This is advantageous because the results appear
most compelling when there is a restricted center of the brain that become activated in
response to the putative cognitive process.
6.3.5 Absolute and Relative Subtraction Analyses
To analyze differences between any two conditions performed in subjects within the
same group, a paired t-test SPM design can be performed. Two different types of
analyses can be conducted: absolute subtraction and relative subtraction analyses. For
absolute subtraction analysis, normalized metabolic PET images are directly subtracted
93
between conditions to yield the SPM maps. This method is only valid for quantitative
PET imaging study in which each voxel of the PET image has units corresponding to
metabolic rate of glucose.
In relative subtraction analysis, the global mean of each PET image is divided out
from the image itself before the condition subtraction. Global mean is the global average
of image intensities of intracerebral tissue, which is also known as global activity. In
neuroimaging, one needs to differentiate between regional and global activity. Regional
activity means the activity measured in a single voxel or a small volume of voxels.
Global activity refers to a global measure of brain activity. The concept of global activity
is important because there are effects in a single voxel that are caused by global effects
[74]. One way to account for global changes is to adjust the data by scaling each scan by
its estimated global activity. This adjustment reduces global nuisance and enhances
weaker differences that are otherwise masked by global variation. By the same token, it
can also reduce activity in brain regions that otherwise would be statistically significant.
Thus both absolute and relative subtraction analyses complement each other and
provide a comprehensive SPM study between any two conditions.
To analyze differences between conditions across two different subject groups, a
two-sample t-test SPM design is carried out. Like the paired-t test design, it also offers
both absolute and relative analyses as described above.
94
6.3.6 Statistical Comparison
Parametric statistical models are assumed at each voxel, using the General Linear Model
to describe the variability in the data in terms of experimental and confounding effects,
and residual variability.
Statistical comparison in functional imaging is more complicated because there
are many voxels, and hence many statistic values. This gives rise to the problem of
multiple comparisons. With an anatomically open hypothesis (i.e. a null hypothesis that
there is no effect anywhere in a specified volume of the brain) a correction for multiple
dependent comparisons is necessary. The Gaussian random field theory provides a way
of adjusting the p-value that takes into account the fact that neighboring voxels are not
independent by virtue of continuity in the original data [80]. Over the years, SPM has
come to refer to the conjoint use of the general linear model and Gaussian random field
theory to analyze and make classical inferences about functional imaging data [76, 269].
The general linear model is used to estimate some parameters that could explain the
spatially continuous data in exactly the same way as in conventional analysis of discrete
data. Gaussian random field theory is used to resolve the multiple comparison problems
that ensue when making inferences over a volume of the brain, and thus provides a
method for correcting p-values.
95
6.3.7 Uncorrected P-values and Regional Hypotheses
When making inferences about regional effects in SPM, one often has some idea about
where the activation should be. In this case, a correction for the entire search volume is
inappropriate. If the hypothesized region contain activated voxels, then inference can be
made using an uncorrected p-value. More stringent uncorrected p-value thresholds, such
as p < 0.01, can be used to control for multiple comparison errors when there is a regional
hypothesis of where the activation will occur [80].
6.3.8 Graphical Representations
Differences in measured brain activity can be represented in a number of ways. In the
simplest form, they can be presented as a table, displaying coordinates that show the most
significant difference in activity between tasks. More often SPM maps are shown as
patches of color on an MRI brain template “slice”, with the colors representing voxels
that have shown statistically significant difference between conditions. The gradient of
color is mapped to statistical values, such as t-value or Z-score. This creates an intuitive
and visually appealing ways of delineating the relative statistical strength of a given area
of activation. Differences in activity can also be represented on a 3-D brain surface model
by superimposing pseudocolor SPM map on the model surface.
96
6.4 Quantitative Analysis using PMOD
Quantitative PET image analyses are performed using the commercial software PMOD
(PMOD Technologies Ltd., Zurich, Switzerland). PMOD is built on a series of user-
friendly and powerful image analysis toolboxes. The base model includes the following
functions:
Loading medical images in different formats, including DICOM;
Viewing the images with different color tables;
Calculating new slice images in arbitrary new orientations;
Performing various image processing and manipulation operations;
Displaying fusion images of matched data sets;
Performing volume-of-interest (VOI) analyses on template and user-defined
regions.
The software allows us to register each PET metabolic image to the reference PET
template using the ICBM 152 standard brain anatomy space. The spatial normalization
method is an elastic matching procedure which has been implemented according to the
methodology used in SPM 99 [79]. Each image is resliced to match the standard anatomy
using an initial scaling in all three dimensions, followed by some elastic transformation
for the fine adjustments. After image normalization, quantative image analyses can then
be carried out using SPM and Broadmann Area maps.
97
Chapter 7: Preliminary Studies
7.1 Epiretinal Prosthesis Implant Subject PET-H
2
O
15
Study
The first retinal prosthesis functionality pilot study was performed on May 21, 2004
using PET-H
2
O
15
on one of the first epiretinal implant subjects. The study was carried out
at the USC PET Molecular Imaging Laboratory. The subject, PB, is a 71 year old female
with an epiretinal prosthesis implanted in her left eye.
Figure 34: Brain activation during epiretinal prosthesis stimulation. SPM results on a single-
subject PET H
2
O
15
brain perfusion study during retinal stimulation. (A) Maximum intensity
projection maps showing a region of activation in the occipital cortex (pointed by the red arrow);
‘R’ and ‘L’ stands for right and left brain hemispheres, respectively; ‘A’ and ‘P’ stands for
anterior and posterior, respectively. (B) Going counterclockwise starting from the upper-right
image are the coronal, sagittal, and transverse planes through the brain showing the location of
activation corresponds to left occipital cortex.
98
Four dynamic PET scans were performed, each beginning simultaneously with a
rapid injection of O-15 water (41-44 mCi) and each consisting of 24 × 5 s frames. The
four scans were accomplished over a time period of about 1 hour. The retinal prosthesis
was either turned off (control scans, #1 and #3) or activated with a train of pulses (tests
scans, #2 and #4). Beginning shortly before injection and continuing throughout each
scan, the patient was given an audio/motor task (click right or left on a computer mouse
depending on whether the pitch of the sound was rising or falling in consecutive beeps)
that continued through the 2 min scan period. The patient was engaged in conversation
between scans. The dynamic image sets were reconstructed by the maximum a posteriori
projection (MAP) iterative statistical technique. Reconstructed images were summed over
the first 70 seconds following initial arrival of activity in the brain as determined by
plotting time-activity curves for regions of interest drawn around the outer boundary of
the brain as seen in the PET scans.
Statistical analysis, using Statistical Parametric Mapping software (SPM 2),
determined a single, highly significant region of activation (p=0.02, 0.32 cc volume) near
the posterior surface of the brain (Figure 34) in the approximate region of the visual
cortex. The region of activation was on the patient’s left occipital cortex relative to the
midline of the brain in a region identified as V1 (Brodmann Area 17) and possibly
extending up into Brodmann Areas 18 and 19 (V2 and V3, respectively). The reason
activation in the occipital cortex was only seen on the left hemisphere could be explained
by the location of the implant. Since the implant was on the temporal hemiretina of the
left eye, only an ipsilateral (left-sided) activation of the visual cortex would be expected.
99
7.2 Normal Control Light Stimulation PET Study
In the first phase of this project, six normal-sighted controls (4 males, 2 females, mean
age 38 ± 9.8, range: 24.2 - 51 years) without history of vision loss or neurologic diseases
were recruited to study cortical response of photic stimulation. Approval for the consent
and protocol for this study was given by the Committees on Research Involving Human
Subjects (CORIHS) at Stony Brook University, New York. The PET study was carried
out at Brookhaven National Laboratory, NY.
Pioneering studies by Phelps et al. [189] using positron computed tomography
and FDG has demonstrated activation of primary visual cortex during white-light
stimulation, and as the complexity of the visual stimulus increases, additional association
visual cortical areas were also recruited. Moreover, regional cerebral blood flow in the
primary visual cortex was shown to have a stimulus rate dependency [72]. Based on these
findings, an optimized flashing visual stimulus with multiple temporal frequencies was
used to elicit maximal occipital activity. The light stimulus, which subtended 10° field of
view on the retina, was a repeating sequence (30-second duration) of a flashing white
square with temporal frequency increasing incrementally from 2.5 to 30 Hz (2.5, 5, 10,
15 and 30 Hz). Two PET FDG scans were performed on separate days for each of the 6
normal subjects. During one scan, controls were blindfolded following FDG injection;
during the other scan, only the subject’s right eye was exposed to a flashing light stimulus
displayed from the computer monitor during the 30-minute FDG uptake phase. Each PET
scan (using CTI/Siemens Model 962 Scanner) was acquired for 20 minutes. Corrected
100
Figure 35: Brain activation during light stimulation in normal sighted subjects. Regions of
increased brain activation (p < 0.01) during light stimulation condition compared to baseline are
overlaid on top of registered MRI human brain template. The “hot” pseudocolor overlay
corresponds to t-values as indicated by the scale bar on the bottom. Top row of axial slices show
the results from absolute analysis, and bottom row the results from relative analysis. Six
contiguous axial slices encompassing the occipital cortex are displayed; the number at the top of
each slice represents slice location, with slice z = 0 corresponding to the axial plane joining the
anterior and posterior commissures. Positive z value represents slice above the commissural
plane, and negative z value represents slice below. L = Left; R = Right V1=primary visual area,
Brodmann Area 17; V2 = Brodmann Area 18; V3 = Brodmann Area 19.
metabolic images were analyzed using SPM 5 (Wellcome Department of Cognitive
Neurology, London, UK) to identify regions of cortical activation. Paired-t test was
conducted. Absolute subtraction analysis between light stimulation and baseline
condition demonstrated significant activation (p < 0.01) in the primary visual cortex (V1
or Brodmann Area 17) on both hemispheres (Figure 35); the activation was localized near
the posterior pole in the right primary visual cortex, and more anteriorly in the left
101
primary visual cortex. Activation was also seen in extrastriate visual areas V2 and V3
(corresponding to Brodmann Areas 18 and 19, respectively) on the left hemisphere.
Similar occipital activation pattern was seen using relative subtraction analysis
(Figure 35). Global normalization performed in relative analysis enhanced the signal in
the left anterior primary visual cortex.
The findings are consistent with monocular photic stimulation results reported by
various investigators (see Section 3.5). It is unlikely that the activation in the left anterior
primary visual cortex resulted from stimulation of peripheral nasal retina that
corresponded to the right temporal crescent in the visual field, because the light stimulus
subtended only the central 10° field of view. One plausible explanation that can account
for our finding is the nasotemporal asymmetry in ganglion cell projection to the primary
visual cortex (see Section 3.5.2).
7.3 Summary
Pilot study of retinal prosthesis functionality on a single implant subject demonstrated
activation in the primary visual cortex using PET H
2
O
15
imaging. Light stimulation in a
group of normal-sighted human subjects demonstrated bilateral primary visual cortex
activation consistent with findings from monocular photic stimulation studies. Activation
in visual areas V2, V3 on the left hemisphere and left anterior calcarine cortex may be
attributed to nasotemporal asymmetry.
102
Chapter 8: Research Design and Methods
8.1 Experimental Methods
8.1.1 Experiment Overview
We have recruited a total of eight normal control subjects and five subjects with retinal
degeneration according to the inclusion and exclusion criteria outlined in Appendix D.
For the normal controls, all eight have undergone both baseline and light stimulation
FDG PET studies, and six of those eight have undergone DTL-Plus TcES, ERG-Jet TcES
or both TcES studies. All five RD subjects have undergone baseline, light stimulation,
and ERG-Jet TcES studies. For both groups, each PET condition was performed on a
separate day, and the order of the study was randomized as described in Section 8.2.3.
Both the normal control and the RD group have undergone baseline and light
stimulation studies. For the baseline study, the subject was kept blindfolded after FDG
injection and throughout the entire 35-minute radiotracer uptake phase before the PET
scan. For the light stimulation study, the subject sat in front of a computer screen
surrounded by a cubicle like structure with black drop-cloth over the cubicle to prevent
exposure to normal ambient light in the PET control room. The subject’s left eye was
blind-folded while the right eye was exposed to a flashing light stimulus from a CRT
monitor during the FDG uptake phase. On two other days of the study, five subjects from
the normal control group underwent transcorneal electrical stimulation (TcES) using
either DTL-Plus or ERG-Jet corneal electrodes during uptake phase of radiotracer before
103
the PET scan. Subjects from the RD group have participated in only one TcES study
using ERG-Jet corneal electrode.
Baseline Study
Light Stimulation Study
DTL-Plus TcES Study
ERG-Jet TcES Study
Figure 36: PET experiment flow diagrams.
104
Immediately following each study condition, subjects were given a PET scan with
both eyes blindfolded throughout the duration of the scan. Each PET scan (using
CTI/Siemens Model 962 Scanner) was acquired for 20 minutes. Flow diagrams for the
four different study conditions (baseline, light stimulation, DTL-Plus TcES, and ERG-Jet
TcES) are shown in Figure 36, and a detailed description of each study is presented
below.
8.1.2 Baseline Study Protocol
Baseline study serves as a reference in subtraction analysis for both light stimulation and
transcorneal electrical stimulation conditions. Subjects were first dark-adapted with a
blindfold for 30 minutes prior to FDG injection. During the baseline study, subject’s both
eyes were blindfolded throughout the entire 35-minute radiotracer uptake phase.
8.1.3 Light Stimulation Study Protocol
Subjects were first dark-adapted with a blindfold for 30 minutes prior to FDG injection.
During light stimulation, the subject’s chin was comfortably placed on the chin rest and
the right eye was exposed to a flashing light stimulus from a CRT computer monitor
during the 30-minute FDG uptake phase. The light stimulus program was run from Dell
OptiPlex model GX1 266MHz desktop computer and displayed on a Viewsonic
OptiQuest Model Q71 17-inch CRT monitor. The photic stimulus, which subtended
central 10° field of view on the retina, is a repeating sequence (30-second duration) of a
105
Figure 37: Light stimulation setup. Subject places his/her chin on the chin rest while fixating on
the flashing light stimulus displayed on the CRT monitor; a high-definition video camera captures
pupillary light reflection to monitor for eye fixation.
flashing white square with temporal frequency increasing incrementally from 2.5 to 30
Hz (2.5, 5, 10, 15 and 30 Hz). The average photopic illuminance of the light stimulus was
measured with IL1700 Research Radiometer (International Light, Inc, Newburyport,
MA) coupled to a SED033 detector (International Light, Inc, Newburyport, MA).
Illuminance was measured by centering the detector at the same distance away from the
CRT monitor as the subject’s right eye during light stimulation and integrating the photon
energy from the flashing stimulus over a period of one minute. The integrated
illuminance was first subtracted from background illuminance and then divide by 60
seconds to yield lumens per square meter (lumen·m
-2
) or lux. By averaging ten
106
measurements, an illuminance value of 1.76 lumen·m
-2
was obtained for the flashing
stimulus. During the light stimulation, a high-definition video camera recorded the
pupilary light reflection from the subject’s right eye. The video served as a quality
control for eye fixation. Any abnormal eye movement during light stimulation was
recorded for future data interpretation. A picture of the setup is shown in Figure 37.
8.1.4 DTL-Plus TcES Study Protocol
Corneal examination was performed at the outset to establish any baseline corneal lesion
that may be present in the subject’s right eye on the day of the study. After instilling 2
drops of Tetracaine anesthetic, a DTL-Plus electrode was placed on the subject’s right
eye, and a return electrode was placed on the temporal skin on the right side of the
subject’s face (Figure 38). The DTL-Plus electrode was connected to the stimulator for a
quick phosphene threshold assessment. The 2 msec pulse duration was selected for
optimal stimulation of the ganglion cell axon nerve fibers in the retina while minimizing
unwanted muscle twitch in the lower eye lid, which was observed for shorter pulse
durations due to stimulation of facial nerve fibers [84].
After a crude threshold current was determined using the “method of ascending
limits”, the “method of constant stimuli” was conducted to fine-tune the threshold (see
Section 4.2.2). Using threshold estimate determined previously as a starting point, a train
of four pulses of the same current strength was sent to the corneal electrode and the
subject was asked to count only the visible phosphenes as the pulses were presented. The
current amplitude was adjusted until the subject counted the number of pulses correctly
107
two out of four times. This current amplitude was used as the threshold. Threshold
current amplitude only provides a very faint phosphene perception. In order to achieve
reasonable activation of the visual cortex, current level 1.5 × threshold current amplitude
was used for the actual stimulation experiment. This stimulation current was individually
determined for each subject, and did not exceed 10 mA. If the stimulation current caused
discomfort in the subject’s eye, lower current amplitude above the phosphene threshold
that did not cause discomfort was used. At the end of the phosphene threshold
assessment, the subject was presented a train of pulses (2 Hz, 2 msec pulse duration,
alternating cathodic and anodic) using the stimulation current and asked to describe the
shape of the phosphene and assign an intensity level (on a scale of 1-10, 10 being the
brightest light ever seen).
Once the stimulation current amplitude was determined, the subject was
blindfolded for 30 minutes with the corneal and the return electrodes still attached.
Immediately before the start of the transcorneal electrical stimulation, two more drops of
Tetracaine anesthetic was instilled in the right eye and FDG was injected. Electrical
stimulation was conducted in a dark room to simulate the condition of light stimulation
done in a previous experiment. During stimulation, human subjects sat quietly in a chair.
A continuous train of pulses (2 Hz, 2 msec pulse duration, alternating cathodic and
anodic current pulses with predetermined current amplitude) was delivered to the
electrode from the stimulator for duration of 30 minutes. To avoid fatigue and adaptation,
stimulus train was turned off for 30 seconds after every five minutes of stimulation to
allow the subject’s eye to rest.
108
Following transcorneal electrical stimulation, the DTL-Plus and return electrode
were removed from the subject. While both eyes still blind-folded, the subject lay quietly
in the PET scanner for a 20 minute scan to acquire the functional brain image.
Immediately following image acquisition, a second corneal exam was performed and
compared to the baseline digital corneal image taken before the experiment to see if any
abrasion was caused by the electrode. In case if corneal lesion was encountered, the
subject was referred to an ophthalmologist for further check up.
8.1.5 ERG-Jet TcES Study Protocol
Corneal examination was performed at the outset to establish any baseline corneal lesion
that may be present in the subject’s right eye on the day of the scan. Two drops of
Tetracaine anesthetic was first instilled in the subject’s right eye, and then the concave
cup of the ERG-Jet electrode was filled with lidocaine gel 3.5% before being placed on
the subject’s right eye. A return electrode was placed on the temporal skin on the right
side of the subject’s face (Figure 39). Both the corneal and the return electrodes were
connected to the neurostimulator for phosphene threshold assessment. The same
threshold determination method was used as described above in Section 8.1.5.
Once the stimulation current was determined, the ERG-Jet electrode was removed
from the subject’s right eye. The subject was then blind-folded for 25 minutes. Five
minutes before the start of TcES, lidocaine gel 3.5% was reapplied to the ERG-Jet
electrode and placed on the subject’s right eye with the aid of low-intensity red headlamp
(Petzl E49P TacTikka Plus 4-LED Headlamp). The subject then sat in a dark room, and
109
Figure 38: DTL-Plus electrode placed on a subject’s right eye. DTL-Plus fiber electrode is
touching the lower corneal limbus of the right eye, and the return electrode is affixed to the right
temporal skin.
Figure 39: ERG-Jet electrode placed on a subject’s right eye. ERG-Jet contact lens electrode
is placed over the cornea of the right eye, and the return electrode is affixed to the right temporal
skin.
110
TcES was started immediately after FDG injection. A continuous train of pulses (2 Hz, 2
msec pulse duration, alternating cathodic and anodic current pulses with predetermined
current amplitude) was delivered to the electrode from the stimulator for duration of 30
minutes. To avoid fatigue and adaptation, stimulation was turned off for 30 seconds after
every five minutes of stimulation to allow the subject’s eye to rest. Following TcES,
ERG-Jet and return electrodes were removed from the subject. With both eyes blind-
folded, the subject underwent the PET scan. Immediately following PET image
acquisition, a second corneal exam was performed and compared to the baseline digital
corneal image before the experiment to see if any abrasion was caused by the electrode.
111
8.1.6 Study Flow Chart
PROCEDURES
Personnel
Screen
Scan
1
Scan
2
Scan
3
Scan
4
Study informed consent obtained SP x
Blood lab and urinanalysis (Chem
Screen; CBC; TSH; Urinalysis)
SP x
Urine toxicology screen (stat) SP x x x x x
Pregnancy test (stat) SP x x x x x
Medical history SP x
Physical exam MD x
Drug history SP x x x x x
Review-inclusion/exclusion MD x x x x x
Brief physical exam MD x x x x
PET subject information SP x x x x
Head-holder SP/RN/PET x
Transmission scan RN/PET x x x x
Lidocaine/prilocaine topical MD/RN x x x x
Hand warming RN x x x x
Arterialized venous catheter
inserted
MD/RN x x x x
Venous catheter inserted RN x x x x
IV drip (0.9% Na Cl with heparin
0.5 units/ml)
RN x x x x
Baseline Study
(Normal Controls & RD)
MD x
Light Stimulation Study
(Normal Controls & RD)
MD x
DTL-Plus TcES Study
(Normal Controls Only)
MD/CP x
ERG-Jet TcES Study
(Normal Controls & RD)
MD/CP x
FDG injection for PET scan MD x x x x
Blood Sampling for
18
FDG; (12 x
10 sec; 2 x 1 min; 3 x 2 min; 4 x 5
min; 1 x 15 min; 1x10 min - total
volume about 17.5 ml)
RN x x x x
Catheters removed MD/RN x x x x
Follow-up GCRC x x x x
Table 2: Study flow chart. CP = Credential personnel for performing TcES; GCRC = General
Clinical Research Center staff; MD = Physician; PET = PET staff; RN = Registered Nurse; SP =
Screening Personnel.
112
8.2 Experimental Techniques
8.2.1 Screening Visit
For confirmation purposes, standard laboratory tests (e.g. Chem Screen, CBC, TSH, and
Urinalysis) have been performed for all subjects when they first arrived at BNL unless
previously done within the last six months. We have obtained lab certifications for copies
of lab tests not performed at BNL to be kept in the investigator file. Urine tests to
identify drugs of abuse (including cocaine, PCP, THC, benzodiazepines, amphetamine,
opiates, and barbiturates) are done on each day of the study. Subjects have their full
medical history obtained and physical examination performed. For the retinitis
pigmentosa patients recruited from Columbia University, we also have a copy of full
report stating the diagnosis and evaluation of their eye condition, as well as digital copies
of the most recent color fundus photographs and optical coherence tomography studies.
8.2.2 FDG PET Scanning Visits
The PET experiment took a total of three or four days to complete depending on whether
the subject underwent one or both TcES conditions. All subjects underwent only one PET
scan per day. Following the initial screening on the first day, subjects received a PET
scan provided that they met the inclusion and exclusion criteria (Section 8.1.1). The four
different studies were baseline, light stimulation and transcorneal electrical stimulation
(TcES) with either DTL-Plus or ERG-Jet electrode. The order of the studies was
randomized in accordance to the randomization procedure described in Section 8.2.3.
113
8.2.3 Randomization
The order of the PET studies was randomized according to the analysis of variance
(ANOVA) because it has been shown from previous FDG studies that metabolic activity
is slightly lower on subsequent visits than on the first visit for baseline scans (ordering
effects). This may be due to anxiety [261]. The anxiety is from the overall procedures
(e.g., study procedure, environment, PET scan, etc...) and not from the radiotracer itself.
We have randomized this study such that increased anxiety experienced on the first day
of the PET scan was spread across the subjects to help counterbalance these ordering
effects. For subjects who underwent three scans, the order was randomized in increments
of 6: (3! = 3 x 2 x 1 = 6). There were 6 possible sequences that could have occurred
between these days (ABC, ACB, BAC, BCA, CAB, CBA). For subjects who underwent
four scans, only the first three scans were randomized and the last scan condition was
performed on the last day of the study.
8.2.4 Arterialized Venous Catheterization
In preparation for the initial scans, two catheters were placed into the subject: one venous
for tracer injection and the other arterialized venous for measurement of total plasma
concentration of radioactivity (FDG). Samples were obtained before, during and at the
end of the study for each scan completed. An initial blood sample was obtained to
measure arterialized blood gases for FDG studies. If the subject’s oxygen percentage was
less than 60%, the subject’s hand was warmed for a longer period and retested until the
114
oxygen level reached 60% prior to isotope injection. Subjects had saline flowing through
each catheter while it is in the arm or wrist.
8.2.5 Radiotracer Synthesis
FDG is synthesized as described previously [98]. FDG is routinely synthesized and used
for human studies at BNL under IND #13,483 approval. However, in the case of
technical difficulties with the FDG production at BNL, FDG can be obtained from a
commercial supplier (Cardinal Health). The company performs quality control on each
batch before the product is released.
8.2.6 PET Image Reconstruction
Final brain reconstruction had a volume of 128×128×63 voxels with pixel size set to 1.72
mm × 1.72 mm (transaxially) and 2.43 mm (axially). The actual spatial resolution after
filtered reconstruction was between 6-7 mm FWHM. Projection data were reconstructed
using “filtered backprojection method” with a Hann filter kernal set to 4.9 mm FWHM.
We corrected the sonogram data by performing attenuation correction, delayed-
coincidence randoms subtraction, and single scatter simulation scatter correction. For
attenuation correction, transmission scan was done using three Ge-68 rods with 2-5 mCi
radioactivities each. The ratio of a blank scan (with no subject in the field of view (FOV))
to the transmission scan (subject inside the FOV) was multiplied to the sonogram data.
115
Chapter 9: Results and Analyses
9.1 Normal Control Subjects
A total of eight right-handed, normal-sighted human controls were recruited for the study.
Table 3 shows the age, gender and PET study conditions for the subjects. ID index in the
first column will be used to refer to normal subjects henceforth.
ID Index Age Gender Light DTL-Plus ERG-Jet
NC 1 47.5 F ● ● ●
NC 2 52.4 M ● ●
NC 3 27.0 M ● ●
NC 4 24.0 M ● ● ●
NC 5 41.8 F ● ● ●
NC 6 31.4 M ● ● ●
NC 7 40.2 M ●
NC 8 25.7 M ●
● Condition performed on the subject
Table 3: Normal controls identifying data.
9.1.1 Subjective Sensation of Stimulation
The central projection of the square photic stimulus duing light stimulation resulted in
perception of a white flashing square in the central 10° field of view of the right eye
(Figure 40a). For TcES, subjects were asked to draw on a visual field grid their subjective
phosphene perception. During TcES using DTL-Plus electrode, subjects reported seeing
phosphene in the upper temporal quadrant of the visual field in the right eye (Figure 40b).
The phosphene was faint and rapid, similar to seeing a lightning bolt from the corner of
116
one’s right eye. On a scale of 1 – 10 (‘10’ being the brightest light ever seen), the subjects
rated the phosphene intensity between 1 and 2. They generally agreed the phosphene was
a white burst of light, but can not describe any other hue. During TcES using ERG-Jet
electrode, all five subjects consistently reported seeing a semi-circular arc-shaped
phosphene in the peripheral temporal visual field of the right eye that extended between
the superior and inferior limits of the visual field (Figure 40c). The phosphene was
likened to the rapid closing of a camera shutter. On a scale of 1 – 10, the subjects rated
the phosphene intensity between 4 and 5. Phosphene color was perceived to be white; one
subject (NC 5) reported seeing white color with a tinge of blue. Control subjects who
underwent both TcES studies consistently reported brighter phosphene sensation using
ERG-Jet than DTL-Plus electrode. For both TcES studies, the alternating monophasic
current pulses resulted in perception of alternating phosphene intensity with the cathodic
pulse reported to be brighter than the anodic pulse.
9.1.2 Light Stimulation vs. Baseline
Statistical parametric analysis was conducted on all eight normal control subjects who
underwent both light stimulation and baseline conditions. Since TcES studies were
performed only on five subjects from the normal control pool, a separate statistical
parametric analysis was carried out on five subjects who participated in TcES. This
allowed for cross comparison between the various stimulation conditions.
Statistical parametric analysis was done using SPM 5 (Section 6.3) on corrected
metabolic PET images (Section 6.1). Both absolute and relative analyses were conducted
117
for each pair of conditions. Resultant SPM maps were superimposed on MRI-T1 axial
slices (ch2bet.nii; MRIcron [210]) as well as 3-dimensional MRI brain surfaces generated
using MRIcron.
Figure 40: Normal control subjective visual perception. (a) Light stimulation; (b) TcES with
DTL-Plus; and (c) TcES with ERG-Jet. Pink overlay on the visual field grid indicates location
and shape of visual perception reported by the normal subjects for each condition.
The results of SPM analysis for all eight normal controls (NC 1 through NC 8,
age: 36.3 ± 10.7 years) are shown in Figures 41–45, and the summary of the finding are
listed in Table 4. Figures 41 and 42 show brain areas that exhibited increased metabolic
activity during light stimulation in comparison to baseline at p-value threshold set to
118
p<0.05 and p<0.01 respectively. Figure 43 displays brain areas with decreased metabolic
activity during light stimulation in comparison to baseline at p < 0.05. Areas of cortical
activation during light stimulation are projected onto 3D rendered brain surfaces as
shown in Figure 44 and cortical deactivation projection is shown in Figure 45.
The resultant SPM analysis shows activation in primary visual cortex (Brodmann
Area 17, abbreviated as BA 17) bilaterally at the posterior occipital pole, with more
significant activation located on the right (axial slices z = 4 to -16, Figures 41-42). There
was also bilateral activation of secondary visual cortices (BA 18 and BA 19) on the left
hemisphere. The strong activation of left extrastriate visual cortex (BA 19) may be
related to visual attention as demonstrated by Kastner et al. [125] who found significantly
larger activation of V3 when the subject attended to the visual stimulus when compared
to looking but not attending to the visual stimulus. Furthermore, observations made using
fMRI indicated that in normal subjects the striate cortex was activated by light stimulus
projected to the contralateral visual field, whereas extrastriate cortical activation occurred
bilaterally, with stronger activation on the contralateral visual cortex [88, 173]. Our
finding of greater left extrastriate visual cortex activation during central light stimulation
of the right eye is consistent with these fMRI findings.
Furthermore, a significant activation locus in the anterior calcarine cortex (BA 17)
was found on the left hemisphere (axial slices z = 4 to 16, Figures 41-42). This
observation is consistent with findings from studies that looked at monocular photic
stimulation of the right eye in which the investigators attributed this phenomenon to nasal
temporal asymmetry of photoreceptor and ganglion cell distribution (Section 3.5).
119
In addition to occipital cortex activation, increase activity was also seen in
anterior prefrontal cortex (BA 10), orbitofrontal cortex (BA 11), dorsal lateral prefrontal
cortex (BA 9), and frontal eye fields (BA 8), with more significant activation on the left
hemisphere (Figure 44). These prefrontal cortical areas are involved in visual attention
and fixation during light stimulation tasks [2, 96, 244]. There was bilateral activation of
middle frontal gyrus (BA 46) which has been observed to be involved in visual attention
and motor planning [244]. There was also a significant activation locus in the inferior
temporal gyrus (BA 20) on the left hemisphere which has been implicated in visual
awareness and perception of shape and pattern [244]. Activation of right lateral
intraparietal cortex (BA 7) may be involved in visual attention [244] and visual spatial
integration [2]. A region of supramarginal gyrus (BA 40) activation on the left
hemisphere was observed. Supramarginal gyrus is involved in somatosensory association
and attention. A small, but significant activation was also found in the rostral part of
parahippocampal gyrus (BA 27) on the left hemisphere as indicated in slices z = - 8 to 8
in Figures 41 and 42. The parahippocampal gyrus has been implicated in visual memory
[244]. Left caudate nucleus was also activated, which has been implicated in selective
visual attention [35]. In summary, light stimulation of the right eye activated primary and
secondary visual cortex as well as cortical areas that subserve functions of visual
attention, eye fixation, pattern perception and visual memory.
Absolute subtraction did not yield brain areas that showed significant decrease in
brain activity during light stimulation in comparison to baseline. Relative subtraction
found a number of cortical regions that showed decreased metabolic activity primarily in
120
auditory association areas (BA 42 and 22) and somatosensory association areas (BA 7, 40
and 43) as shown in Figures 43 and 45. The observed deactivations can be attributed to
cross-modal suppression [104, 139, 221], the idea that when attention is fixated on one
sensory input (i.e. vision), cortical processing of other sensory modalities (i.e. auditory
and somatosensory inputs) are depressed.
A summary of brain regions activated and deactivated during light stimulation in
comparison to baseline is tabulated in Tables 4. The coordinates conform to the Talairach
space [239] and denote local maximum for each corresponding brain region. The ‘Side’
entry denotes either left (‘L’) or right (‘R’) brain hemisphere. Both T-value and the
corresponding uncorrected p-value are listed for each brain region.
Light stimulation data from a subgroup of normal controls (NC 1 through NC 5;
age: 38.5 ± 12.5 years) are also analyzed. Since the same subgroup of subjects also
underwent TcES (DTL-Plus) study, a direct cross-comparison between light stimulation
and TcES conditions can be made. The graphical results from SPM analysis are shown in
Figures 46 – 49, and summarized in Table 5.
The pattern of brain activation for this group of five normal subjects is very
similar to the eight-subject SPM analysis presented above. Brain activations were
observed in bilateral primary visual cortex (BA 17), with more significant activation in
the right occipital pole. Activation was also seen in bilateral secondary visual cortex (BA
18), left secondary visual cortex (BA 19), and left anterior calcarine cortex (BA 17). In
addition, activations were also seen in bilateral orbitofrontal cortex (BA 11), left dorsal
lateral prefrontal cortex (BA 9 and 46), left supramarginal gyrus (BA 40), left inferior
121
Figure 41: Eight normal controls light stimulation activation SPM maps (p < 0.05). Selected
Brodmann areas that exhibit activation during light stimulation are labeled in green.
122
Figure 42: Eight normal controls light stimulation activation SPM maps (p < 0.01). Selected
Brodmann areas that exhibit activation during light stimulation are labeled in green.
123
Figure 43: Eight normal controls light stimulation deactivation SPM maps (p < 0.05).
Selected Brodmann areas that exhibit deactivation during light stimulation are labeled in green.
124
Figure 44: Eight normal controls light stimulation activation 3D SPM projection. Selected
Brodmann areas that exhibit activation during light stimulation are labeled in yellow.
125
Figure 45: Eight normal controls light stimulation deactivation 3D SPM projection. Selected
Brodmann areas that exhibit deactivation during light stimulation are labeled in yellow.
temporal gyrus (BA 20), and left rostral parahippocampal gyrus (BA 27). These activated
brain areas all have statistical significance better than p < 0.05, but they did not reach as
high a significance level as those seen in the eight subject analysis by virtue of a smaller
sample size.
Absolute subtraction did not find any brain region that exhibited statistically
significant decrease in activity (Figure 47 and 49). Relative subtraction found decreased
activity in bilateral primary and association auditory cortex (Heschl’s gyrus, BA 42), in
left somatosensory association cortex (BA 7), and in right supramarginal gyrus (BA 40).
126
Moreover, deactivation was observed in the inferior temporal gyrus (BA 20) near the
anterior temporal pole, which plays a role in multisensory association and emotion.
Decreased activity was also observed in supplementary motor cortex (BA 6), pars
triangularis (BA 45), fusiform gyrus (BA 37), and temporopolar area (BA 38).
In summary, light stimulation in the five normal controls activated primary and
secondary visual cortices as well as brain areas that subserve functions of visual attention,
fixation, pattern perception and visual memory. Deactivation seen in auditory and
somatosensory association cortical areas can be ascribed to cross-modal suppression.
127
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
Light Stimulation > Baseline (Absolute Subtraction)
Angular gyrus 39 L -40 -53 22 3.30 0.0060
Calcarine cortex 17 L -18 -74 10 4.67 0.0011
17 L -16 -100 -10 4.75 0.0010
17 R 18 -102 -2 5.06 0.0007
Caudate nucleus L -16 20 8 3.77 0.0036
Dorsolateral prefrontal
cortex
9 L -30 36 50 4.94 0.0008
Inferior temporal gyrus 20 L -52 -32 -26 6.30 0.0002
Middle frontal gyrus 46 L -32 46 40 4.77 0.0010
46 R 48 40 20 3.57 0.0045
Middle occipital gyrus 19 L -48 -80 6 3.96 0.0028
Orbitofrontal cortex 11 L -6 65 -26 7.82 0.0001
11 R 14 48 -22 4.65 0.0012
Parahippocampal gyrus 27 L -16 -38 -2 4.12 0.0022
Superior frontal gyrus 10 L -8 66 -12 4.63 0.0012
Supramarginal gyrus 40 L -60 -50 38 3.10 0.0086
Light Stimulation > Baseline (Relative Subtraction)
Calcarine cortex 17 L -14 -74 8 6.06 0.0003
17 R 20 -102 -6 4.31 0.0018
Middle occipital gyrus 19 L -48 -72 2 2.72 0.0150
Orbitofrontal cortex 11 L -6 54 -26 4.24 0.0019
Baseline > Light Stimulation (Absolute Subtraction)
None
Baseline > Light Stimulation (Relative Subtraction)
Heschl’s gyrus 42 R 46 -14 8 4.63 0.0012
Subcentral area 43 R 56 -8 26 4.60 0.0012
Superior parietal gyrus 7 L -28 -76 46 2.76 0.0140
7 R 8 -72 58 2.87 0.0120
Superior temporal gyrus 22 R 52 -6 -4 5.19 0.0006
R 46 -14 8 5.57 0.0004
Supramarginal gyrus 40 R 52 -34 34 4.76 0.0010
Table 4: Eight normal controls light stimulation vs. baseline SPM summary.
128
Figure 46: Five normal controls light stimulation activation SPM maps. Selected Brodmann
areas that exhibit activation during light stimulation are labeled in green.
129
Figure 47: Five normal controls light stimulation deactivation SPM maps. Selected
Brodmann areas that exhibit deactivation during light stimulation are labeled in green.
130
Figure 48: Five normal controls light stimulation activation 3D SPM projection. Selected
Brodmann areas that exhibit activation during light stimulation are labeled in yellow.
Figure 49: Five normal controls light stimulation deactivation 3D SPM projection. Selected
Brodmann areas that exhibit deactivation during light stimulation are labeled in yellow.
131
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
Light Stimulation > Baseline (Absolute Subtraction)
Angular gyrus 39 L -40 -50 26 2.59 0.0300
Calcarine cortex 17 L -18 -72 8 4.21 0.0068
17 R 18 -102 -4 3.63 0.0110
Dorsolateral prefrontal
cortex
9 L -30 36 50 3.17 0.0168
Inferior temporal gyrus 20 L -58 -30 -30 5.56 0.0026
Lingual gyrus 18 L -24 -94 -20 5.45 0.0028
Middle frontal gyrus 46 L -32 50 36 3.46 0.0128
Middle occipital gyrus 19 L -42 -80 6 2.35 0.0400
Middle temporal gyrus 21 L -64 -6 -10 3.59 0.0110
21 L -54 -36 -8 3.16 0.0170
Orbitofrontal cortex 11 L -4 54 -26 5.01 0.0037
11 R 12 42 -26 4.38 0.0060
Parahippocampal gyrus 27 L -18 -42 -4 3.39 0.0136
Supramarginal gyrus 40 L -56 -50 28 2.47 0.0340
Light Stimulation > Baseline (Relative Subtraction)
Calcarine cortex 17 L -16 -72 6 8.65 0.0005
18 L -12 -98 -10 4.44 0.0057
17 R 16 -102 2 4.43 0.0057
Inferior temporal gyrus 20 L -60 -26 -16 6.00 0.0019
Middle temporal gyrus 21 L -62 0 -14 3.05 0.0188
Orbitofrontal cortex 11 L -6 54 -24 10.32 0.0003
11 R 12 42 -26 5.44 0.0028
Parahippocampal gyrus 27 L -16 -40 -6 6.91 0.0011
Superior parietal gyrus 7 R 28 -72 54 3.55 0.0119
Baseline > Light Stimulation (Absolute Subtraction)
None
Baseline > Light Stimulation (Relative Subtraction)
Fusiform gyrus 37 L -56 -64 -8 3.09 0.0181
Heschl’s gyrus 42 L -36 -24 4 9.61 0.0003
42 R 46 -16 -2 4.80 0.0043
Inferior frontal gyrus 45 R 50 22 24 5.75 0.0023
Inferior parietal gyrus 7 L -36 -70 28 3.07 0.0186
Inferior temporal gyrus 20 L -36 -14 -38 3.28 0.0150
20 R 34 0 -42 3.32 0.0146
Superior temporal gyrus 22 R 46 -16 -2 4.80 0.0040
Supplementary motor 6 R 50 -2 30 4.06 0.0077
Supramarginal gyrus 40 R 6 -32 32 8.32 0.0006
Temporopolar area 38 R 34 10 -22 15.62 0.0001
Table 5: Five normal controls light stimulation vs. baseline SPM summary.
132
9.1.3 TcES (DTL-Plus) vs. Baseline
Five normal controls (NC 1, 2, 4, 5, 6; age: 39.4 ± 11.6 years) underwent TcES using
DTL-Plus corneal electrode. The average phosphene threshold current was 0.59 ± 0.23
mA, with a range between 0.35 mA and 0.95 mA. The average stimulation current used
during the 30-minute TcES was 0.93 ± 0.33 mA.
Activation SPM results are shown in Figures 50 and 52 and the findings are
summarized in Tables 6. Compared to baseline, brain activation was observed in
secondary visual cortex (BA 18) bilaterally and secondary visual cortex (BA 19) on the
left hemisphere. Activation was more significant and extensive on the left occipital cortex
than that of the right, and most visual cortex activation was confined to below the
calcarine fissure. No activation was observed in the primary visual cortex (calcarine
cortex) of both hemispheres at p < 0.05 statistical threshold.
In addition to visual cortex activation, extensive increase in metabolic activity
was observed in the left orbitofrontal cortex (BA 11), left frontal eye fields (BA 8), and
bilateral anterior prefrontal cortex (BA 10). These cortical areas play a role in visual
attention and fixation [2, 96, 244]. Activation was also observed in the left supramarginal
gyrus (BA 40) and left pars triangularis (BA 45), which are also implicated in attention.
Activation in the left inferior temporal gyrus (BA 20) and left parahippocampal gyrus
(BA 27) were observed, and they are involved in form perception and memory [244].
Left caudate nucleus was also activated, whose function has been implicated in selective
visual attention [35]. Comparing to light stimulation condition, TcES (DTL-Plus) resulted
in activation of more extensive ventral medial, dorsal lateral and anterolateral prefrontal
133
cortex; this may be attributed to more attention required for perceiving the faint and
peripherally located phosphene. Similar to the light stimulation condition, the activations
were predominantly localized on the left hemisphere.
Cortical areas that exhibited decreased metabolic activity during TcES (DTL-
Plus) are shown in Figures 51 and 53, with the summary of findings listed in Tables 6.
Absolute subtraction did not yield brain areas that showed any significantly decreased
activity during TcES. Relative subtraction found decreased cortical activity in the right
auditory association cortex (BA 42 and 22), bilateral somatosensory association cortex
(BA 5), and left somatosensory association cortex (BA 7). Decreased activity was also
found in the right supramarginal gyrus (BA 40), which has been implicated in
somatosensory association [221]; bilateral fusiform gyrus (BA 37); and right
temporopolar area (BA 38), which is involved in integration of sensory and limbic inputs.
Decreases in auditory and somatosensory cortex can be ascribed to cross-modal
suppression [104, 139, 221].
In our analysis, TcES (DTL-Plus) did not elicit significant activation (p<0.05) in
the primary visual cortex. It is possible the weak phosphene sensation did not strongly
activate the primary visual cortex to achieve statistical threshold in SPM. In order to
demonstrate retinotopic mapping of electrical retinal stimulation in the primary visual
cortex, a better electrode paradigm for TcES that can elicit higher phosphene intensity
and lead to more significant activation in the primary visual cortex is needed. ERG-Jet
contact lens electrode is thus chosen for TcES because of its ability to achieve higher
subjective phosphene intensity.
134
Figure 50: Five normal controls TcES (DTL-Plus) activation SPM maps. Selected Brodmann
areas that exhibit activation during TcES (DTL-Plus) are labeled in green.
135
Figure 51: Five normal controls TcES (DTL-Plus) deactivation SPM maps. Selected
Brodmann areas that exhibit deactivation during TcES (DTL-Plus) are labeled in green.
136
Figure 52: Five normal controls TcES (DTL-Plus) activation 3D SPM projection. Selected
Brodmann areas that exhibit activation during TcES (DTL-Plus) are labeled in yellow.
Figure 53: Five normal controls TcES (DTL-Plus) deactivation 3D SPM projection. Selected
Brodmann areas that exhibit deactivation during TcES (DTL-Plus) are labeled in yellow.
137
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
TcES DTL-Plus > Baseline (Absolute Subtraction)
Caudate nucleus L -10 10 8 2.45 0.035
Frontal eye field 8 L -32 26 56 2.79 0.025
Inferior frontal gyrus 45 L -56 32 4 3.73 0.010
Inferior occipital gyrus 18 L -22 -100 -10 3.66 0.011
19 L -38 -90 -8 4.75 0.005
18 R 32 -94 -10 3.25 0.016
Inferior temporal gyrus 20 R 64 -44 -16 3.78 0.009
20 R 42 4 -46 3.14 0.017
20 L -52 -34 -16 4.23 0.007
Orbitofrontal cortex 11 R 28 56 -2 3.15 0.017
11 L -26 64 2 3.14 0.017
Parahippocampal gyrus 27 L -16 -42 -2 3.48 0.013
Superior temporal gyrus 22 L -56 -10 -8 3.23 0.016
Supramarginal gyrus 40 L -56 -40 34 2.94 0.021
TcES DTL-Plus > Baseline (Relative Subtration)
Inferior frontal gyrus 45 L -54 38 4 2.48 0.034
Orbitofrontal cortex 11 L -10 60 -18 2.58 0.030
Parahippocampal gyrus 27 L -14 -36 0 4.91 0.004
Superior frontal gyrus 10 L -12 56 2 4.61 0.005
10 R 22 40 0 19.23 0.00002
Baseline > TcES DTL-Plus (Absolute Subtraction)
None
Baseline > TcES DTL-Plus (Relative Subtraction)
Fusiform gyrus 37 L -52 -66 -4 3.15 0.017
37 R 52 -62 -18 3.25 0.016
Heschl’s gyrus 42 R 58 -24 16 3.97 0.008
Posterior insular cortex L -36 -8 -6 4.39 0.006
R 44 0 -4 16.52 0.00004
Precuneus cortex 5 L -4 -48 72 4.65 0.004
Superior parietal gyrus 7 L -26 -74 40 2.41 0.036
Superior temporal gyrus 22 R 62 -6 8 3.48 0.013
Superior temporal pole 38 R 56 4 -2 4.06 0.007
Supramarginal gyrus 40 R 64 -28 24 5.00 0.004
Table 6: Five normal controls TcES (DTL-Plus) vs. baseline SPM summary.
138
9.1.4 Light Stimulation vs. TcES (DTL-Plus)
To compare differences in the pattern of activation between light stimulation and TcES
(DTL-Plus), absolute and relative subtraction analyses were conducted for the same five
normal control subjects (NC 1 – NC 5) who underwent both conditions. SPM results are
shown in Figures 54 - 57 and the summary of finding is listed in Tables 7. Absolute
analysis did not yield any brain area that was more activated during light stimulation than
TcES (DTL-Plus). However, relative analysis showed more activation in bilateral
primary visual cortex (BA 17) and bilateral secondary visual cortices (BA 18 and 19). In
addition, more activation was observed in the right supramarginal gyrus (BA 40) and
anterior region of the right inferior temporal gyrus (BA 20). The finding suggests light
sitmulation activated more strongly areas of the visual cortex than TcES (DTL-Plus).
Brain areas more activated during TcES (DTL-Plus) than light stimulation were
found predominantly in the left prefrontal cortex (Figure 55). More activation was seen in
the left orbitofrontal area (BA 11), left frontopolar area (BA 10), left pars triangularis
(BA 45), left pars opercularis (BA 44), left prefrontal gyrus (BA 47), bilateral
supramarginal gyrus (BA 40), bilateral angular gyrus (BA 39), and bilateral inferior
temporal gyrus (BA 20). In addition, bilateral supplementary motor cortex (BA 6) and
right frontal eye fields (BA 8) also showed more activation. These cortical areas have
been implicated in saccades and eye movements [2]. The finding suggests TcES (DTL-
Plus) activated more extensive brain areas involved in visual attention, fixation and
saccades than light stimulation.
139
Figure 54: Five normal controls SPM map - Light stimulation > TcES (DTL-Plus). Selected
Brodmann areas that demonstrate more activation during light stimulation than TcES (DTL-Plus)
are labeled in green.
140
Figure 55: Five normal controls SPM map - TcES (DTL-Plus) > Light stimulation. Selected
Brodmann areas that demonstrate more activation during TcES (DTL-Plus) than light stimulation
are labeled in green.
141
Figure 56: Normal control 3D SPM projection - Light stimulation > TcES (DTL-Plus).
Selected Brodmann areas that exhibit more activation during light stimulation than TcES (DTL-
Plus) are labeled in yellow.
Figure 57: Normal control 3D SPM projection - TcES (DTL - Plus) > Light stimulation.
Selected Brodmann areas that exhibit more activation during TcES (DTL-Plus) than light
stimulation are labeled in yellow.
142
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
Light Stimulation > TcES DTL-Plus (Absolute Subtraction)
None
Light Stimulation > TcES DTL-Plus (Relative Subtration)
Calcarine cortex 17 L -6 -92 -4 8.90 0.0004
17 L -10 -72 6 4.03 0.0079
17 R 6 -86 -2 8.51 0.0005
17 L -14 -64 20 14.85 0.0001
Cuneus cortex 19 R 14 -82 38 5.52 0.0026
Fusiform gyrus 37 R 32 -54 0 10.48 0.0002
37 R 50 -72 -2 5.00 0.0037
Inferior frontal gyrus 44 R 56 6 6 9.45 0.00035
Inferior occipital gyrus 18 L -18 -98 -10 19.93 0.00002
18 R 24 -94 -6 18.83 0.00002
Middle occipital gyrus 19 L -48 -73 2 3.12 0.0170
19 L -26 -58 4 4.65 0.0048
Middle temporal gyrus 21 R 64 -52 -2 4.04 0.0078
Putamen R 34 0 2 4.54 0.0052
Supramarginal gyrus 40 R 54 -28 18 27.60 0.00001
Temporopolar area 38 R 32 12 -28 7.49 0.0009
TcES DTL-Plus > Light Stimulation (Absolute Subtraction)
Angular gyrus 39 L -42 -60 38 3.23 0.0160
39 R 52 -60 42 3.17 0.0170
Frontal eye fields 8 R 8 26 62 3.13 0.0170
Inferior frontal gyrus 44 L -52 8 10 12.11 0.0001
44 L -40 16 30 6.93 0.0012
Inferior prefrontal gyrus 47 L -38 38 0 6.39 0.0015
Inferior temporal gyrus 20 L -50 0 -32 11.59 0.0002
20 R 38 -4 -42 3.45 0.0130
Middle temporal gyrus 21 L -48 -8 -14 7.89 0.0007
Precuneus cortex 5 R 2 -50 70 6.06 0.0019
Superior frontal gyrus 10 L -14 58 12 3.99 0.0080
Superior parietal gyrus 7 L -24 -56 36 8.40 0.0006
Supplementary motor 6 L -37 -2 37 4.65 0.0048
6 R 12 -16 58 3.31 0.0148
Supramarginal gyrus 40 L -44 -22 26 8.18 0.0006
40 R 42 -36 52 3.14 0.0170
TcES DTL-Plus > Light Stimulation (Relative Subtraction)
Frontal eye field 8 R 18 22 64 4.29 0.0064
Inferior frontal gyrus 45 L -48 26 2 3.61 0.0112
Inferior temporal gyrus 20 L -32 -6 -34 5.17 0.0033
20 R 42 -6 -38 5.23 0.0032
Middle frontal gyrus 10 L -6 46 4 13.07 0.0001
Orbitofrontal cortex 11 L -28 56 -4 14.48 0.0001
Table 7: Five normal controls light stimulation vs. TcES (DTL-Plus) SPM summary.
143
9.1.5 TcES (ERG-Jet) vs. Baseline
Five normal controls (NC 1, 3, 4, 5, 6; age: 34.3 ± 10.0 yrs) underwent TcES using ERG-
Jet corneal electrode. The average phosphene threshold current was 0.72 ± 0.18 mA, with
a range between 0.55 mA and 1.00 mA. The average stimulation current used during the
30-minute TcES was 1.08 ± 0.28 mA.
Activation SPM results are shown in Figures 58 and 60 and the findings are
summarized in Table 8. Compared to baseline, an activation locus was observed in the
anterior calcarine cortex on the left hemisphere (BA 17 labeled in axial slice 8 of Figure
58). This region of the primary visual cortex receives visual input from the nasal
hemiretina of the right eye, and corresponded well to right peripheral phosphene
sensation reported by the normal controls. There was also activation in bilateral
secondary visual cortex (BA 18) and left secondary visual cortex (BA 19).
In addition to visual cortex activation, extensively increased activity was observed
in bilateral orbitofrontal cortex (BA 11), right anterior prefrontal cortex (BA 10), left
frontal eye fields (BA 8), left pars opercularis (BA 44), and left caudate nucleus. These
brain areas play a role in visual attention, fixation and saccade [2, 35, 96, 244]. Strong
activation was also observed in the left primary and association sensory cortices (BA 3, 2,
1), left subcentral area (BA 43) and left supramarginal gyrus (BA 40). Activation in these
areas reflects sensation of ERG-Jet contact lens electrode on the eye during TcES.
Activation of bilateral inferior temporal gyrus (BA 20) and left parahippocampal gyrus
(BA 27) were observed, and they are involved in form perception and memory [244].
144
Figure 58: Five normal controls TcES (ERG-Jet) activation SPM maps. Selected Brodmann
areas that exhibit activation during TcES (ERG-Jet) are labeled in green.
145
Figure 59: Five normal controls TcES (ERG-Jet) deactivation SPM maps. Selected
Brodmann areas that exhibit deactivation during TcES (ERG-Jet) are labeled in green.
146
Figure 60: Five normal controls TcES (ERG-Jet) activation 3D SPM projection. Selected
Brodmann areas that exhibit activation during TcES (ERG-Jet) are labeled in yellow.
Figure 61: Five normal controls TcES (ERG-Jet) deactivation 3D SPM projection. Selected
Brodmann areas that exhibit deactivation during TcES (ERG-Jet) are labeled in yellow.
147
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
TcES ERG-Jet > Baseline (Absolute Subtraction)
Calcarine cortex 17 L -18 -74 10 2.49 0.033
17 R 18 -84 6 2.45 0.035
Caudate nucleus L -12 8 10 2.40 0.036
Frontal eye field 8 L -28 32 54 2.93 0.021
Fusiform gyrus 37 L -56 -50 -22 3.45 0.013
Inferior occipital gyrus 18 L -16 -102 -6 3.17 0.017
18 R 26 -100 -8 2.46 0.034
Inferior frontal gyrus 44 R 46 8 14 2.88 0.022
Inferior parietal gyrus 40 L -58 -50 44 2.84 0.023
Inferior temporal gyrus 20 L -62 -50 -18 4.36 0.006
20 R 64 -32 -22 2.75 0.025
Middle frontal gyrus 10 R 40 58 4 3.29 0.015
Middle occipital gyrus 19 L -50 -78 0 2.8 0.024
Orbitofrontal cortex 11 L -8 50 -28 5.26 0.003
11 L -22 68 0 3.42 0.013
Parahippocampal gyrus 27 L -14 -36 -12 2.64 0.028
Postcentral gyrus 3,2,1 L -48 -24 60 2.85 0.023
Precuneus cortex 5 L -16 -40 52 3.42 0.013
5 R 4 -42 74 2.71 0.026
7 L -10 -66 58 2.42 0.036
Subcentral area 43 L -64 -2 24 3.43 0.013
TcES ERG-Jet > Baseline (Relative Subtraction)
Caudate R 6 14 -2 5.48 0.003
Inferior temporal gyrus 20 L -64 -36 -14 3.4 0.014
Orbitofrontal cortex 11 L -2 48 -28 6.89 0.001
11 R 8 38 -24 3.43 0.013
Baseline > TcES ERG-Jet (Absolute Subtraction)
None
Baseline > TcES ERG-Jet (Relative Subtraction)
Fusiform gyrus 37 R 54 -66 -14 3.44 0.013
Heschl’s gyrus 42 R 48 -16 6 2.28 0.040
Inferior temporal gyrus 20 R 38 -14 -38 5.39 0.003
Putamen R 22 10 -10 7.00 0.001
Superior temporal gyrus 22 R 50 -34 6 2.39 0.037
Supplementary motor 6 L -6 4 74 4.52 0.005
6 R 8 -10 78 3.17 0.017
Supramarginal gyrus 40 R 42 -38 48 3.66 0.011
Table 8: Five normal controls TcES (ERG-Jet) vs. baseline SPM summary.
148
Absolute subtraction did not yield any brain region that exhibited statistically
significant decrease in activity (Figure 59 and 61). Relative subtraction found decreased
activity in the right auditory association cortex (BA 42, 22), right supramarginal gyrus
(BA 40). Moreover, deactivation was observed in the right inferior temporal gyrus (BA
20) near the anterior temporal pole, which plays a role in multisensory association and
emotion. Decreased activity was also observed in bilateral supplementary motor cortex
(BA 6) and right fusiform gyrus (BA 37).
9.1.6 Light Stimulation vs. TcES (ERG-Jet)
Results from the SPM comparison between light stimulation and TcES (ERG-Jet) in the
normal control group are shown in Figures 62 - 64 and the summary listed in Table 9.
Absolute analysis did not yield any brain area that was statistically more activated during
light stimulation than TcES (ERG-Jet). However, relative analysis showed stronger
activation in the occipital cortex including left anterior calcarine cortex (BA 17), bilateral
secondary visual cortex (BA 18), and right secondary visual cortex (BA 19). Stronger
activation was also observed in the right middle temporal gyrus (BA 21), bilateral
fusiform gyrus (BA 37) and right parahippocampal gyrus (BA 27). Brain areas more
activated during TcES (ERG-Jet) than light stimulation were found predominantly in the
right prefrontal cortex (BA 9, 10, 44, 45, and 47), right frontal eye field (BA 8), left
orbitofrontal cortex (BA 11). The finding suggests light stimulation activated more
strongly areas of the visual cortex than TcES (ERG-Jet), and TcES (ERG-Jet) activated
more strongly areas involved in visual attention and saccades than light stimulation.
149
Figure 62: Five normal controls SPM map – Light Stimulation > TcES (ERG-Jet). Selected
Brodmann areas that exhibit more activation during light stimulation than TcES (ERG-Jet) are
labeled in green.
150
Figure 63: Five normal controls SPM map – TcES (ERG-Jet) > Light Stimulation. Selected
Brodmann areas that exhibit more activation during TcES (ERG-Jet) than light stimulation are
labeled in green.
151
Figure 64: Normal control 3D SPM projection - Light Stimulation vs. TcES (ERG-Jet).
Selected Brodmann areas are labeled in yellow.
152
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
Light Stimulation > TcES ERG-Jet (Absolute Subtraction)
None
Light Stimulation > TcES ERG-Jet (Relative Subtraction)
Calcarine cortex 17 L -20 -58 18 3.04 0.019
Fusiform gyrus 37 L -52 -62 4 2.51 0.033
37 R 56 -64 -4 3.77 0.010
Inferior occipital gyrus 18 R 28 -88 -14 15.34 0.00005
19 R 42 -72 -6 4.16 0.007
Lingual gyrus 18 L -20 -94 -16 7.29 0.001
Middle temporal gyrus 21 R 56 -42 0 6.92 0.001
21 R 52 0 -26 2.89 0.022
Parahippocampal gyrus 27 R 28 -38 -4 9.13 0.0004
TcES ERG-Jet > Light Stimulation (Absolute Subtraction)
Angular gyrus 39 R 54 -62 38 3.70 0.010
Caudate L -6 2 8 4.19 0.007
Dorsolateral prefrontal
cortex
9 R 22 44 38 3.78 0.010
Frontal eye field 8 R 22 26 60 4.02 0.008
Fusiform gyrus 37 L -60 -54 -14 2.75 0.025
Inferior frontal gyrus 44 R 56 18 12 3.27 0.015
Inferior prefrontal gyrus 47 R 34 36 2 4.71 0.005
Orbitofrontal cortex 11 L -18 62 -12 3.67 0.011
Precuneus 7 L -4 -68 60 4.64 0.005
5 L -8 -40 54 4.16 0.007
7 R 4 -60 66 4.63 0.005
Subcentral area 43 L -66 -6 14 5.08 0.004
Superior frontal gyrus 10 R 20 50 4 3.79 0.010
Superior frontal gyrus 46 R 24 50 18 3.19 0.017
TcES ERG-Jet > Light Stimulation (Relative Subtraction)
Fusiform gyrus 37 L -60 -56 -18 4.83 0.004
Inferior frontal gyrus 45 R 52 24 -2 3.34 0.014
Inferior prefrontal gyrus 47 L -34 48 -4 5.90 0.002
Inferior temporal gyrus 20 L -30 -4 -48 6.60 0.001
20 L -64 -36 -14 2.43 0.035
20 R 48 -12 -22 3.89 0.009
Middle frontal gyrus 10 R 8 46 0 4.11 0.007
Orbitofrontal cortex 11 L -10 42 -24 8.26 0.001
Table 9: Five normal controls light stimulation vs. TcES (ERG-Jet) SPM summary.
153
9.2 Retinal Degeneration Subjects
Five subjects with retinal degeneration (RD) were recruited for the study (age: 43.7 ±
10.5 yrs). All five subjects met the inclusion/exclusion criteria outlined in Section 8.1.1
and have had at least ten years of light-perception or near light-perception vision. Table
10 shows the identifying data for all five subjects. To minimize potential confounding
factors, the subjects all have early-onset visual loss occurring before adulthood. The
location and extent of remaining vision at the time of PET study varied between the five
subjects: RD 1 reported seeing best in her right mid peripheral visual field of her right
eye; RD 2 reported losing mostly central vision with preservation of peripheral sight; RD
3 reported tunnel vision, with light-perception vision in a pinhole near the central visual
field of his right eye; RD 4 and 5 all reported severe loss of peripheral vision with more
preservation of central vision. RD 2 appears to have the best vision in the group and RD 4
the worst.
DNA testing was conducted at Columbia University Medical Center and the
results are shown in Table 10. One subject has a novel form of CRX mutation, and a
second has a mutation in the ABCA4 gene. The remaining three subjects have mutations
that have not yet been identified based on ARRPgenetest (Asper Ophthalmics, Estonia).
Optical coherence tomography (OCT) study conducted on the RD subjects
showed primarily loss of photoreceptors in the outer nuclear layer while the retinal nerve
fiber layer was preserved. The OCT results for RD 1 are shown in Figure 65. OCT line
raster (Figure 65b) demonstrated loss of photoreceptor in the fovea as well as perifoveal
region of the retina. The average normal retinal thickness and retinal volume are 250 µm
154
and 10 mm
3
, respectively. For RD 1 subject, an average retinal thickness of 234 µm and
retinal volume of 7.5 mm
3
are below normal (Figure 65d), and the reduction in both
parameters are attributed to photoreceptor loss in the outer retina. Retinal nerve fiber
layer thickness for RD 1 was within normal range in all four retinal quadrants (Figure
65c), indicating preservation of retinal ganglion cell axons.
Figure 65: OCT study of RD 1’s right eye. (a) Color fundus photo of the right eye; blue dashed
line represents location of the OCT line raster plot; (b) corresponding OCT raster plot; yellow
arrowhead indicates photoreceptor loss in the foveal pit; pink arrowheads indicate perifoveal
photoreceptor loss; (c) retinal nerve fiber layer (RNFL) measurement in all four quadrants of the
retina; (d) retinal thickness, as measured from inner limiting membrane (ILM) to retinal pigment
epithelium (RPE). I: inferior; N: nasal; S: superior; T: temporal.
155
156
9.2.1 Subjective Sensation of Stimulation
Duing light stimulation, the RD subjects generally perceived a diffuse, white flashing
light from their right eye. The quality of the light sensation was described as ‘cloudy.’
RD 1 reported seeing the light best in her right mid peripheral visual field; RD 2 reported
seeing a circular rim of flashing light around his mid periphery; RD 3 reported seeing
pulsing light the size of a pinhead in the center of his visual field; and both RD 4 and RD
5 reported seeing the light stimulus in their central visual field. When asked to grade the
light intensity on a scale of 1 – 10 (‘10’ being the brightest light the subject has ever
seen), they gave a rating between 5 and 6 when the photic stimulation was turned on and
between 1 and 2 when it was turned off.
Following TcES using ERG-Jet electrode, subjects were asked to describe the
shape and size of the phosphene by touching a piece of Play-Doh kneaded into a circular
torus to represent their visual field (Figure 66, upper left). All five RD subjects perceived
phosphene in their right peripheral visual field during TcES, although there is some
variation in reported phosphene shape and size (Figure 66). The phosphene was reported
to be faint and rapid, but readily perceived. On an intensity scale of 1-10, the RP subjects
rated the phosphene between 2 and 3. Four subjects (RD 1, 2, 3, and 5) also noted streak-
like radiation of the phosphene eminating from the location of highest intensity to along
the right visual periphery and towards the center of the visual field. Subjects also reported
seeing the brightest phosphene immediately after the stimulator was turned on, and the
intensity gradually faded with subsequent electrical pulses, but remained perceptible. The
157
alternating monophasic pulse train resulted in perception of alternating phosphene
intensity; cathodic pulse was perceived to be brighter than anodic pulse.
Figure 66: Phosphene sensation reported by RD subjects during TcES (ERG-Jet). One
subject describing phosphene sensation using Play-Doh (upper left); pink overlay on the visual
field grid indicates location and shape of phosphene reported by each RD subject.
9.2.2 Light Stimulation vs. Baseline
Light activation SPM results for the RD subjects are shown in Figures 67 and 69 and the
findings are summarized in Table 11. Compared to baseline, light stimulation activated a
small region of primary visual cortex (BA 17) in the occipital pole on the right
158
hemisphere. There was also extensive secondary visual cortex (BA 18 and 19) activation
on the left hemisphere. In addition to visual cortex, significant activation was observed in
the bilateral orbitofrontal cortex (BA 11), bilateral inferior prefrontal cortex (BA 47), and
right dorsolateral prefrontal cortex (BA 46). These cortical areas have been shown to play
a role in visual attention and fixation [2, 96, 244]. Activation was also observed in
bilateral insular cortex and right temporopolar area (BA 38); these brain areas have been
implicated in emotion.
Brain areas that exhibited decreased activity during light stimulation are shown in
Figures 68 and 70, with the summary of finding listed in Table 11. Absolute subtraction
did not yield any brain area that showed significantly decreased activity during light
stimulation. Relative subtraction found decreased cortical activity in the right primary
auditory cortex (BA 42), left supramarginal gyrus (BA 40) and left somatosensory cortex
(BA 3). Decreased activity was also found in the left subcentral area (BA 43), which has
been implicated in somatosensory association [221]; anterior portion of left inferior
temporal gyrus (BA 20) and left temporopolar area (BA 38), which are involved in
integration of sensory and limbic inputs. There was also deactivation in areas of
secondary visual cortex (BA 19) bilaterally near the parieto-occipital junction. These
areas are involved in conveying visual spatial information to higher integrative areas in
the parietal cortex where other sensory modalities converge to form a map for spatial
orientation. In summary, decreased activity in auditory and somatosensory association
cortical areas during light stimulation can be ascribed to cross-modal suppression [104,
139, 221].
159
Figure 67: Five RD subjects light stimulation activation SPM maps. Selected Brodmann areas
that exhibit activation during light stimulation are labeled in green.
160
Figure 68: Five RD subjects light stimulation deactivation SPM maps. Selected Brodmann
areas that exhibit deactivation during light stimulation are labeled in green.
161
Figure 69: Five RD subjects light stimulation activation 3D SPM projection. Selected
Brodmann areas that exhibit activation during light stimulation are labeled in yellow.
Figure 70: Five RD subjects light stimulation deactivation 3D SPM projection. Selected
Brodmann areas that exhibit deactivation during light stimulation are labeled in yellow.
162
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
Light Stimulation > Baseline (Absolute Subtraction)
Calcarine cortex 17 R 20 -96 0 3.37 0.0140
Fusiform gyrus 37 R 54 -63 -18 2.54 0.0320
Inferior occipital gyrus 18 L -28 -86 -6 4.91 0.0040
19 L -48 -74 -14 4.36 0.0060
Inferior prefrontal gyrus 47 L -32 36 -14 4.90 0.0040
47 R 36 38 -16 3.37 0.0140
Insular cortex L -30 2 16 3.29 0.0150
R 44 8 4 3.54 0.0120
Lingual gyrus 18 L -12 -90 -16 3.45 0.0130
18 R 34 -86 -18 3.16 0.0170
Middle temporal gyrus 46 R 38 52 16 2.46 0.0350
Orbitofrontal cortex 11 L -8 38 -16 2.94 0.0210
11 R 26 36 -18 5.95 0.0020
Temporopolar area 38 R 40 18 -34 3.74 0.0100
Thalamus R 8 -16 12 3.16 0.0170
Light Stimulation > Baseline (Relative Subtraction)
Fusiform gyrus 37 R 58 -58 -18 2.33 0.0400
Inferior occipital gyrus 19 L -42 -80 -16 4.01 0.0080
Middle frontal gyrus 46 R 36 58 12 5.60 0.0025
Orbitofrontal cortex 11 L -34 50 -16 5.32 0.0030
11 R 32 58 -10 5.23 0.0032
Temporopolar area 38 R 40 18 -34 2.30 0.0420
Baseline > Light Stimulation (Absolute Subtraction)
None
Baseline > Light Stimulation (Relative Subtraction)
Angular gyrus 39 L -40 -60 32 5.79 0.0022
Cuneus cortex 19 R 10 -88 34 5.22 0.0032
Inferior prefrontal gyrus 47 L -44 24 -2 3.75 0.0100
Inferior temporal gyrus 20 L -46 -4 -38 3.54 0.0120
Middle cingulum 23 L -8 -44 36 5.53 0.0026
Postcentral gyrus 3 L -44 -30 60 2.43 0.0360
Precuneus cortex R 12 -46 44 4.55 0.0052
Subcentral area 43 L -58 -8 24 6.32 0.0016
Superior occipital gyrus 19 L -24 -78 30 3.10 0.0180
Superior temporal gyrus 42 R 60 -38 22 4.55 0.0052
Supramarginal gyrus 40 L -60 -48 32 4.32 0.0062
Temporopolar area 38 L -48 16 -16 5.87 0.0021
Table 11: RD subjects light stimulation vs. baseline SPM summary.
163
9.2.3 TcES (ERG-Jet) vs. Baseline
TcES (ERG-Jet) activation SPM results for the RD subjects are shown in Figures 71 and
73 and the findings are summarized in Table 12. Compared to baseline, TcES activated a
small region of anterior calcarine cortex in the left hemisphere (BA 17, slice 12 of Figure
71). This area of primary visual cortex receives visual input from the peripheral nasal
retina of the right eye and is consistent with the sensation of right peripheral phosphene
reported by the RD subjects. There was also bilateral activation in secondary visual
cortex (BA 18) and activation of left secondary visual cortex (BA 19) near the parieto-
occipital junction. In addition to visual cortex, significant activation was observed in the
bilateral frontal eye field (BA 8), right dorsolateral prefrontal cortex (BA 9), and left
inferior prefrontal gyrus (BA 47); activation in these brain areas is possibly involved in
visual attention and saccade [2, 96, 244]. Activation was also seen in bilateral primary
motor area (BA 4), supplementary motor cortex (BA 6), and somatosensory cortex (BA
3, 2, 1), with stronger activation on the left hemisphere; activation in these brain areas
reflect sensation of ERG-Jet contact lens on the right eye during TcES and fine motor
control exerted by the subjects to keep the electrode in place on the cornea.
TcES (ERG-Jet) deactivation SPM results are shown in Figures 72 and 74, and
summarized in Tables 12. Absolute subtraction did not yield significantly decreased
activity in any brain area. Relative subtraction found decreased activity in the right
primary auditory cortex (BA 42), left somatosensory association cortex (BA 7), left
supramarginal gyrus (BA 40), and dorsal occipital cortex (BA 19). Decreases in auditory
and somatosensory cortex can be ascribed to cross-modal suppression [104, 139, 221].
164
Figure 71: Five RD subjects TcES (ERG-Jet) activation SPM maps. Selected Brodmann areas
that exhibit activation during TcES (ERG-Jet) are labeled in green.
165
Figure 72: Five RD subjects TcES (ERG-Jet) deactivation SPM maps. Selected Brodmann
areas that exhibit deactivation during TcES (ERG-Jet) are labeled in green.
166
Figure 73: Five RD subjects TcES (ERG-Jet) activation 3D SPM projection. Selected
Brodmann areas that exhibit activation during TcES (ERG-Jet) are labeled in yellow.
Figure 74: Five RD subjects TcES (ERG-Jet) deactivation 3D SPM projection. Selected
Brodmann areas that exhibit deactivation during TcES (ERG-Jet) are labeled in yellow.
167
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
TcES (ERG-Jet) > Baseline (Absolute Subtraction)
Calcarine cortex 17 L -18 -56 12 2.60 0.030
17 R 12 -96 -2 4.37 0.006
Cuneus cortex 18 R 14 -92 24 2.48 0.034
Dorsolateral prefrontal
cortex (DLPFC)
9 R 28 38 42 4.21 0.0068
Frontal eye field 8 L -32 12 62 3.75 0.010
8 R 20 32 54 4.17 0.007
Inferior prefrontal gyrus 47 L -46 26 -8 4.50 0.0055
Inferior temporal gyrus 20 L -48 0 -36 2.55 0.032
Lingual gyrus 18 L -14 -96 -12 2.90 0.022
18 R 16 -94 -18 4.01 0.008
Middle occipital gyrus 18 L -20 -92 6 2.60 0.030
19 L -38 -78 32 3.37 0.014
Orbitofrontal cortex 11 L -8 62 -18 2.45 0.036
Postcentral gyrus 3,2,1 L -34 -34 64 2.38 0.038
3,2,1 R 24 -38 72 2.21 0.046
Posterior cingulate cortex 23 R 5 -48 30 2.77 0.025
Precentral gyrus 4 L -34 -24 62 3.23 0.016
4 R 36 -22 62 2.35 0.039
Precuneus cortex 5 L -8 -46 74 2.29 0.042
Superior temporal gyrus 22 R 56 -28 10 2.95 0.021
Supplementary motor 6 L -8 14 68 2.43 0.036
6 R 8 16 66 4.32 0.006
Temporopolar area 38 R 56 6 -2 3.30 0.015
TcES (ERG-Jet) > Baseline (Relative Subtraction)
DLPFC 9 L -20 34 46 2.78 0.0250
Frontal eye field 8 R 22 28 52 6.43 0.0015
Inferior temporal gyrus 20 L -36 10 -40 4.60 0.0050
Lingual gyrus 18 L -14 -90 -14 4.91 0.0040
Middle occipital gyrus 19 L -38 -78 34 2.95 0.0210
Precentral gyrus 4 L -36 -24 62 4.91 0.0040
4 R 40 -18 60 3.10 0.0180
4 R 14 -30 60 3.37 0.0140
Supplementary Motor 6 L -30 -14 68 5.95 0.0020
6 L -40 8 58 5.66 0.0024
6 R 12 20 60 6.68 0.0013
Temporopolar area 38 L -48 24 -10 3.64 0.0110
Baseline > TcES (ERG-Jet) (Absolute Subtraction)
None
Table 12: RD subjects TcES (ERG-Jet) vs. baseline SPM summary.
168
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
Baseline > TcES (ERG-Jet) (Relative Subtraction)
Inferior parietal gyrus 7 L -28 -58 40 4.55 0.005
Middle occipital gyrus 19 L -24 -76 30 2.82 0.024
Superior occipital gyrus 19 R 16 -88 34 3.23 0.016
Superior temporal gyrus 42 R 54 -38 20 5.42 0.003
Supramarginal gyrus 40 R 54 -38 50 2.63 0.029
Table 12 (Continued): RD subjects TcES (ERG-Jet) vs. baseline SPM summary.
9.2.4 Light Stimulation vs. TcES (ERG-Jet)
Results from the SPM comparison between light stimulation and TcES (ERG-Jet) for the
RD subjects are shown in Figures 75 - 77 and summarized in Table 13. Absolute analysis
found only one region of the orbitofrontal cortex (BA 11) on the right hemisphere that
was more activated during light stimulation than TcES (ERG-Jet). Relative analysis
showed large areas in the occipital cortex including bilateral secondary visual cortex (BA
18), left secondary visual cortex (BA 19) more strongly activated during light stimulation
than TcES (ERG-Jet). In addition, stronger activation was observed in the fusiform gyrus
(BA 37), supramarginal gyrus (BA 40), angular gyrus (BA 39), inferior prefrontal gyrus
(BA 47), and anterior regions of inferior temporal gyrus (BA 20). The finding suggests
light stimulation activated more strongly areas of the secondary visual cortex than TcES
(ERG-Jet).
Absolute analysis did not yield any brain area more activated during TcES (ERG-
Jet) than light stimulation. Relative analysis found more activation in the right orbito-
169
Figure 75: Five RD subjects SPM map – Light Stimulation > TcES (ERG-Jet). Selected
Brodmann areas that exhibit more activation during light stimulation than TcES (ERG-Jet) are
labeled in green.
170
Figure 76: Five RD subjects SPM map – TcES (ERG-Jet) > Light Stimulation. Selected
Brodmann areas that exhibit more activation during TcES (ERG-Jet) than light stimulation are
labeled in green.
171
Figure 77: RD subjects 3D SPM projection - Light Stimulation vs. TcES (ERG-Jet). Selected
Brodmann areas are labeled in yellow.
172
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
Light Stimulation > TcES ERG-Jet (Absolute Subtraction)
Orbitofrontal cortex 11 R 16 12 -16 2.89 0.022
Light Stimulation > TcES ERG-Jet (Relative Subtration)
Angular gyrus 39 R 52 -58 42 2.97 0.020
Fusiform gyrus 37 R 56 -68 6 5.96 0.002
Inferior occipital gyrus 18 L -22 -90 -10 4.28 0.006
19 L -44 -80 -10 5.96 0.002
18 R 30 -86 -14 9.93 0.0003
Inferior prefrontal gyrus 47 L -48 28 -10 2.44 0.036
Inferior temporal gyrus 20 L -28 -14 -30 2.97 0.020
20 R 42 -12 -28 3.10 0.018
Middle temporal gyrus 21 R 54 2 -20 3.31 0.015
Orbitofrontal gyrus 11 R 4 58 -6 2.52 0.032
Superior frontal gyrus 10 L -12 70 2 2.97 0.020
10 R 28 62 18 2.56 0.031
Superior occipital gyrus 19 L -26 -82 44 5.05 0.004
18 R 16 -100 6 3.57 0.012
Supramarginal gyrus 40 L -40 -38 46 10.93 0.0002
TcES ERG-Jet > Light Stimulation (Absolute Subtraction)
None
TcES ERG-Jet > Light Stimulation (Relative Subtraction)
Orbitofrontal cortex 11 R 12 64 -2 3.88 0.009
Putamen R 16 8 -8 6.09 0.002
Superior temporal gyrus 22 R 54 -22 2 8.14 0.0006
Table 13: RD subjects light stimulation vs. TcES (ERG-Jet) SPM summary.
frontal cortex (BA 11), superior temporal gyrus (BA 22) and putamen during TcES. The
finding, however, does not indicate a general trend in the difference of activation pattern
between TcES (ERG-Jet) and light stimulation to imply specific functional significance.
Normal control comparison (see Section 9.1.6) yielded comparable finding,
except TcES (ERG-Jet) activated more strongly regions in the prefrontal cortex involved
in visual fixation and saccade. This discrepancy may be attributed to the difference in
metabolic activity increase in brain areas associated with visual attention between normal
controls and RD subjects.
173
9.3 Normal Control and Retinal Degeneration Subject Comparison
9.3.1 Phosphene Threshold Comparison
Five subjects from the normal control group participated in each of the two TcES studies,
one with DTL-Plus fiber electrode, and the other with ERG-Jet contact lens electrode (see
Table 3). The average phosphene threshold for DTL-Plus electrode was 0.59 ± 0.23 mA
with a range between 0.35 and 0.95 mA; the average for ERG-Jet electrode was 0.72 ±
0.18 mA with a range between 0.55 and 1.0 mA (Figure 78). Although the average
phosphene threshold current was slightly higher for ERG-Jet than DTL-Plus, the
difference was not statistically significant (p<0.05). The slight difference in phosphene
threshold between the two electrodes can be explained in part by the larger conducting
surface area of the ERG-Jet compared to DTL-Plus electrode and also differences in the
conductivity between silver and gold. Miyake et al. [161], using a corneal contact
electrode, found threshold of 0.29 mA in healthy individuals at 5-ms pulse duration.
Dorfman et al. [53], using a Burian-Allen contact lens electrode, found thresholds
between 2 and 3 mA at 1-ms pulse duration in four healthy individuals. Gekeler et al.
[84] reported an average threshold of 0.3 mA for healthy subjects using DTL electrode.
Our average of 0.59 mA for DTL-Plus electrode and 0.72 mA for ERG-Jet at 2-ms pulse
duration is comparable to their findings. The discrepancy may be due to different
objective methods used for threshold determination. Gekeler used psychophysical
method, whereas Miyake and Dorfman used cortical potential as a guage for making
174
threshold determination. In addition, different types of electrodes used in the study can
also contribute to variations in threshold values.
Five RD subjects participated in TcES using ERG-Jet electrode. The average
phosphene threshold current for the RD group was 3.08 ± 2.01 mA with a range between
1.0 and 5.8 mA. Individual phosphene threshold comparison between normal controls
and RD subjects is shown in Figure 79. Although the threshold current seems to increase
directly with age by virtue of longer duration of RD at older age, it is better correlated
with severity of visual loss. Our findings are comparable to reported values of 4 mA at 4
ms pulse duration using Burian-Allen corneal contact electrode and 2.63 mA at 4 ms
pulse duration using DTL electrode in patients with retinal degeneration [84]. Two
sample t-test showed the threshold current using ERG-Jet was significantly higher
(p<0.05) for RD group than the normal control (Figure 78).
Figure 78: Comparison of phosphene threshold current between normal and RD subjects.
NC-DTL: normal control (TcES using DTL-Plus); NC-JET: normal control (TcES using ERG-
Jet); RD-JET: retinal degeneration subjects (TcES using ERG-Jet). Mean threshold current is
marked by the square; vertical lines represent standard deviation above and below the mean; ‘*’
indicates statistically significant difference (p<0.05).
175
Figure 79: Individual threshold current comparison between normal and RD subjects
9.3.2 Light Stimulation Cross-Comparison
Cross-comparison is performed using two-sample t-test method. Metabolic image from
light stimulation study is first subtracted from the baseline image to yield quantitative
subtraction image for each subject in the two study groups. These subtraction images are
then inputted into SPM for statistical two-sample comparisons.
Results from the light stimulation cross-comparison analysis are shown in Figures
81 – 83, and summarized in Table 14. Absolute analysis reveals only one small region of
the primary visual cortex (BA 17) on the right hemisphere more activated in the normal
176
control than the RD group. Relative analysis shows regions in bilateral primary visual
cortex near the occipital pole (BA 17) and left anterior calcarine cortex (BA 17) to be
more activated in the normal group. This finding is consistent with individual group
analysis (Section 9.1.2 and 9.2.2). Larger area of the right primary visual cortex was
significantly activated during light stimulation in the normal controls than the RD
subjects. This discrepancy can be explained by differences in the extent and location of
visual loss between the RD subjects. To examine this difference in more detail, we
compared each normalized light stimulation subtraction metabolic image (light
stimulation minus baseline) at the same axial slice location (slice = 0 mm) for every
subject in both the normal and RD group (Figure 80). All five normal control metabolic
subtraction images (N1 – N5) reveal consistently increased activity in the right posterior
occipital pole region (enclosed in the blue circle in Figure 80), whereas more variations
are observed in the same occipital region for the five RD subjects. RD 1 reported seeing
light stimulus in the right mid-peripheral visual field of the right eye, which resulted in
increased metabolic activity along the calcarine cortex in the left hemisphere, while no
increase was observed in the right occipital pole. RD 2 reported loss of central vision;
therefore, central light stimulation did not result in increased activity in bilateral occipital
pole. RD 3 saw the light stimulus through a pin-hole in the central visual field, which
resulted in only slightly increased activity in the right occipital pole. Both RD 4 and RD 5
reported seeing more light in the central visual field, which resulted in more significant
metabolic increase in the right occipital pole. When the subtraction images are grouped
for SPM analysis, the same occipital region may not reach statistical significance because
177
of these inconsistencies. Similar argument can be made for a region in the left anterior
calcarine cortex.
Light stimulation resulted in stronger activation in bilateral secondary visual
cortex (BA 19), right inferior temporal gyrus (BA 20), and left inferior prefrontal cortex
(BA 47) in the group of RD subjects than normal controls.
Figure 80: Light stimulation subtraction metablic image comparison. Blue circles in SPM
overlay images enclose primary visual cortex (BA 17) activation.
178
Figure 81: Light stimulation comparison SPM – Normal controls > RD subjects. Selected
Brodmann areas that exhibit more activation in normal controls than RD subjects during light
stimulation are labeled in green.
179
Figure 82: Light stimulation comparison SPM - RD subjects > Normal controls. Selected
Brodmann areas that exhibit more activation in RD subjects than normal controls during light
stimulation are labeled in green.
180
Figure 83: Light stimulation comparison 3D SPM projection. Selected Brodmann areas are
labeled in yellow.
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
Normal Control > RD Subjects (Absolute Subtraction)
Calcarine cortex 17 R 20 -102 -10 1.91 0.045
Normal Control > RD Subjects (Relative Subtraction)
Calcarine cortex 17 L 22 -98 -2 2.23 0.027
17 L -18 -70 8 2.07 0.035
17 R 20 -98 -4 2.34 0.022
RD Subjects > Normal Control (Absolute Subtraction)
Inferior prefrontal gyrus 47 L -36 54 -14 2.35 0.023
Inferior temporal gyrus 20 R 38 12 -42 2.62 0.015
RD Subjects > Normal Control (Relative Subtraction)
Inferior occipital gyrus 19 L -48 -70 -18 2.90 0.010
Inferior prefrontal gyrus 47 L -34 58 -8 2.56 0.017
Inferior temporal gyrus 20 R 38 0 -46 3.64 0.003
Lingual gyrus 19 R 42 -78 -22 2.53 0.017
Table 14: Light stimulation normal controls vs. RD subjects SPM summary.
181
9.3.3 TcES (ERG-Jet) Cross-Comparison
Cross-comparison analysis is performed using two-sample t-test method. Metabolic
image from TcES (ERG-Jet) study is first subtracted from the baseline image to yield
quantitative subtraction image for each subject in the two study groups. These subtraction
images are then inputted into SPM for statistical two-sample comparisons.
Results from TcES (ERG-Jet) cross-comparison SPM analysis between the
normal controls and the RD subjects are shown in Figures 84 – 86, and summarized in
Table 15. Absolute analysis reveals a region in the left anterior calcarine cortex (BA 17)
that shows more activation in the normal control than the RD group. This finding is
consistent with individual group analysis (Section 9.1.5 and 9.2.3) in which the normal
controls exhibit a larger activation locus in this region of the calcarine cortex than the RD
subjects. Comparing to the RD subjects, normal controls also demonstrate more
statistically significant activation in bilateral secondary visual cortex (BA 18), left
secondary visual cortex (BA 19), left fusiform gyrus (BA 37), left supramarginal gyrus
(BA 40), left primary motor cortex (BA 4), left somatosensory cortex (BA 3,2,1), left
supplementary motor cortex (BA 6), and left frontal eye field (BA 8). Relative analysis
yields similar regions that have more activation in the normal than the RD group. The
finding demonstrates stronger activation in the visual cortex (BA 17, 18, 19), visual
attention and saccade-related brain areas during TcES in normal than RD subjects.
TcES (ERG-Jet) in RD subjects does not yield any brain area that exhibited
statistically stronger activation than normal controls (Figure 85 and 86).
182
Figure 84: TcES (ERG-Jet) comparison SPM – Normal controls > RD subjects. Selected
Brodmann areas that exhibit more activation in normal controls than RD subjects during TcES
(ERG-Jet) are labeled in green.
183
Figure 85: TcES (ERG-Jet) comparison SPM - RD subjects > Normal controls. No brain
region exhibits significant activation in RD subjects than normal controls during TcES (ERG-Jet).
184
Figure 86: TcES (ERG-Jet) comparison 3D SPM projection. Selected Brodmann areas are
labeled in yellow.
185
Brain Region Brodmann
Area
Side Coordinates [mm]
x y z
T p-value
(uncorrected)
Normal Control > RD Subjects (Absolute Subtraction; p < 0.05)
Calcarine cortex 17 L -24 -74 10 2.14 0.032
Dorsolateral prefrontal
cortex
9 R 38 34 42 2.03 0.038
Frontal eye field 8 L -28 32 54 2.70 0.013
Fusiform gyrus 37 L -62 -52 -16 2.55 0.017
Inferior occipital gyrus 18 L -18 -102 -6 2.68 0.014
18 R 24 -100 -10 2.09 0.035
Inferior temporal gyrus 20 L -48 -2 -46 2.29 0.025
Middle occipital gyrus 19 L -46 -82 0 2.28 0.026
Postcental gyrus 3,2,1 L -20 -38 50 2.45 0.020
Precentral gyrus 4 L -38 -26 66 2.42 0.021
Subcentral area 43 L -64 0 22 2.36 0.023
Superior parietal gyrus 7 L -28 -66 62 2.21 0.029
Supplementary Motor 6 L -32 10 52 2.07 0.035
Supramarginal gyrus 40 L -58 -50 44 2.55 0.017
Normal Control > RD Subjects (Relative Subtraction; p < 0.05)
Calcarine cortex 17 L -16 -62 10 2.19 0.030
Inferior occipital gyrus 18 L -30 -96 -10 2.93 0.009
Middle occipital gyrus 18 L -20 -102 2 2.43 0.020
19 L -44 -84 -2 2.40 0.021
Supramarginal gyrus 40 L -58 -50 44 2.62 0.015
RD Subjects > Normal Control (Absolute Subtraction; p <0.05)
None
RD Subjects > Normal Control (Relative Subtraction; p < 0.05)
None
Table 15: TcES (ERG-JET) normal controls vs. RD subjects SPM summary.
9.3.4 Higher-level Visual Area Comparison
A close comparison of brain activation maps between normal controls and RD subjects
during both light stimulation and TcES (ERG-Jet) reveals a consistent difference in the
activation of higher-level visual processing areas. As shown in Figure 87, there are
186
primarily three areas that exhibited this difference: left inferior temporal gyrus (BA 20),
left fusiform gyrus (BA 37), and left parahippocampal gyrus (BA 27). Both inferior
temporal gyrus and fusiform gyrus reside in the temporal lobe and process visual
information that flows along the ventral stream. They have been shown to participate in
visual perception of shape, form and pattern [35, 244]. The parahippocampal gyrus has a
function in visual perception and memory [244].
Strong activation in all three regions is seen in the normal control group during
both light stimulation and TcES. In fact, for TcES ERG-Jet, all three areas show
statistically significant increase in activity compared to baseline, while parahippocampal
gyrus and inferior temporal gyrus show activation that yielded statistical significance
(p<0.05) during light stimulation. The same areas are not shown to be significantly
activated (p<0.05) during both light stimulation and TcES in the RD group. It can be
argued that SPM analysis does not achieve statistical significance in these regions for the
RD subjects, even though these regions do show small, but consistent activation. To
dispel any ambiguity, a quantitative comparison of activation in the three regions is
conducted for both subject groups, and the finding is shown in Table 16. Individual
values along with group average for each brain region and stimulation condition are
displayed for comparison. The values in the table represent the average increase in
metabolic activity from a region of interest drawn around the three brain areas based on
the light stimulation activation map for the normal controls, and they are derived by
subtracting individual subject metabolic PET image for each stimulation condition from
the baseline. A red star next to the group mean indicates the average increase in
187
activation is signifantly different from a mean of zero at p < 0.05. The normal controls
demonstrate consistent positive increase in metabolic activity in all three brain regions
during light stimulation and TcES (ERG-Jet). The increases in both individual and mean
values in the three brain regions are greater for TcES than light stimulation, possibly due
to increased attention placed on observing the phosphene. For the RD group, the pattern
is not consistent and there is significant fluctuation in the values; in fact, some values are
negative, indicating decreases in metabolic activity compare to baseline. The average
activation across all RD subjects does not yield statistical significance in all three higher-
level visual areas.
The quantitative analysis confirms the finding that significant increase in
metabolic activity in the three higher-level visual areas occurs only in the normal
controls, but not the RD subjects. Since these visual areas are intimately involved in
processing of complex visual detail, it is possible that light-perception-only input from
the retina of the RD subjects may not have enough spatial detail to significantly activate
these regions. Based on previous imaging studies addressing visual plasticity, it is also
possible that chronic, severe visual impairment may induce plastic remodeling in higher-
order visual processing areas and render them less responsive to visual input.
188
Figure 87: Higher-level visual area comparison between normal and RD subjects. Selected
Brodmann areas are labeled in yellow. Note the absence of activation in higher-level visual areas
(BA 20, 27, and 37) for the RD subjects.
189
Table 16: Quantitative comparison of activation in higher-level visual areas. Values listed
have units of µmol (glucose)/minute/gram tissue. For each Brodmann area, the increase in
metabolic rate of glucose is listed for each individual in the group along with mean and standard
deviation. NC: normal control group; RD: retinal degeneration group; ‘*’ indicates the group
average is statistically different from a mean of zero (at p<0.05).
190
9.4 Results Summary
9.4.1 Normal Controls
Comparing to baseline, light stimulation of the right eye in the normal controls resulted in
activation of primary and association visual cortex in both hemispheres. Activation of the
primary visual cortex is found both above and below the calcarine sulcus at the occipital
pole on both hemispheres, which has been shown to process visual input from the macula
region of the retina. The light stimulus is a flashing square that subtended central 10°
field of view of the right eye; consequently we expect activation of the primary visual
cortex at the occipital pole that receives projection from the macula. Activation in the
primary visual cortex was more significant on the right than the left hemisphere. In
addition, there was a very significant region of activation in the anterior calcarine cortex
on the left hemisphere above the calcarine sulcus. Activation in this region of the
occipital cortex is consistent with findings from studies that looked at monocular photic
stimulation of the right eye in humans [158, 159, 209, 245], and this phenomenon may be
explained by nasotemporal asymmetry in photoreceptor and ganglion cell distribution
(see Section 3.5.2). In addition to primary visual cortex activation, association visual
areas (BA 18 and 19) are also activated, with stronger activation more towards the left
hemisphere.
TcES (DTL-Plus) resulted in activation of mainly bilateral association visual
cortex (BA 18) and an area of left association visual area (BA 19). Similar to the pattern
of light stimulation, extrastriate visual cortex exhibited significantly stronger activation
191
on the left hemisphere. Studies using fMRI have shown that when light stimulus was
presented in one half of the visual field, activation in extrastriate visual cortex occurred
bilaterally, with weaker activation on the same side of the brain as the light stimulus [87,
88, 173]. The subjective light sensation in the upper right visual field of the right eye
during TcES (DTL-Plus) is analogous to projecting light stimulus onto the inferior nasal
retina. Since output from the nasal retina of the right eye predominantly crosses onto the
left hemisphere, more extensive activation was observed in the contralateral (left) visual
cortex. TcES (DTL-Plus) did not achieve primary visual cortex activation (p < 0.05);
possibly because the peripherally located faint phosphene was suboptimal for eliciting
activity in the primary visual cortex. The phosphene sensation during TcES (DTL-Plus)
was consistently reported to be in the upper right visual field, which resulted from
electrical stimulation of inferior nasal retina as corroborated by TcES modeling.
Activation was confined to visual cortex below the calcarine fissure which process input
from the upper visual field.
TcES (ERG-Jet) resulted in activation of primary visual area (BA 17) in the left
anterior calcarine cortex which receives visual input from the nasal hemiretina of the
right eye. Normal controls reported seeing phosphene in the right peripheral visual field
of the right eye during TcES using ERG-Jet, and this subjective phosphene sensation is
corroborated by the modeling results showing preferential stimulation of retinal ganglion
cell axon fibers along the peripheral nasal retina of the right eye. In addition to primary
visual cortex activation, there was also bilateral association visual cortex (BA 18)
192
activation, with activation extending significantly into association visual area (BA 19)
and higher-level visual processing areas (BA 20 and 37) on the left hemisphere.
Light stimulation and TcES (DTL-Plus and ERG-Jet) also resulted in activation in
ventromedial frontal lobe (including BA 10, 11), dorsolateral prefrontal cortex (including
BA 46, 47) and anterolateral prefrontal cortex (including BA 9, 46). These areas play a
significant role in visual attention and fixation.
9.4.2 Retinal Degeneration Subjects
Light stimulation in the group of RD subjects activated primary visual cortex (BA 17) in
the right occipital pole and association visual cortex (BA 18 and 19) on the left
hemisphere. Comparing to normal controls, a much smaller region of activation in the
primary visual cortex reached statistical significance in the RD group. This can be
explained by differences in the extent and location of visual loss between the RD
subjects. TcES using ERG-Jet activated a region of anterior calcarine cortex (BA 17) in
the left hemisphere that receives visual input from the peripheral nasal retina of the right
eye. This activation result is consistent with the sensation of peripheral temporal
phosphene in the right eye reported by the RD subjects. There was also bilateral
activation in association visual cortex (BA 18) and a region of left association visual
cortex (BA 19) near the parieto-occipital junction.
Similar to normal controls, both light stimulation and TcES (ERG-Jet) resulted in
activation in areas of prefrontal cortex in the RD subjects, reflecting the involvement of
visual attention during the stimulation studies.
193
9.4.3 Normal and Retinal Degeneration Group Comparison
Graphical comparison of activation regions between normal controls and RD subjects
under the two stimulation conditions is shown in Figure 88, and the corresponding
activated Brodmann areas are summarized in Table 17. Light stimulation and TcES
(ERG-Jet) activated areas involved in primary vision and visual attention in both normal
and RD groups. In addition, TcES (ERG-Jet) invoked additional activation in the frontal
eye field (BA 8) which is involved in saccade, and sensorimotor areas which reflect
sensation of the ERG-Jet contact lens on the cornea as well as motor control exerted by
the subjects to keep the electrode in place.
The main difference between normal and RD groups lies in the activation pattern
of higher-level visual processing areas. Visual areas involved in form vision and visual
memory are consistently activated in normal controls during both light stimulation and
TcES (ERG-Jet), whereas activation in the same areas does not reach statistical
significance in RD subjects. One additional difference is the activation of emotion-related
brains areas (BA 38 and insular cortex) for RD subjects during both light stimulation and
TcES (ERG-Jet), which is not observed in the normal control group. It is possible that
visual perception evokes a strong emotional response in subjects with severe visual loss.
194
Figure 88: 3D SPM activation comparison between normal and RD subjects. The top two
rows of 3D SPM projections show cortical areas that exhibit more activation during light
stimulation compared to baseline in normal (row 1) and RD subjects (row 2); The bottom two
rows of 3D SPM projections show cortical areas that exhibit more activation during TcES (ERG-
Jet) compared to baseline in normal (row 3) and RD subjects (row 4). Selected Brodmann areas
are labeled in yellow.
195
Table 17: Comparison of activated brain regions between normal and RD subjects. Numbers
represent Brodmann areas. NC: normal controls; RD: retinal degeneration subjects.
196
Chapter 10: Conclusion
The results presented in this thesis demonstrate the feasibility of using transcorneal
electrical stimulation method to elicit visual cortex response in both normal controls and
subjects with retinal degeneration. We have shown activation of primary and association
visual cortex in RD subjects during both light stimulation and TcES. Transcorneal
electrical stimulation with ERG-Jet electrode on the right eye resulted in consistent
phosphene sensation in the peripheral temporal visual field in both the normal controls
and RD subjects. This subjective phosphene sensation is corroborated by TcES modeling
which demonstrates preferential stimulation along the peripheral nasal hemiretina in the
right eye. TcES (ERG-Jet) resulted in retinotopically-matched primary visual area
activation in the anterior calcarine cortex on the left hemisphere, which receives visual
input from the nasal hemiretina of the right eye. We found less consistent pattern of
primary visual cortex activation from light stimulation between the normal and RD
groups, which can be explained by varying extent and location of visual loss between the
RD subjects. Higher-level visual processing areas involed in object vision and visual
memory are not significantly activated in RD subjects during both light stimulation and
TcES. It is possible that bare light-perception input from the retina may not offer enough
spatial detail to significantly activate these regions that are more involved in complex
visual processing. Another possibility is that chronic, severe visual impairment may
induce plasticity in higher-order visual processing areas and render them less responsive
to visual input.
197
10.1 Contributions
We have shown that electrical stimulation of the retina in both normal and RD subjects
elicit similar pattern of visual cortex activation; this result demonstrates the preservation
of retinotopic map in the primary visual cortex despite chronic, severe visual loss. The
study lends additional proof to the idea that the neuronal network between the retina and
the visual cortex can remain intact following severe outer retinal degeneration, an
important premise for the validity of therapeutic intervention using stem-cell and
artificial retinal prostheses. Knowledge gained from this study can serve as a basis for
future functional imaging assessment of retinal prosthesis functionality in implant
subjects.
10.2 Suggestions for Future Work
To extend along this research work, fMRI can be adapted to study brain response during
transcorneal electrical stimulation. The low invasiveness, lack of radiation exposure, and
higher spatial and temporal resolution makes fMRI a superior imaging modality over
PET to map functional activation in the brain in subjects with retinal degeneration. Brain
activation SPM maps not only can be achieved at higher spatial resolution than PET, it
can be potentially completed in one time setting for a single individual. Since the
progression and severity of retinal dystrophy vary between RD patients, fMRI studies can
be individually tailored to achieve more accurate visual cortex activation map.
Challenges to make the corneal electrode and the wiring MRI-compatible need to be
198
overcome, but can be accomplished with currently available technology. Since the
electrode is placed on the eye and away from the brain tissue, any artifact in the brain
introduced by the effect of the magnetic field on the electrode will be kept to a minimum.
With the advent of new hybrid PET-MRI scanner, it is possible to capture
functional PET data and superimpose it on top of high-resolution morphological MRI
images all in one scanning session. This will allow us not only to achieve greater
accuracy in functional mapping of brain activation, but also directly compare the results
between PET and fMRI.
In addition, diffusion tensor MRI methodology can be used to measure anatomical
connection between the eye and the visual cortex. It will allow us to study how different
forms of retinal degeneration disrupt the normal organization and integrity of the visual
pathway over time.
Ultimately we would like to measure brain response to direct retinal electrical
stimulation in vivo in subjects who have been implanted with a retinal prosthesis. This
allows us to have control over the location and size of retinal stimulation, and look for
corresponding activation in the visual cortex. This approach will directly evaluate the
functionality of retinal prosthesis in vivo and potentially open new doorways to enrich
our understanding about brain plasticity.
199
Bibliography
[1] Alavi A and Reivich M. The conception of FEG-PET imaging.
[http://tracerkinetics.engr.iupui.edu/HistoricalNotes/TheConceptionofFDGAAlavi.htm ]
[2] Anderson TJ, Jenkins IH, Brooks DJ, Hawken MB, Frackowiak RSJ, Kennard C. (1994)
Cortical control of saccades and fixation in man A PET study. Brain 117:1073-1084.
[3] App E, Debus G. (1998) Saccadic velocity and activation: development of a diagnostic tool
for assessing energy regulation. Ergonomics 41:689-697.
[4] Armel KC and Ramachandran VS. (1999) Acquired synesthesia in retinitis pigmentosa.
Neurocase 5:293-296.
[5] Armitage DW, LeVeen HH, Pethig R. (1983) Radiofrequency-induced hyperthermia:
computer simulation of specific absorption rate distributions using realistic anatomical
models. Phys. Med. Biol. 28:31-42.
[6] Arno P, De Volder AG, Vanlierde A, Wanet-Defalque MC, Streel E, Robert A, Sanabria-
Bohorquez S, and Veraart C. (2001) Occipital activation by pattern recognition in the early
blind using auditory substitution for vision. NeuroImage 13:632–645.
[7] Astafiev SV, Shulman GL, Stanley CM, Snyder AZ, Van Essen DC, Corbetta M. (2003)
Functional organization of human intraparietal and frontal cortex for attending, looking,
and pointing. J. Neurosci. 23:4689– 4699.
[8] Baker CI, Peli E, Knouf N, Kanwisher NG. (2005) Reorganization of visual processing in
macular degeneration. The Journal of Neuroscience 25(3):614-618.
[9] Bartels A, Zeki S. (2000) The architecture of the colour centre in the human visual brain,
new results and a review. Eur. J. Neurosci. 12:172– 193.
[10] Bavelier D, Neville HJ. (2002) Cross-modal plasticity: Where and how? Nat. Rev. Neurosci.
3:443–452.
[11] Bear MF, Connors B, and Paradiso M. (2006) Neuroscience: Exploring the Brain.
Lippincott Williams & Wilkins.
[12] Beauchamp MS, Petit L, Ellmore TM, Ingeholm J, Haxby JV. (2001) A parametric fMRI
study of overt and covert shifts of visuospatial attention. NeuroImage 14:310– 321.
[13] Beckers G, Zeki S. (1995) The consequences of inactivating areas V1 and V5 on visual
motion perception. Brain 118:49-60.
[14] Berardi N, Pizzorusso T, Ratto GM, and Maffei L. (2003) Molecular basis of plasticity in
the visual cortex. Trends Neurosci. 26:369-378.
200
[15] Berson EL. (1993) Retinitis pigmentosa. The Friedenwald Lecture. Invest. Ophthalmol. Vis.
Sci. 34(5):1659–1676.
[16] Berthezene Y, Truy E, Morgon A, Giard MH, Hermier M, Franconi JM, Froment JC. (1997)
Auditory cortex activation in deaf subjects during cochlear electrical stimulation.
Evaluation by functional magnetic resonance imaging. Invest. Radiol. 32:297-301.
[17] Boynton GM, Engel SA, Glover GH, Heeger DJ. (1996) Linear systems analysis of
functional magnetic resonance imaging in human V1. J. Neurosci. 16(13):4207–4221.
[18] Brian A. Wandell. (1999) Computational neuroimaging of human visual cortex. Annu. Rev.
Neurosci. 22:145-173.
[19] Brindley GS, Lewin WS. (1968) The sensations produced by electrical stimulation of the
visual cortex. J. Physiol. 196:479–93.
[20] Britten KH and Heuer HW. (1999) Spatial summation in the receptive fields of MT neurons.
J. Neurosci. 19:5074–5084.
[21] Burkhalter A, Felleman DJ, Newsome WT, Van Essen DC. (1986) Anatomical and
physiological asymmetries related to visual areas V3 and VP in macaque extrastriate
cortex. Vis. Res. 26(1):63–80.
[22] Burkhalter A, Bernardo KL. (1989) Organization of corticocortical connections in human
visual cortex. Proc. Natl. Acad. Sci. USA 86:1071-1075.
[23] Burton H, Sinclair RJ, McLaren DG. (2004) Cortical activity to vibrotactile stimulation: an
fMRI study in blind and sighted individuals. Hum. Brain Mapp. 23:210–228.
[24] Carman G, Drury H, Van Essen D. (1995) Computational methods for reconstructing and
unfolding the cerebral cortex. Cereb. Cortex 5(6):506–517.
[25] Chaurasia BD, Mathur BB. (1976) Eyedness. Acta Anat (Basel) 96(2):301-305.
[26] Chavis PS, Al-Hazmi A, Clunie D, Hoyt WF. (1997) Temporal crescent syndrome with
magnetic resonance correlation. J. Neuroophthalmol. 17:151-155.
[27] Cherry SR, Sorenson JA, Phelps ME. (2003) Physics in Nuclear Medicine, 3
rd
ed. Saunders.
[28] Chow AY, Peachey NS. (1998) The subretinal microphotodiode array retinal prosthesis.
Ophthalmic Res. 30:195-198.
[29] Chow AY, Pardue MT, Chow VY, et al. (2001) Implantation of silicon chip
microphotodiode arrays into the cat subretinal space. IEEE Trans. Neural Syst. Rehabil.
Eng. 9(1):86-95.
[30] Chow AY, Pardue MT, Perlman JL, Ball SL, Chow VY, Hetling JR, Peyman GA et al.
(2002) Subretinal implantation of semiconductor-based photodiodes: durability of novel
implant designs. J. Rehabil. Res. Dev. 39(3):313-321.
201
[31] Chow AY, Chow VY, Packo KH, et al. (2004) The artificial silicon retina microchip for the
treatment of vision loss from retinitis pigmentosa. Arch Ophthalmol. 122(4):460-469.
[32] Churchland MM, Priebe NJ, Lisberger SG. (2005) Comparison of the spatial limits on
direction selectivity in visual areas MT and V1. J. Neurophysiol. 93:1235-1245.
[33] Clarke S, Miklossy J. (1990) Occipital cortex in man: organization of callosal connections,
related myelo- and cytoarchitecture, and putative boundaries of functional visual areas. J.
Comp. Neurol. 298:188-214.
[34] Cohen LG, Celnick P, Pascual-Leone A, Corwell B, Faiz L, Dambrosia J, et al. (1997)
Functional relevance of cross-modal plasticity in blind humans. Nature 389:180–183.
[35] Corbetta M, Miezin FM, Dobmeyer S, Shulman GL, Petersen SE. (1991) Selective and
divided attention during visual discriminations of shape, color, and speed: functional
anatomy by positron emission tomography. The Journal of Neuroscience 11(8):2383-2402.
[36] Corbetta M, Akbudak E, Conturo TE, Snyder AZ, Ollinger JM, Drury HA, Linenweber MR,
Petersen SE, Raichle ME, Van Essen DC, Shulman GL. (1998) A common network of
functional areas for attention and eye movements. Neuron 21:761– 773.
[37] Courtney SM, Ungerleider LG. (1997) What fMRI has taught us about human vision. Curr.
Opin. Neurobiol. 7:554-561.
[38] Curcio C, Sloan KR, Kalina RE, Hendrickson AE. (1990a) Human photoreceptor
topography. J. Comp. Neurol. 300:497-523.
[39] Curcio C, Allen KA. (1990b) Topography of ganglion cells in human retina. J. Comp.
Neurol. 300:5-25.
[40] Curcio CA, Medeiros NE, Millican CL. (1996) Photoreceptor loss in age-related macular
degeneration. Invest. Ophthalmol. Vis. Sci. 37(7):1236–1249.
[41] Dale AM, Sereno MI. (1993) Improved localization of cortical activity by combining EEG
and MEG with MRI cortical surface reconstruction: a linear approach. J. Cogn. Neurosci.
5:162–176.
[42] Daw NW. (1995) Visual development. New York: Plenum Press. p.228
[43] Dawson WW, Trick GL, Litzkow CA. (1979) Improved electrode for electroretinography.
Invest. Ophthalmol. Vis. Sci. 18(9):988-991.
[44] Deans SR. (1983) The Radon Transform and Some of its Applications. New York: Wiley.
[45] De Leo D, Hickey PA, Meneghel G, Cantor CH. (1999) Blindness, fear of sight loss, and
suicide. Psychosomatics 40:339–44.
202
[46] De Volder AG, Catalan-Ahumada M, Robert A, Bol A, Labar D, Coppens A, Michel C,
Veraart C. (1999) Changes in occipital cortex activity in early blind humans using a
sensory substitution device. Brain Research 826:128–134.
[47] Delbeke J, Pins D, Michaux G, Wanet-Defalque MC, Parrini S, Veraart C. (2001) Electrical
stimulation of anterior visual pathways in retinitis pigmentosa. Invest. Ophthalmol. Vis. Sci.
42:291–297.
[48] Delbeke J, Wanet-Defalque MC, Gerard B et al. (2002) The microsystems based visual
prosthesis for optic nerve stimulation. Artif. Organs 26(3):232-234.
[49] Delbeke J, Oozeer M, Veraart C. (2003) Position, size and luminosity of phosphenes
generated by direct optic nerve stimulation. Vision Res. 43(9):1091-1102.
[50] Desimone R, Schein SJ. (1987) Visual properties of neurons in area V4 of the macaque:
sensitivity to stimulus form. J. Neurophysiol. 57:835-868.
[51] DeYoe EA, Carman GJ, Bandettini P, Glickman S, Wieser J, Cox R, Miller D, Neitz J.
(1996) Mapping striate and extrastriate visual areas in human cerebral cortex. Proc. Natl.
Acad. Sci. USA 93:2382– 2386.
[52] Dobelle WH. (2000) Artificial vision for the blind by connecting a television camera to the
visual cortex. ASAIO J. 46:3–9.
[53] Dorfman LJ, Gaynon M, Ceranski J, Louis AA and Howard JE. (1987) Visual electrical
evoked potentials: evaluation of ocular injuries. Journal of Neurology 37:123–128.
[54] Dougherty RF, Koch VM, Brewer AA, Fischer B, Modersitzki J, Wandell BA. (2003)
Visual field representations and locations of visual areas V1/2/3 in human visual cortex.
Journal of Vision 3(10):586-598.
[55] Dreher B, Burke W, and Calford MB. (2001) Cortical plasticity revealed by circumscribed
retinal lesions or artificial scotomas. Prog. Brain Res. 134:217-246.
[56] Drury HA, Van Essen DC, Anderson CH, Lee CW, Coogan TA, Lewis JW. (1996)
Computerized mappings of the cerebral cortex: a multiresolution flattening method and a
surfacebased coordinate system. J. Cogn. Neurosci. 8:1–28.
[57] Dubner R and Zeki SM. (1971) Response properties and receptive fields of cells in an
anatomically defined region of the superior temporal sulcus in the monkey. Brain Res.
35:528–532.
[58] Duffy CJ, Wurtz RH. (1995) Response of monkey MST neurons to optic flow stimuli with
shifted centers of motion. J. Neurosci. 15:5192-5208.
[59] Easty LE, Sparrow JM. (1999) Oxford textbook of ophthalmology. Oxford University Press,
Oxford, UK.
203
[60] Eberdt M, Brown PK, Lazzi G. (2003) Two-dimensional SPICE-linked multiresolution
impedance method for low-frequency electromagnetic interactions. IEEE Trans. on
Biomedical Engineering 50(7):881-889.
[61] Ehrenstein WH, Arnold-Schulz-Gahmen BE, Jaschinski W. (2005) Eye preference within
the context of binocular functions. Graefes Arch. Clin. Exp. Ophthalmol. 243(9):926-932.
[62] Elfar SD, Cottaris NP, Iezzi R, Abrams GW. (2004) Rapid mapping of cortical multi-
electrode arrays and its application for the evaluation of retinal prostheses. Invest.
Ophthalmol. Vis. Sci. 45:3403.
[63] Engel SA, Rumelhart DE, Wandell BA, et al. (1994) fMRI of human visual cortex. Nature
369:525.
[64] Engel SA, Glover GH, Wandell BA. (1997a) Retinotopic organization in human visual
cortex and the spatial precision of functional MRI. Cereb. Cortex 7(2):181–192.
[65] Fahle M, Schmid M. (1988) Naso-temporal asymmetry of visual perception and of the
visual cortex. Vision Res. 28:293-300.
[66] Fariss RN, Li ZY, and Milam AH. (2000) Abnormalities in rod photoreceptors, amacrine
cells, and horizontal cells in human retinas with retinitis pigmentosa. Am. J. Ophthalmol.
129:215-223.
[67] Felleman DJ, Van Essen DC. (1991) Distributed hierarchical processing in the primate
cerebral cortex. Cereb. Cortex 1:1–47.
[68] Felleman DJ, Burkhalter A, Van Essen DC. (1997) Cortical connections of areas V3 and VP
of macaque extrastriate visual-cortex. J. Comp. Neuro. 379(1):21–47.
[69] Ffytche DH, Guy CN, Zeki S. (1995) The parallel visul motion inputs into areas V1 and V5
of human cerebral cortex. Brain 118:1375-1394.
[70] Fine I, Wade AR, Brewer AA, May MG, Goodman DF, et al. (2003) Longterm deprivation
affects visual perception and cortex. Nat. Neurosci. 6:915–916.
[71] Finney EM, Fine I, and Dobkins KR. (2001) Visual stimuli activate auditory cortex in the
deaf. Nat. Neurosci. 4:1171-1173.
[72] Fox PT, Raichle ME. (1984) Stimulus rate dependence of regional cerebral blood flow in
human striate cortex, demonstrated by positron emission tomography. Journal of
Neurophysiology 51(5):1109-1120.
[73] Frackowiak RSJ, Friston KJ, Frith C, Dolan R, Price CJ, Zeki S, Ashburner J, and Penny
WD. (2003) Human Brain Function, 2
nd
edition, Chapter 1, Figure 1. Academic Press.
[74] Frackowiak RSJ, Friston KJ, Frith C, Dolan R, Price CJ, Zeki S, Ashburner J, and Penny
WD. (2003) Human Brain Function, 2
nd
edition, Chapter 1, pp. 25. Academic Press.
204
[75] Friston KJ, Frith CD, Liddle PF, and Frackowiak RSJ. (1991) Comparing functional (PET)
images: the assessment of significant change. J. Cereb. Blood Flow Metab. 11:690-699.
[76] Friston KJ, Worsley KJ, Frackowiak RSJ, Mazziotta JC, and Evans AC. (1994a) Assessing
the significance of focal activations using their spatial extent. Hum. Brain Mapp. 1:214-
220.
[77] Friston KJ, Ashburner J, Frith CD, Poline JB, Heather JD, and Frackowiak RSJ. (1995a)
Spatial registration and normalization of images. Hum. Brain Mapp. 2:165-189.
[78] Friston KJ, Holmes AP, Worsley KJ, Poline JB, Frith CD, and Frackowiak RSJ. (1995b)
Statistical Parametric Maps in functional imaging: A general linear approach. Hum. Brain
Mapp. 2:189-210.
[79] Friston KJ, Williams S, Howard R, Frackowiak RS, and Turner R. (1996) Movement-
related effects in fMRI time-series. Magn. Reson. Med., 35:346-355.
[80] Friston KJ, Ashburner JT, Kiebel SJ, Nichols TE and Penny WD. (2006) “Statistical
Parametric Mapping: The Analysis of Functional Brain Images” 1st edition. Chapter 18:
Random field theory, pp. 232. Academic Press.
[81] Fujikado T, Morimoto T, Matsushita K, Shimojo H, Okawa Y and Tano Y. (2006) Effect of
transcorneal electrical stimulation in patients with nonarteritic ischemic optic neuropathy or
traumatic optic neuropathy. Japanese Journal of Ophthalmology 50(3):266-273.
[82] Fujikado T, Morimoto T, Kanda H, Kusaka S, Nakauchi K, et al. (2007) Evaluation of
phosphenes elicited by extraocular stimulation in normals and by suprachoroidal-
transretinal stimulation in patients with retinitis pigmentosa. Graefe’s Arch. Clin. Exp.
Ophthalmol. 245:1411-1419.
[83] Fukuda Y, Sawai H, Watanabe M, Wakakuwa K and Morigiwa K. (1989) Nasotemporal
overlap of crossed and uncrossed retinal ganglion cell projections in the Japanese monkey
(Macaca fuscata). The Journal of Neuroscience 9(7):2353-2373.
[84] Gekeler F, Messias A, Ottinger M, Bartz-Schmidt KU and Zrenner E. (2006) Phosphenes
electrically evoked with DTL electrodes: a study in patients with retinitis pigmentosa,
glaucoma, and homonymous visual field loss and normal subjects. Investigative
Ophthalmology & Visual Science 47(11):4966-4974.
[85] Giraud AL, Truy E, Frackowiak R. (2001) Imaging plasticity in cochlear implant patients.
Audiol. Neurootol. 6:381-393.
[86] Gizewski ER, Gasser T, de Greiff A, Boehm A, Forsting M. (2003) Cross-modal plasticity
for sensory and motor activation patterns in blind subjects. Neuroimage 19:968–975.
[87] Goebel R, Muckli L, Zanella FE, Singer W, Stoerig P. (2001) Sustained extrastriate cortical
activation without visual awareness revealed by fMRI studies of hemianopic patients.
Vision Research 41(10-11):1459-1474.
205
[88] Gonzalez F, Relova JL, Prieto A, Peleteiro M, Romero MC. (2006) Hemifield dependence
of responses to color in human fusiform gyrus. Vision Research 46(16):2499-2504.
[89] Goodale & Milner (1992) Separate pathways for perception and action. Trends in
Neuroscience 15:20-25.
[90] Gougoux F, Lepore F, Lassonde M, Voss P, Zatorre RJ, Belin P. (2004) Neuropsychology:
Pitch discrimination in the early blind. Nature 430:309.
[91] Gougoux F, Zatorre RJ, Lassonde M, Voss P, Lepore F. (2005) A functional neuroimaging
study of sound localization: visual cortex activity predicts performance in early-blind
individuals. PLoS. Biol. 3:324–333.
[92] Graziano M, Anderson R, Snowden R. (1994) Tuning of MST neurons to spiral motions. J.
Neurosci. 14:54-56.
[93] Grigsby SS, Tsou BH. (1994) Grating and flicker sensitivity in the near and far periphery:
Nasotemporal asymmetries and binocular summation. Vision Res. 34:2841-2848.
[94] Grumet AE, Wyatt JL Jr, Rizzo JF. (2000) Multielectrode stimulation and recording in the
isolated retina. J. Neurosci. Methods 101(1):31-42.
[95] Guenther E, Troger B, Schlosshauer B, Zrenner E. (1999) Long-term survival of retinal cell
cutures on retinal implant materials. Vision Res. 39(24):3988-3994.
[96] de Haan B, Morgan PS, Rorden C. (2008) Covert orienting of attention and overt eye
movements activate identical brain regions. Brain Res. 1204:102-111.
[97] Haines DE. (2000) Neuroanatomy: an atlas of structures, sections, and systems. 5
th
edition.
Lippincott William & Wilkins.
[98] Hamacher K, Coenen HH, Stocklin G. (1986) Efficient stereospecific synthesis of no carrier
added 1-
18
F-Fluoro-2-deoxy-D-glucose. J. Nucl. Med. 27:235-238.
[99] Hambrecht FT. (1995) Visual prostheses based on direct interfaces with the visual system.
Baillieres Clin. Neurol. 4(1):147-165.
[100] Hamilton RH, Keenan JP, Catala M, Pascual-Leone A. (2000) Alexia for Braille following
bilateral occipital stroke in an early blind woman. Cogn. Neurosci. 11:237–240.
[101] Hari R, Pelizzone M, Makela JP, Hallstrom J, Huttunen J, Knuurila J. (1988)
Neuromagnetic responses from a deaf subject to stimuli presented through a multichannel
cochlear prosthesis. Ear Hear. 9:148-152.
[102] Hartong DT, Berson EL, Dryja TP. (2006) Retinitis Pigmentosa. Lancet 368(9549):1795-
1809.
206
[103] Haxby JV, Grady CL, Horwitz B, Ungerleider LG, Mishkin M, Carson RE, Herscovitch P,
Schapiro MB, Rapoport SI. (1991) Dissociation of object and spatial visual processing
pathways in human extrastriate cortex. Proc. Natl. Acad. Sci. USA 88:1621– 1625.
[104] Haxby JV, Horwitz B, Ungerleider LG, Maisog JM, Pietrini P, Grady CL. (1994) The
functional organization of human extrastriate cortex: a PET-rCBF study of selective
attention to faces and locations. The Journal of Neuroscience 14(11):6336-6353.
[105] Heckenlively JR, Boughman J, Friedman L. (1988) Diagnosis and classification of retinitis
pigmentosa. In Retinitis Pigmentosa, ed. JR Heckenlively, p.21. Philadelphia, PA:
Lippincott.
[106] Heckenlively JR, Arden GB. (2006) Principles and Practice of Clinical Electrophysiology
of Vision. 2
nd
Edition. Massachusetts Institute of Technology Press.
[107] Heide W, Binkofski F, Seitz RJ, Posse S, Nitschke MF, Freund HJ, Kömpf D. (2001)
Activation of frontoparietal cortices during memorized triple-step sequences of saccadic
eye movements: an fMRI study. Eur. J. Neurosci. 13:1177– 1189.
[108] Hensch TK. (2004) Critical period regulation. Annu. Rev. Neurosci. 27:549–579.
[109] Herzog H, Lamprecht A, Kuhn A, Roden W, Vosteen KH, Feinendegen LE. (1991)
Cortical activation in profoundly deaf patients during cochlear implant stimulation
demonstrated by H2(15)O PET. J. Comput. Assist. Tomogr. 15:369-375.
[110] Heywood CA, Gadotti A, Cowey A. (1992) Cortical area V4 and its role in the perception
of color. J. Neurosci. 12:4056-4065.
[111] Holmes G. (1918) Disturbances of vision by cerebral lesions. Br. J. Opthalmol. 2:353–384.
[112] Holmes G. (1944) The organization of the visual cortex in man. Proc. R. Soc. London Ser.
B 132:348–361.
[113] Horner PJ, Gage FH. (2000) Regenerating the damaged central nervous system. Nature
407: 963–70.
[114] Horton J, Hoyt W. (1991a) The representation of the visual field in human striate cortex.
Arch. Opthalmol. 109:816–824.
[115] Horton JC, Hoyt WF. (1991b) Quandrantic visual field defects: A hallmark of lesions in
extrastriate (V2/V3) cortex. Brain 114:1703-1718.
[116] Huang SC, Phelps ME, Hoffman EJ, et al. (1980) Noninvasive determination of local
cerebral metabolic rate of glucose in man. Am. J. Physiol. 238:E69-E82.
[117] Humayun MS, Propst R, de Juan E Jr, McCormick K, Hickingbotham D. (1994) Bipolar
surface electrical stimulation of the vertebrate retina. Arch. Ophthalmol. 112(1):110-116.
207
[118] Humayun MS, de Juan E, Dagnelie G, Greenberg RJ, Propst RH, Phillips DH. (1996)
Visual perception elicited by electrical stimulation of retina in blind humans. Arch.
Ophthalmol. 114:40–46.
[119] Humayun MS, Prince M, de Juan EJ, Barron Y, Moskowitz M, et al. (1999a) Morphometric
analysis of the extramacular retina from postmortem eyes with retinitis pigmentosa. Invest.
Ophthalmol. Vis. Sci. 40(1):143–148.
[120] Humayun MS, de Juan EJ, Weiland JD, Dagnelie G, Katona S, et al. (1999b) Pattern
electrical stimulation of the human retina. Vis. Res. 39:2569–2576.
[121] Humayun MS. (2001) Intraocular retinal prosthesis. Trans. Am. Ophthalmol. Soc. 99:271-
300.
[122] Ito J. (1993) Auditory cortex activities in severely hearing-impaired and cochlear implant
patients. Positron emission tomographic study. Adv. Otorhinolaryngol. 48:29-34.
[123] Jones KE, Campbell PK, Normann RA. (1992) A glass/silicon composite intracortical
electrode array. Ann. Biomed. Eng. 20(4):423-437.
[124] Jones BW, Watt CB, Frederick JM, Baehr W, Chen CK, Levine EM, Milam AH, Lavail
MM, and Marc RE. (2003) Retinal remodeling triggered by photoreceptor degenerations.
Journal of Comparative Neurology 464:1-16.
[125] Kastner S, Ungerleider LG. (2001) The neural basis of biased competition in human visual
cortex. Neuropsychologia 39:1263-1276.
[126] Kandel ER, Schwartz JH, Jessell TM. (2000) Principles of Neural Science, 4
th
Edition.
McGraw-Hill.
[127] Kanwisher N, Chun MM, McDermott J, Ledden PJ. (1996) Functional imaging of human
visual recognition. Cogn. Brain Res. 5:55– 67.
[128] Kawamura A. (1986) Study on characteristics of electrically evoked response (EER) in the
rabbit eyes. Journal of the Japanese Ophthalmology Society 90:1417–1421.
[129] Kim S, Sadda S, Pearlman J, Humayun MS, de Juan E Jr, et al. (2002) Morphometric
analysis of the macula in eyes with disciform age-related macular degeneration. Retina
22(4):471–477.
[130] Kinomura S, Larsson J, Gulyas B, Roland PE. (1996) Activation by attention of the human
reticular formation and thalamic intralaminar nuclei. Science 271(5248):512-515.
[131] Knighton RW. (1975) An electrically evoked slow potential of the frog’s retina. II.
Identification with PII component of electroretinogram. J. Neurophysiol. 38:198-209.
[132] Komatsu H, Ideura Y. (1993) Relationships between color, shape, and pattern selectivities
of neurons in the inferior temporal cortex of the monkey. J. Neurophysiol. 70:677-694.
208
[133] Krumpaszky HG, Klauss V. (1996) Epidemiology of blindness and eye disease.
Ophthalmologica 210:1–84.
[134] Kwong KK, Belliveau JW, Chesler DA, Goldberg IE,Weisskoff RM, et al. (1992) Dynamic
magnetic resonance imaging of human brain activity during primary sensory stimulation.
Proc. Natl. Acad. Sci. USA 89:5675–5679.
[135] Lai S, Hopkins AL, Haacke EM, Li D, WassermanBA, et al. (1993) Identification of
vascular structures as a major source of signal contrast in high resolution 2D and 3D
functional activation imaging of the motor cortex at 1.5T: preliminary results. Mag. Res.
Med. 30:387–392.
[136] Lambert V, Laloyaux C, Schmitt C, Gerard B, Delbeke J, Veraart C. (2003) Localization,
discrimination, and grasping of daily life objects with an implanted optic nerve prosthesis.
Invest. Ophthalmol. Vis. Sci. 44:3585.
[137] Landau K, Wichmann W, Valavanis A. (1995) The missing temporal crescent. Am. J.
Ophthalmol. 119:345-349.
[138] Langner J. (2003) Master Thesis: Development of a parallel computing optimized head
movement correction method in PET. [ http://www.jens-langner.de/ftp/MScThesis.pdf ]
[139] Laurienti PJ, Burdette JH, Wallace MT, Yen YF, Field AS, Stein BE. (2002)
Deactivation of sensory-specific cortex by cross-modal stimuli. Journal of Cognitive
Neuroscience 14(3):420-429.
[140] Leclerc C, Saint-Amour D, Lavoie ME, Lassonde M, Lepore F. (2000) Brain functional
reorganization in early blind humans revealed by auditory event-related potentials.
Neuroreport 11:545–550.
[141] Lee DS, Lee JS, Oh SH, Kim SK, Kim JW, Chung JK, Lee MC, and Kim CS. (2001)
Cross-modal plasticity and cochlear implants. Nature 409:149-150.
[142] Lessard N, Paré M, Lepore F, Lassonde M. (1998) Early-blind human subjects localize
sound sources better than sighted subjects. Nature 395:278–280.
[143] Liu W, Sivaprakasam M, Singh PR, Bashirullah R, Wang G. (2003) Electronic visual
prosthesis. Artif. Organs 27(11):986-995.
[144] Loeb GE. (1990) Cochlear prosthetics. Annu. Rev. Neurosci. 13:357-71.
[145] Lueschow A, Miller EK, Desimone R. (1994) Inferior temporal mechanisms for invariant
object recognition. Cereb. Cortex 4:523-531.
[146] Lumer ED, Rees G. (1999) Covariation of activity in visual and prefrontal cortex associated
with subjective visual perception. Proc. Natl. Acad. Sci. (Neurobiology) 96:1669-1673.
209
[147] Majji AB, Humayun MS, Weiland JD et al. (1999) Long-term histological and
electrophysiological results of an inactive epiretinal electrode array implantation in dogs.
Invest. Ophthalmol. Vis. Sci. 40(9):2073-2081.
[148] Malach R, Reppas JB, Benson RR, Kwong KK, Jiang H, Kennedy WA, Ledden PJ, Brady
TJ, Rosen BR, Tootell RB. (1995) Object-related activity revealed by functional magnetic
resonance imaging in human occipital cortex. Proc. Natl. Acad. Sci. USA 92:8135– 8139.
[149] Marc RE, Jones BW, Watt CB, Strettoi E. (2003a) Neural remodeling in retinal
degeneration. Prog. Retin. Eye Res. 22(5):607–655.
[150] Marc RE, Jones BW. (2003b) Retinal remodeling in inherited photoreceptor degenerations.
Mol. Neurobiol. 28(2):139–147.
[151] Margalit E, Maia M, Weiland JD, Greenberg RJ, Humayun MS et al. (2002) Retinal
Prosthesis for the blind. Survey of Ophthalmology 47(4):335-356.
[152] Margalit E, Sadda SR. (2003) Retinal and optic nerve diseases. Artif. Organs 27(11):963–
974.
[153] Maunsell JH and Van Essen DC. (1983) Functional properties of neurons in middle
temporal visual area of the macaque monkey. I. Selectivity for stimulus direction, speed,
and orientation. J. Neurophysiol. 49:1127–1147.
[154] Maunsell JH, Sclar G, Nealey TA, DePriest DD. (1991) Extraretinal representations in area
V4 in the macaque monkey. Vis. Neurosci. 7:561-573.
[155] Maurer D, Lewis TL, Mondloch CJ. (2005) Missing sights: Consequences for visual
cognitive development. Trends Cogn. Sci. 9:144–151.
[156] Maynard EM. (2001) Visual prostheses. Annu. Rev. Biomed. Eng. 3: 145–68.
[157] McDermott HJ. (2004) Music perception with cochlear implants: a review. Trends Amplif.
8:49-82.
[158] Miki A, Liu GT, Raz J, van Erp TGM, Liu CJ, Leuthardt EC, Haselgrove JC. (2000)
Contralateral monocular dominance in anterior visual cortex confirmed by functional
magnetic resonance imaging. American Journal of Ophthalmology 130(6):821-824.
[159] Miki A, Liu GT, Englander SA, van Erp TGM, Bonhomme GR, Aleman DO, Liu CJ,
Haselgrove JC. (2001) Functional magnetic resonance imaging of eye dominance at 4
Tesla. Ophthalmic Res. 33:276-282.
[160] Mitchell RN, Cotran RS. (1997) Chapter 3: Repair: cell regeneration, fibrosis, and wound
healing. In: Kumar V, Cotran RS, Robbins SL (eds). Basic Pathology, 6th edn.
Philadelphia: W.B. Saunders p.47–59.
210
[161] Miyake Y, Yanagida K and Yagasaki K. (1980a) Clinical application of EER (electrically
evoked response) (1) analysis of EER in normal subjects. Journal of the Japanese
Ophthalmology Society 84:354–360.
[162] Miyake Y, Yanagida K and Yagasaki K. (1980b) Clinical application of EER (electrically
evoked response) (2) analysis of EER in patients with dysfunctional rod or cone visual
pathway. Journal of the Japanese Ophthalmology Society 84:502–509.
[163] Miyake Y, Yanagida K and Yagasaki K. (1980c) Clinical application of EER (electrically
evoked response) (3) analysis of EER in patients with central retinal arterial occlusion.
Journal of the Japanese Ophthalmology Society 84:587–593.
[164] Miyake Y, Yanagida, K and Yagasaki K. (1980d) Clinical application of EER (electrically
evoked response) (4) analysis of EER in patients with optic nerve disease. Journal of the
Japanese Ophthalmology Society 84:2047–2052.
[165] Miyake K, Yoshida M, Inoue Y and Hata Y. (2007) Neuroprotective effect of transcorneal
electrical stimulation on the acute phase of optic nerve injury. Invest. Opthalmol. Vis. Sci.
48(5):2356-2361.
[166] Moran J, Desimone R. (1985) Selective attention gates visual processing in the extrastriate
cortex. Science 229(4715):782-784.
[167] Morimoto T, Miyoshi T, Matsuda S, Tano Y, Fujikado T and Fukuda Y. (2004)
Transcorneal electrical stimulation rescues axotomized retinal ganglion cells by activating
endogenous retinal IGF-1 system. Investigative Ophthalmology and Visual Science 46(6):
2147-2155.
[168] Morimoto T, Fukui T, Matsushita K, Okawa Y, Shimojyo H, Kusaka S, Tano Y, Fujikado
T. (2006) Evaluation of residual retinal function by pupillary constrictions and phosphenes
using transcorneal electrical stimulation in patients with retinal degeneration. Graefe Arch.
Clin. Exp. Ophthalmol. 244:1283–1292.
[169] Morimoto T, Fujikado T, Choi JS, Kanda H, Miyoshi T, Fukuda Y, Tano Y. (2007)
Transcorneal electrical stimulation promotes the survival of photoreceptors and preserves
retinal function in royal college of surgeons rats. Invest. Ophthalmol. Vis. Sci. 48(10):4725-
4732.
[170] Moseley M, Glover GH. (1995) Functional MR imaging: capabilities and limitations. In
Functional Neuroimaging, ed. J Kucharczyk, M Moseley, T Roberts, W Orrison. North
American Imaging Clinics.
[171] Mower GD, Berry D, Burchfiel JL, Duffy FH. (1981) Comparison of the effects of dark
rearing and binocular suture on development and plasticity of cat visual cortex. Brain Res.
220:255–267.
[172] Naples GG, Mortimer JT, Schneiner A, Sweeney JD. (1988) A spiral nerve cuff electrode
for peripheral nerve stimulation. IEEE Trans. Biomed. Eng. 35(11):905-916.
211
[173] Nelles G, Widman G, de Greiff A, Mesitrowitz A, Dimitrova A, Weber J. (2002) Brain
representation of hemifield stimulation in poststroke visual field defects. Stroke 33:1286-
1293.
[174] Nishimura H, Hashikawa K, Doi K, Iwaki T, Watanabe Y, Kusuoka H, Nishimura T, and
Kubo T. (1999) Sign language ‘heard’ in the auditory cortex. Nature 397:116.
[175] Nishimura H, Doi K, Iwaki T, Hashikawa K, Oku N, Teratani T, Hasegawa T, Watanabe A,
Nishimura T, Kubo T. (2000) Neural plasticity detected in short- and long-term cochlear
implant users using PET. Neuroreport 11:811-815.
[176] Noppeney U, Friston KJ, Ashburner J, Frackowiak R, Price CJ. (2005) Early visual
deprivation induces structural plasticity in gray and white matter. Curr. Biol. 15:R488–
R490.
[177] Normann RA, Maynard EM, Rousche PJ, Warren DJ. (1999) A neural interface for a
cotical vision prosthesis. Vision Research 39:2577-2587.
[178] Normann RA, Warren DJ, Ammermuller J, Fernandez E, Guillory S. (2001) High-
resolution spatiotemporal mapping of visual pathways using multi-electrode arrays. Vision
Res. 41(10-11):1261-1275.
[179] Normann RA. (2007) Technology insight: future neuroprosthetic therapies for disorders of
the nervous system. Nature Clinical Practice (Neurology) 3(8):444-452.
[180] Ogawa S, Tank D, Menon R, Ellermann J, Kim S, et al. (1992) Intrinsic signal changes
accompanying sensory stimulation: functional brain mapping with magnetic resonance
imaging. Proc. Natl. Acad. Sci. USA. 89:5951–5955.
[181] Okazawa H, Naito Y, Yonekura Y, Sadato N, Hirano S, Nishizawa S, Magata Y, Ishizu K,
Tamaki N, Honjo I, Konishi J. (1996) Cochlear implant efficiency in pre- and post-
lingually deaf subjects. A study with H2(15)O and PET. Brain 119:1297-1306.
[182] Orban GA, Fize D, Peuskens H, Denys K, Nelissen K, Sunaert S, Todd J, Vanduffel W.
(2003) Similarities and differences in motion processing between the human and macaque
brain: evidence from fMRI. Neuropsychologia 41:1757–1768.
[183] Pardue MT, Stubbs EB Jr, Perlman JI, et al. (2001) Immunohistochemical studies of the
retina following long-term implantation with subretinal microphotodiode arrays. Exp. Eye
Res. 73(3):333-343.
[184] Pelizzone M, Hari R, Makela J, Kaukoranta E, Montandon P. (1986) Activation of the
auditory cortex by cochlear stimulation in a deaf patient. Neurosci. Lett. 68:192-196.
[185] Pena C, Bowsher K, and Samuels-Reid J. (2004) FDA-approved neurologic devices
intended or use in infants, children, and adolescents. Neurology 63:1163-1167.
212
[186] Petit L, Clark VP, Ingeholm J, Haxby JV. (1997) Dissociation of saccade-related and
pursuit-related activation in human frontal eye fields as revealed by fMRI. J. Neurophysiol.
77:3386–3390.
[187] Petitto LA, Zatorre RJ, Gauna K, Nikelski EJ, Dostie D, Evans AC. (2000) Speech-like
cerebral activity in profoundly deaf people processing signed languages: implications for
the neural basis of human language. Proc. Natl. Acad. Sci. USA. 97:13961-13966.
[188] Peyman G, Chow AY, Liang C, et al. (1998) Subretinal semiconductor microphotodiode
array. Ophthalmic Surg. Lasers 29:234-241.
[189] Phelps ME, Kuhl DE, Mazziotta JC. (1981) Metabolic mapping of the brain’s response to
visual stimulation: Studies in humans. Science 211:573-589.
[190] Phelps ME, Mazziotta JC, Schelbert HR. (1986) Positron Emission Tomography and
Autoradiography: Principles and Applications for the Brain and Heart. Raven Press, New
York, pp 287-346.
[191] Podzebenko K, Egan GF, Watson JDG. (2002) Widespread dorsal stream activation during
a parametric mental rotation task, revealed with functional magnetic resonance imaging.
NeuroImage 15:547–558.
[192] Poggel DA, Toth LJ, Kim DS, and Rizzo JF. (2007) Brain activation maps in patients with
retinal degeneration. ARVO, poster 935-B910.
[193] Porac C, Coren S. (1975) Is eye dominance a part of generalized laterality? Percept and
Motor Skills 40(3):763–769.
[194] Posner MI, Raichle ME. (1994) Images of Mind. New York: Freeman.
[195] Potts AM, Inoue J and Buffum D. (1968) The electrically evoked response of the visual
system (EER). Investigative Ophthalmology and Visual Science 7:269–278.
[196] Potts AM, Inoue J. (1969) The electrically evoked response (EER) of the visual system II.
Effect of adaptation and retinitis pigmentosa. Invest. Ophthalmol. 8(6):605–612.
[197] Potts AM, Inoue J. (1970) The electrically evoked response of the visual system (EER) III.
Further consideration to the origin of the EER. Invest. Ophthalmol. 9:814-819.
[198] Ptito M, Kupers R. (2005) Cross-modal plasticity in early blindness. Journal of Integrative
Neuroscience 4:479-488.
[199] Rattay F. (1986) Analysis of models for external stimjulation of axons. IEEE Trans. on
Biomedical Engineering 33(10):974-977.
[200] Rattay F. (1989) Analysis of models for extracellular fiber stimulation. IEEE Trans. on
Biomedical Engineering 36(7):676-682.
213
[201] Rauschecker JP. (1995) Compensatory plasticity and sensory substitution in the cerebral
cortex. Trends Neurosci. 18:36–43.
[202] Reiss MR. (1997). Ocular dominance: some family data. Laterality 2(1):7-16.
[203] Rizzo M, Nawrot M, Blake R, Damasio A. (1992) A human visual disorder resembling area
V4 dysfunction in the monkey. Neurology 42:1175-1180.
[204] Rizzo JF, Socha M, Edell D, Antkowiak B, Brock D. (1994) Development of a silicon
retinal implant – surgical methods and mechanical design. Invest. Ophthalmol. Vis. Sci.
35(4):1535-1535.
[205] Rizzo JF, Wyatt J, Loewenstein J, Kelly S, Shire D. (2003) Perceptual efficacy of electrical
stimulation of human retina with a microelectrode array during short-term surgical trials.
Invest. Ophthalmol. Vis. Sci. 44(12):5362-5369.
[206] Rizzo JF, Wyatt J, Loewenstein J, Kelly S, Shire D, Wyatt J. (2003b) Methods and
perceptual thresholds for short-term electrical stimulation of human retina with
microelectrode arrays. Invest. Ophthalmol. Vis. Sci. 44(12):5355-5361.
[207] Röder B, Teder-Sälejärvi W, Sterr A, Rösler F, Hillyard SA, Neville HJ. (1999) Improved
auditory spatial tuning in blind humans. Nature 400:162–166.
[208] Roig-Melo EA, Alfaro III DV, Apple DJ. (2000) Pathophysiology of the retinal pigment
epithelium and choroid. In Macular Surgery, ed. H Quiroz-Mercado, DV Alfaro III, PE
Ligget, Y Tano, E de Juan Jr, pp. 14–23. Philadelphia: Lippincott, Williams & Wilkins.
[209] Rombouts SARB, Barkhof F, Sprenger M, Valk J, Scheltens P. (1996) The functional basis
of ocular dominance: functional MRI (fMRI) finding. Neuroscience Letters 221(1):1-4.
[210] Rorden C. MRIcroN, Version beta 12, Feb 2008. www.mricro.com.
[211] Sack RL, Lewy AJ, Blood ML, Keith LD, Nakagawa H. (1992) Circadian rhythm
abnormalities in totally blind people: incidence and clinical significance. J. Clin.
Endocrinol. Metab. 75: 127–34.
[212] Sadato N, Pascual-Leone A, Grafman J, Ibanez V, Deiber MP, Dold G, et al. (1996)
Activation of the primary visual cortex by Braille reading in blind subjects. Nature
380:526–528.
[213] Sadato N, Pascual-Leone A, Grafman J, Deiber MP, Ibanez V, Hallett M. (1998) Neural
networks for Braille reading by the blind. Brain 121: 1213-1229.
[214] Sadato N, Okada T, Honda M, Yonekura Y. (2002) Critical period for cross-modal
plasticity in blind humans: a functional MRI study. Neuroimage 16:389–400.
[215] Sadato N, Okada T, Honda M, Matsuki KI, Yoshida M, Kashikura KI, Takei W, Sato T,
Kochiyama T, Yonekura Y. (2004) Cross-modal integration and plastic changes revealed
214
by lip movement, random-dot motion and sign languages in the hearing and deaf. Cereb.
Cortex 15:1113-1122.
[216] Sadato N. (2005) How the blind “see” Braille: lessons from functional magnetic resonance
imaging. The Neuroscientist 11:577-582.
[217] Santos A, Humayun MS, de Juan EJ, Greenburg RJ, Marsh MJ, et al. (1997) Preservation
of the inner retina in retinitis pigmentosa. A morphometric analysis. Arch. Ophthalmol.
115(4):511–515.
[218] Schanze T, Wilm M, Eger M, Hesse L, Eckhorn R. (2002) Activation zones in cat visual
cortex evoked by electrical retina stimulation. Graefes Arch. Clin. Exp. Ophthalmol.
240(11):947-954.
[219] Schmidt EM, Bak MJ, Hambrecht FT et al. (1996) Feasibility of a visual prosthesis for the
blind based on intracortical microstimulation of the visual cortex. Brain 119(2):507-522.
[220] Schoups A, Vogels R, Qian RN, and Orban G. (2001) Practising orientation identification
improves orientation coding in V1 neurons. Nature 412:549-553.
[221] Schulman GL, Corbetta M, Buckner RL, Raichle ME, Fiez JA, Miezin FM, Petersen SE.
(1997) Top-down modulation of early sensory cortex. Cerebral Cortex 7:193-206.
[222] Schwahn HN, Gekeler F, Kohler K, 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):961-967.
[223] Sereno MI, Dale AM, Reppas JB, Kwong KK, Belliveau JW, Brady TJ, Rosen BR, Tootell
RBH. (1995) Borders of multiple visual areas in humans reveled by functional magnetic
resonance imaging. Science 268: 889– 893.
[224] Shandurina AN, Panin AV, Sologubova EK, et al. (1996) Results of the use of therapeutic
periorbital electrostimulation in neurological patients with partial atrophy of the optic
nerves. Neurosci. Behav. Physiol. 26:137-142.
[225] Sharma RK, Ehinger B. (1999) Management of hereditary retinal degenerations: present
status and future directions. Surv. Ophthalmol. 43(5):427–444.
[226] Shimazu K, Miyaki Y and Watanabe S. (1999) Retinal ganglion cell response properties in
the transcorneal electrically evoked response of the visual system. Vision Res. 39:2251-
2260.
[227] Shimony JS, Burton H, Epstein AA, McLaren DG, Sun SW, et al. (2006) Diffusion tensor
imaging reveals white matter reorganization in early blind humans. Cereb. Cortex 16:1653–
1661.
[228] Shulman RG. (1996) Interview with Robert G Shulman. J. Cogn. Neurosci. 8(5):474–480.
215
[229] Sinha P, Ostrovsky Y, Meyers E. (2006) Parsing visual scenes via dynamic cues. J. Vis.
6:95.
[230] Sokoloff L, Reivich M, Kennedy C, et al. (1977) The [C-14] deoxyglucose mehod for the
measurement of local cerebral glucose utilization: theory, procedure, and normal values in
the conscious and anesthetized albino rat. J. Neurochem. 28:897-916.
[231] Sokoloff L. (1984) Metabolic probes of central nervous system activity in experimental
animals and man. Sunderland, MA:Sinauer.
[232] Sorenson JA and Phelps ME. (1987) Physics in Nuclear Medicine. 2nd ed. Orlando: Grune
& Stratton Inc.
[233] Stanford Vision and Imaging Science and Technology. http://white.stanford.edu/
[234] Stett A, Barth W, Weiss S, Haemmerle H, Zrenner E. (2000) Electrical multi-site
stimulation of the isolated chicken retina. Vision Res. 40(13):1785-1795.
[235] Stone JL, Barlow WE, Humayun MS, de Juan EJ, Milam AH. (1992) Morphometric
analysis of macular photoreceptors and ganglion cells in retinas with retinitis pigmentosa.
Arch. Ophthalmol. 110(11):1634–1639.
[236] Sullivan RKP, WoldeMussie E, and Pow DV. (2007) Dendritic and synaptic plasticity of
neurons in the human age-related macular degeneration retina. Investigative
Ophthalmology & Visual Science 48(6):2782-2791.
[237] Sunaert S, Van Hecke P, Marchal G, Orban GA. (1999) Motionresponsive regions of the
human brain. Exp. Brain Res. 127:355–370.
[238] Takei K. (1988) Localization of the components of the electrically evoked response (EER)
of the visual system. Journal of the Japanese Ophthalmology Society 92:1682–1686.
[239] Talairach P and Tournoux J. (1988) “A Stereotactic coplanar atlas of the human brain”
Stuttgart Thieme.
[240] Tanghe SJ, Najafi K, Wise KD. (1990) A planar multichannel stimulating electrode for use
in neural prostheses. Sensors Actuators B. Chemical 1(1-6):464-467.
[241] Tassicker GE. (1956) Preliminary report on a retinal stimulator. Br. J. Physiol. Opt. 13:
102–105.
[242] Teo P, Sapiro G, Wandell BA. (1998) Creating connected representations of cortical gray
matter for functional MRI visualization. IEEE Med. Trans. 16:852–863.
[243] Ter-Pogossian MM, Raichle ME, and Sobel BE. (1980) Positron-Emission Tomography.
Scientific American 243:170-181.
[244] Tong F. (2003) Primary visual cortex and visual awareness. Nature Reviews (Neuroscience)
4:219-229.
216
[245] Toosy AT, Werring DJ, Plant GT, Bullmore ET, Miller DH, Thompson AJ. (2001)
Asymmetrical activation of human visual cortex demonstrated by functional MRI with
monocular stimulation. Neuroimage 14(3):632-641.
[246] Tootell RB, Mendola JD, Hadjikhani NK, et al. (1997) Functional analysis of V3A and
related areas in human visual cortex. J. Neurosci. 17:7060-7078.
[247] Toyoda J, Fujimoto M. (1984) Application of transretinal current stimulation for the study
of bipolar-amacrine transmission. J. Gen. Physiol. 84:915-925.
[248] Treue S, Maunsell JH. (1996) Attentional modulation of visual motion processing in
cortical areas MT and MST. Nature 382:539-541.
[249] Trojano L, Grossi D, Linden DE, Formisano E, Goebel R, Cirillo S, Elefante R, Di Salle F.
(2002) Coordinate and categorical judgements in spatial imagery. An fMRI study.
Neuropsychologia 40:1666–1674.
[250] Ungerleider LG, Desimone R. (1986) Cortical connections of visual area MT in the
macaque. J. Comp. Neurol. 248:190-222.
[251] U.S. Department of Energy Office of Science: http://artificialretina.energy.gov/
[252] U.S. Department of Energy’s Historical Commitment to Improved Healthcare through
Nuclear Medicine. [ http://www.doemedicalsciences.org/mnm/timeline.shtml ]
[253] Vaina LM. (1989) Selective impairment of visual motion interpretation following lesions of
the right occipito-parietal area in humans. Biol. Cybern. 61:347-359.
[254] Van Essen DC, Zeki SM. (1978) The topographic organization of Rhesus monkey prestriate
cortex. J. Physiol. 277:193–226.
[255] Veraart C, Raftopoulos C, Mortimer JT et al. (1998) Visual sensations produced by optic
nerve stimulation using an implanted selfsizing spiral cuff electrode. Brain Res. 813: 181–
186.
[256] Veraart C, WanetoDefalque MC, Delbeke J, et al. (2001) Assessment of the MIVIC optic
nerve visual prosthesis. Invest. Ophthalmol. Vis. Sci. 42(suppl):S942.
[257] Veraart C, Want-Defalque M, Gerard B, Vanlierde A, Delbeke J. (2003) Pattern recognition
with the optic nerve visual prosthesis. Artif. Organs. 27: 996–1004.
[258] Voss P, Lassonde M, Gougoux F, Fortin M, Guillemot JP, Lepore F. (2004) Early and late-
onset blind individuals show supra-normal auditory abilities in far-space. Curr. Biol.
14:1734–1738.
[259] Voss P, Gougoux F, Lassonde M, Zatorre RJ, Lepore F. (2006) A positron emission
tomography study during auditory localization by late-onset blind individuals. Neuroreport
17:383-388.
217
[260] Wandell BA. (1999) Computational neuroimaging of human visual cortex. Annu. Rev.
Neurosci. 22:145-173.
[261] Wang GJ, Volkow ND, Overall J, Hitzemann RJ, Pappas N, Pascani K, Fowler JS. (1996)
Reproducibility of regional brain metabolic responses to lorazepam. J. Nucl. Med.
37(10):1609-1613.
[262] Watanabe M. (1986) Visual information processing from the retina to the prefrontal cortex.
Shinrigaku Kenkyu 56:365-378.
[263] Weeks R, Horwitz B, Aziz-Sultan A, Tian B, Wessinger CM, Cohen LG, et al. (2000) A
Positron emission tomographic study of auditory localization in the congenitally blind. J.
Neurosci. 20:2664–2672.
[264] Weih L, McCarty CA, Taylor HR. (2000) Functional implications of vision impairment.
Clin. Experiment Ophthalmol. 28: 153–155.
[265] Weisz JM, O’Connell SR, Bressler NM. (2000) Treatment guidlines for age-related
macular degeneration based upon results from the macular photocoagulation study. In
Macular Surgery, ed. H Quiroz-Mercado, DV Alfaro III, PE Ligget, Y Tano, E de Juan Jr,
pp. 201–211. Philadelphia: Lippincott, Williams & Wilkins.
[266] Westheimer G. (2003) Entoptic phenomena. Adler’s Physiology of the Eye. Kaufman L,
Alm A, eds. St. Louis: Mosby; p.441–452.
[267] Wiesel TN, Hubel DH. (1965) Extent of recovery from the effects of visual deprivation in
kittens. J. Neurophysiol. 28:1060–1072.
[268] Wilms M, Eger M, Schanze T, Eckhorn R. (2003) Visual resolution with epi-retinal
electrical stimulatiion estimated from activation profiles in cat visual cortex. Vis. Neurosci.
20(5):543-555.
[269] Worsley KJ, Marrett S, Neelin P, Vandal AC, Friston KJ, and Evans AC. (1996) A unified
statistical approach or determining significant signals in images of cerebral activation.
Hum. Brain Mapp. 4:58-73.
[270] Yanai D, Lakhanpal RR, Weiland, JD, et al. (2003) The value of preoperative tests in the
selection of blind patients for a permanent microelectronic implant. Trans. Am.
Ophthalmol. Soc. 101:223–228.
[271] Zeki SM. (1969) Representation of central visual fields in prestriate cortex of monkey.
Brain Res. 14:271–291.
[272] Zeki S, Watson JD, Lueck CJ, Friston KJ, Kennard C, Frackowiak RS. (1991) A direct
demonstration of functional specialization in human visual cortex. J. Neurosci. 11:641-649.
[273] Zeki S. (1993) A Vision of the Brain. Oxford, UK: Blackwell Science.
218
[274] Zrenner E, Miliczek KD, Gabel VP et al. (1997) The development of subretinal
microphotodiodes for replacement of degenerated photoreceptors. Ophthalmic Res.
29(5):269-280.
[275] Zrenner E, Strett A, Weiss S, et al. (1999) Can subretinal microphotodiodes successfully
replace degenerated photoreceptors? Vision Res. 39:2555-2567.
[276] Zrenner E. (2002) Will retinal implants restore vision? Science 295(5557):1022-1025.
219
Appendices
Appendix A
Three-Compartmental Model for FDG Metabolism
The Michaelis-Menten hypothesis states:
Where S is the substrate, P is the product, E is the enzyme, SE is the enzyme-substrate
complex and k
1
, k
2
, and k
3
are the rate constants. The reaction rate R is:
The term V
m
is the maximum rate of the reaction, and K
m
is the concentration of S that
produces a reaction rate one half the maximum value. If more than one substrate is
competing for the enzyme, then the reaction rates R and R’ for the competing processes
can be written as:
If S’ represents a tracer of the original substrate S, the [S’] is a much lower value than
[S]. Therefore:
1 ) / ' ( ) / (
' / ' '
'
1 ) / ' ( ) / (
/
m m
m m
m m
m m
K S K S
K S V
R
K S K S
K S V
R
m
m
K S
S V
R
(A.2)
S + E
SE P + E
S + E
SE SE
k
1
k
2
k
3
(A.1)
(A.4)
(A.3)
220
Divide equation A.6 by A.5:
Now consider the Sokoloff deoxyglucose method [230]. If glucose metabolism in the
tissue of interest is assumed to be in a steady state, then the rate of the hexokinase
reaction will be the rate of the entire process of glycolysis. Let MRGLc* (µmol/min/g)
refer to the metabolic rate of FDG and MRGlc (µmol/min/g) be the metabolic rate for
glucose. If C
t
is the concentration of free (unphosphorylated) glucose in the tissue space
and C
*
t
is the corresponding concentration for FDG, then, from equation A.7 and the fact
that the rate of phosphorylation equals k
3
×C
t
for glucose and k
3
×C*
t
for FDG, it follows
that:
If the plasma glucose (C
p
) and plasma FDG (C
*
p
) concentrations are constant and both
sides of equation A.8 are multiplied by C
p
/C
*
p
, and divided by blood flow (F) then:
The ratio of terms on the left side of equation A.9 is defined as the lumped constant (LC)
of the FDG model. It is simply the net extraction of FDG divided by the net extraction of
glucose. In steady state, the flux of FDG into tissue is balanced by the flux out:
LC
C C K V
C C K V
F C MRGlc
F C MRGlc
p t m m
p t m m
p
p
/
/
/
/ *
*
* * * *
t
t
t m m
t m m
C k
C k
C K V
C K V
MRGlc
MRGlc
3
* *
3
*
* * *
S K V
S K V
R
R
m m
m m
'
' ' '
'
' '
/
' S
K S
K K V
R
K S
S V
R
m
m m m
m
m
(A.5)
(A.6)
(A.8)
(A.9)
(A.7)
221
Given equation A.8, A.9 and A.12, with some algebraic manipulation it can be shown
that MRGlc is given by:
Equation A.13 produces a local estimate of MRGlc if C
p
(steady-state plasma glucose
value) is measured, if LC is known and if the rate constants for FDG (k
*
1
, k
*
2
, and k
*
3
)
are determined for each region of organ of interest. In actual practice, the Sokoloff
operational equation of the FDG model given by equation A.14 is used.
T T
t k k T k k
T
t k k T k k
dt e e dt LC
dt e e k
MRGlc
0 0
) ( ) (
0
) ( ) ( *
1
*
3
*
2
*
3
*
2
*
3
*
2
*
3
*
2
) ( (
p
*
p
p
*
p
*
p
*
i
C
t C
C
t) C
t C T C
(A.14)
C
*
i
(T) is the total
18
F tissue concentration at time T. In this approach, only the quantities
in bold type in Equation A.14 are measured. Average estimates of the rate constants and
LC [116, 190] are used as a part of the routine calculation of MRGlc using Equation
A.14.
*
3
*
2
*
3
*
1
k k
k k
LC
C
MRGlc
p
flux in = flux out
*
*
*
3
*
2
*
1
* *
3
* *
2
* *
1
p
t
t t p
C
C
k k
k
C k C k C k
(A.10)
(A.11)
(A.12)
(A.13)
222
Appendix B
General Linear Model Parameter Estimation
The General Linear Model can be written in matrix notation as:
Y is the column vector of observations; ɛ is the column vector of error terms; β is the
column vector of parameters; and X is the design matrix which has one row per
observation, and one column (explanatory variable) per model parameter.
Once an experiment has been completed, we have J observations of the random
variable Y
j
, and L number of parameters, β. Because the number of parameter, L, is
generally chosen to be less than the number of observations J, the linear set of equations
can not be solved exactly. Some method of estimating parameters that “best fit” the data
is required. This is done with least squares.
We denote a set of parameter estimates as:
These parameters lead to fitted values:
Giving residual error:
The residual sum-of-squares (S) is the sum of the square differences between the actual
and fitted values:
J
j
T
j
e e e S
1
2
X Y
(B.1)
T
L
~
,....,
~ ~
1
(B.2)
~ ~
,...,
~ ~
1
X Y Y Y
T
J
(B.3)
X Y Y Y e e e
T
J
~
,...,
1
(B.4)
(B.5)
223
S can also be written in full as:
This is minimized when:
This equation is the Ɩ
th
row of
If (X
T
X) is invertible, which it is if and only if the design matrix X is of full rank, then
the least squares estimates are:
2
1
1 1
)
~
...
~
(
L jL
J
j
j j
x x Y S
(B.6)
(B.7)
0
~
...
~
2
~
1
1 1
J
j
L jL j j jl
l
x x Y x
S
(B.8)
~
X X Y X
T T
(B.9)
Y X X X
T T
1 ~
224
Appendix C
Activating Function Derivation
The derivation of activating function is based on the work of Frank Rattay [199, 200] in
his attempt to present a mathematical model for the analysis as well as for the computer
simulation of the stimulus/response characteristics of nerve or muscle fibers.
Figure C.1 Electrical network to simulate the currents in an axon. Unmyelinated (a) and
myelinated fibers (b) are segmented into cylinders of length Δx. For the myelinated case, Δx is
given by the internodal distance, whereas Δx depends only on computational accuracy for
unmyelinated fibers. Within one segment the membrane is active in the purple-colored region of
length L which is the nodal width in (c), and L = Δx in the unmyelinated fiber (a). An electrical
circuit (c) is used to model the axons, consisting of capacity C
m
, resting membrane voltage E
r
,
membrance conductance G
m
, and axoplasmic conductance G
a
. External potential is denoted by
‘e’, and internal (intracellular) potential by ‘׀’. Red-color dot in (c) indicates node ‘n’ with current
(red arrows) flowing out from it; node ‘n’ is represented as purple-colored segments in (a) & (b).
225
Symbol Definition Unit
C
m
capacity of membrane [µF]
c
m
capacity of membrane per cm
2
[µF/cm
2
]
G
a
conductance of axoplasm [mS/cm]
ρ
i
resistivity of axoplasm [kΩ·cm]
V
i,n
internal potential at n
th
node [mV]
V
e,n
external potential at n
th
node [mV]
V
rest
internal resting potential [mV]
V
n
reduced membrane voltage at n
th
node [mV]
Δx
segmentation length of fiber [cm]
L
active length of membrane [cm]
d
fiber diameter [cm]
I
ionic,n
ionic current at n
th
node [µA]
i
ionic,n
ionic current density at n
th
node [µA/cm
2
]
t
time [ms]
Table C.1 Definition of symbols with corresponding units.
As shown in Figure C.1, the stimulating electrode for both the unmyelinated and
myelinated axon causes a space and time dependent potential V
e
in the medium where
nerve fibers are imbedded. Consider the purple-color segment of the axon that is situated
directly under the electrode, it can be modeled as node ‘n’ in the electrical circuit (Fig
C.1c). The current flow for the n
th
segment of the fiber at the point marked with a red dot
is caused by the voltages between the different points of the network and consists of a
capacitance current, an ionic current and two currents along the inside of the fiber.
(Symbols and units are listed in Table C.1) By applying Kirchoff’s Current Law at n
th
node:
0 ) ( ) (
) (
1 , , 1 , , ,
, ,
n i n i a n i n i a n ionic
n e n i
m
V V G V V G I
dt
V V d
C (C.1)
226
We introduce the reduced voltages where V
rest
is the resting membrane potential:
rest n e n i n
V V V V
, ,
(C.2)
Substitute C.2 into C.1, and find:
m n ionic n e n e n e n n n a
n
C I V V V V V V G
dt
dV
/ ) 2 2 (
, 1 , , 1 , 1 1
(C.3)
By inserting G
a
= πd
2
/4ρ
i
Δx and C
m
= πdLc
m
, and introducing the ionic current
density i
ionic,n
, Equation C.3 is transformed to:
m n ionic
n e n e n e
n n n
i
n
c i
x
V V V
x
V V V
L
x d
dt
dV
,
2
1 , , 1 ,
2
1 1
2 2
4
(C.4)
The ionic currents will be described by further differential equations, e.g., the
Hodgkin-Huxley equations for the unmyelinated case or the Frankenhaeuser-Huxley
equations for the myelinated fiber [200]. Equation C.4 shows that the influence of
extracellular current sources is given by:
2
1 , , 1 ,
2
) (
x
V V V
t f
n e n e n e
n
(C.5)
f
n
(t) is the second difference quotient of the extracellular potential along the fiber.
In the case of unmyelinated fiber, L = Δx simplifies (C.4), and with Δx→0, (C.5)
becomes:
2
2
) , (
) , (
x
t x V
t x f
e
(C.6)
Where x is the length coordinate of the fiber. ) , ( t x f in (C.6) is called the
activating function because it is responsible for activating a fiber by extracellular
stimulation. To obtain an action potential in a fiber which is at rest, the reduced voltage
227
V
n
of (C.2) has to become positive. At rest the reduced voltage anywhere along the fiber
is approximately the same, thus 0 2
2
1 1
x V V V
n n n
. Therefore, according to
(C.4), an action potential can be generated at the fiber coordinate x n x
n
if ) (t f
n
is
positive.
For the unmyelinated axon, the activating function becomes the second spatial
derivative of the external potential in the direction of the axon, whereas the activating
function for myelinated axons is given by the second difference quotient of the external
potential along the axon.
228
Appendix D
Human Subject Inclusion and Exclusion Criteria
Normal Control Inclusion Criteria: 1) Ability to understand and give informed consent; 2)
Males and females 18-80 years of age. Inclusion criterion #1 will be determined through
verbal communication with the subject. Inclusion criterion #2 will be verified by medical
history.
Normal Control Exclusion Criteria: 1) Urine positive for psychoactive drugs (including
cocaine, PCP, THC, benzodiazepines, amphetamine, opiates, and barbiturates); 2)
Pregnant; 3) Breast feeding; 4) Use of prescription (non-psychiatric) medication(s) that
can affect brain function, except for birth control, in the past two weeks (including but
not limited to antihistamines, which can affect brain function); 5) History of uncontrolled
diabetes (i.e., blood glucose >180 on any day of study as measured on Beckman Glucose
Analyzer II); 6) History of acute or chronic medical illness that in the opinion of
physicians may affect brain function (e.g. endogenous depression or mental illness); 7)
History of claustrophobia or fear of small enclosed space. Exclusion criterion #1 are
determined by immediate urodynamic study (STAT UDS) on each day of study.
Exclusion #2 by immediate urinary chorionic gonadotropin (STAT UCG) for females of
childbearing potential on each day of study. Exclusion criterion #’s 3-7 are determined
by obtaining a medical history and through verbal communication with the subject.
229
Retinal Degeneration (RD) Subjects Inclusion Criteria: 1) Males and females 18-80
years of age; 2) Subject exhibits retinal dystrophy as diagnosed by an ophthalmologist; 3)
Greater than 10 years of light perception or near light perception only best visual acuity;
4) Subject does not have implantable devices (e.g. pacemaker); and 5) Must be fairly
mobile (suitable for traveling to Brookhaven, NY via a limousine). All Inclusion criteria
are verified by medical history and source documents provided to BNL from referring
physician(s). Also, since the subjects are visually impaired, an individual not associated
with this study must read the informed consent to the subject. This person must sign the
consent form as a witness to the informed consent process.
Retinal Degeneration (RD) Subjects Exclusion Criteria: In addition to the exclusion
criteria specified for the normal controls, RD subjects must meet the following two
exclusion criteria: 1) No significant life-threatening health problem (i.e. seizure,
congestive heart failure); and 2) No significant ocular disease (e.g. cataract, glaucoma,
optic neuropathy, diabetic retinopathy) besides RD. All exclusion criteria are verified by
medical history and source documents provided to BNL from referring physician(s).
RD subjects were screened and recruited by co-investigator Dr. Steven Tsang and his
associates at Columbia University, NY.
Abstract (if available)
Abstract
Objective measures that demonstrate activation of the retina, primary visual cortex, and possibly higher cortical association areas are necessary to illustrate the functionality of retinal prosthesis in sight-impaired subjects who have been implanted with such a device. Neural remodeling that occurs in outer retinal degenerative diseases can have an impact on the usefulness of retinal prostheses.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Modeling retinal prosthesis mechanics
PDF
Manipulation of RGCs response using different stimulation strategies for retinal prosthesis
PDF
Towards a high resolution retinal implant
PDF
Effect of continuous electrical stimulation on retinal structure and function
PDF
An in vitro model of a retinal prosthesis
PDF
Saliency based image processing to aid retinal prosthesis recipients
PDF
Cortical and subcortical responses to electrical stimulation of rat retina
PDF
Understanding the degenerate retina's response to electrical stimulation: an in vitro approach
PDF
Characterization of visual cortex function in late-blind individuals with retinitis pigmentosa and Argus II patients
PDF
Intraocular and extraocular cameras for retinal prostheses: effects of foveation by means of visual prosthesis simulation
PDF
Improving stimulation strategies for epiretinal prostheses
PDF
Prosthetic visual perception: retinal electrical stimulation in blind human patients
PDF
Therapeutic electrical stimulation strategies for neuroregeneration and neuroprotection of retinal neurons
PDF
Viewing the picture we paint
PDF
Stimulation strategies to improve efficiency and temporal resolution of epiretinal prostheses
PDF
Prosthetic vision in blind human patients: Predicting the percepts of epiretinal stimulation
PDF
Next generation neural interfaces for neural modulation and prosthesis
PDF
Electrical stimulation of degenerate retina
PDF
Investigation of the electrode-tissue interface of retinal prostheses
PDF
Intraocular camera for retinal prostheses: refractive and diffractive lens systems
Asset Metadata
Creator
Xie, John Zhong
(author)
Core Title
PET study of retinal prosthesis functionality
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Biomedical Engineering
Publication Date
09/08/2009
Defense Date
08/11/2009
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
blindness,functional imaging,light stimulation,OAI-PMH Harvest,PET,positron emission tomography,quantitative PET,retinal degenerative diseases,retinal prosthesis,transcorneal electrical stimulation,visual cortex plasticity
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Humayun, Mark (
committee chair
), Jadvar, Hossein (
committee member
), Khoo, Michael C.K. (
committee member
), Sampath, Alapakkam P. (
committee member
), Singh, Manbir (
committee member
), Weiland, James D. (
committee member
)
Creator Email
jian_xie@hotmail.com,jianxie@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m2591
Unique identifier
UC1314234
Identifier
etd-Xie-3229 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-262952 (legacy record id),usctheses-m2591 (legacy record id)
Legacy Identifier
etd-Xie-3229.pdf
Dmrecord
262952
Document Type
Dissertation
Rights
Xie, John Zhong
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
Tags
functional imaging
light stimulation
PET
positron emission tomography
quantitative PET
retinal degenerative diseases
retinal prosthesis
transcorneal electrical stimulation
visual cortex plasticity