Impaired and Enhanced CMOS Phototransistors"
Multiple-Unit Artificial Retina Chipset
Dr. Elliot McGucken
University of North Carolina at Chapel Hill
NC State University
A computer-chip based device that can provide limited-resolution vision
for people with retinal-based blindness. Beneficiaries would be 10,000,000
people worldwide suffering from forms of blindness including retinal
pigmentosa and age-related macular degeneration.
MERRILL LYNCH "INNOVATION GRANTS" AWARDED TO FIVE DOCTORAL STUDENTS
DOCTORAL RESEARCH YIELDS GROUNDBREAKING PROPOSALS RANGING FROM NEW
COMPUTER CHIPS TO A MALE ORAL CONTRACEPTIVE
NEW YORK, Sept.16 -- The Merrill Lynch Forum today announced the first
winners of the Innovation Grants Competition -- its global competition
challenging doctoral students to craft commercial applications of their
dissertation research. The winners were recognized at an awards dinner at
Merrill Lynch headquarters last night (Sept. 15), hosted by Merrill Lynch
Chairman and CEO David H. Komansky.
Dr. Jan Mark Noworolski, from the University of California at Berkeley,
received the top prize in the competition for creating a new type of power
converter, a key element in virtually all electronic devices. This
technology would greatly reduce the size, parts count and weight of power
supplies for the increasingly pervasive array of portable electronic
products such as cell phones and laptop computers, as well as enabling the
design of new mobile electronic products. "Power management is one of the
major constraints in personal electronics," he said. "An integrated design
using this technology could offer a 10-fold improvement in device
A total of 213 proposals from 16 countries were submitted to the
competition, which was open to new Ph.D. recipients in the sciences,
liberal arts, and engineering disciplines. Entries were judged by a
distinguished panel of nine entrepreneurs, venture capitalists,
journalists, and innovators and were considered without knowledge of the
applicants' identity or academic affiliation.
"Academic research is a significant and often untapped source of
intellectual capital in our society, and a tremendous economic resource,"
said Merrill Lynch Chairman and CEO David H. Komansky. "The winning
proposals from this competition are all excellent examples of how new
knowledge can be transformed into new value simply by encouraging
researchers to look at their research from a different perspective. We
hope that these Innovation Grants will help foster a closer interaction
between world-class science and the world of commerce," Mr. Komansky
The judging panel consisted of:
John Seely Brown, Chief Scientist, Xerox Corporation, and Director, Xerox
Palo Alto Research Center Edgar W. K. Cheng, former Chairman, The Stock
Exchange of Hong Kong John Doerr, Partner, Kleiner Perkins Caufield &
Byers Esther Dyson, Chairman, EDventure Holdings, Inc. Peter C. Goldmark,
Chairman and Chief Executive, The International Herald Tribune William
Haseltine, Chairman & CEO, Human Genome Sciences, Inc. John Markoff,
Technology Correspondent, The New York Times Edward McKinley, President,
E.M. Warburg, Pincus & Company International, Ltd. Arati Prabhakar, former
Chief Technology Officer, Raychem Corporation In evaluating the
applications, the judges sought to identify proposals with the potential
to affect real change in industries and in the way people live their
lives. "The Innovation Grants Competition is a terrific idea," said judge
John Doerr, of venture-capital firm Kleiner Perkins Caufield & Byers. "I
was impressed with many of the proposals and thought that several of the
ideas would merit a venture-capital follow-up."
The five winning entries:
First Place, $50,000 -- Single-Chip Power Converter. Dr. Jan Mark
Noworolski, University of California at Berkeley. A unique, one-chip power
converter that uses electromechanical energy instead of inductive energy
storage. This technology could dramatically reduce the size and complexity
of portable electronic devices such as laptop computers, cellular phones,
Second Place, $20,000 -- Membrane Chips. Dr. Jay T. Groves, Stanford
University. A technology that enables biological membranes to be
incorporated into computer chips. These chips could be used by the medical
diagnostic industry, particularly for AIDS research, and leukemia.
Second Place, $20,000 -- Multiple-Unit Artificial Retina Chipset (MARC).
Dr. Elliot McGucken, University of North Carolina at Chapel Hill/NC State
University. A computer-chip based device that can provide
limited-resolution vision for people with retinal-based blindness. This
device could benefit the more than 10,000,000 people worldwide suffering
from blindness originating from various causes.
Third Place, $10,000 -- Male Oral Contraceptive. Dr. Bruce Lahn, Whitehead
Institute of Biomedical Research, Massachusetts Institute of Technology.
This research led to the development of an understanding of the role of
the gene CDY in producing an essential enzyme for sperm production. This
research could produce a male oral contraceptive that would chemically
inhibit the production of the sperm-producing enzyme.
Third Place, $10,000 -- Artificially Engineered Quantum Solid Materials.
Dr. Alexander Balandin, University of Notre Dame. This study of new
materials based on quantum confinement properties suggests opportunities
for the engineering of a new generation of electronic devices. The most
significant market application would be the improvement of devices such as
semiconductor lasers, CD players, digital cameras, and optical drives.
Additional grants of $5,000 were awarded to each of the winners'
universities and discretionary grants of $3,000 each were awarded to five
The 1998 Innovation Grants Competition was directed by Michael Schrage, a
Research Associate at the MIT Media Lab, and a leading expert on issues
surrounding innovation and new business development. "What fuels the 'new
economy' of the information age is ideas," said Schrage. "This competition
takes great ideas that might otherwise have languished for years in
academia and brings them to the attention of people who can translate them
into transformative technologies. Anyone looking at these proposals can
see that they contain truly exciting possibilities."
The competition was open to doctoral students who successfully defended
their dissertations between January 1, 1996, and July 1, 1998. Entrants
were required to submit a 3,000-word explanation of how their dissertation
topic could be translated into a commercial product or service. The
description had to include: a summary of the dissertation, a description
of the most significant commercial idea embodied in it, an analysis of the
potential market for the product or service, and a discussion of technical
steps necessary to bring the innovation to market.
The Merrill Lynch Forum is a "virtual" think tank established by the
global financial services company to bring together leading experts to
consider and explore issues of worldwide importance in the areas of
technology, economics, and international relations.
Those interested in additional information, should visit the Competition's
web site, http://www.ml.com/innovation, or call 1-888-33Forum. Additional
information is also available by sending e-mail to:InnovationGrants@ml.com
DR. ELLIOT MCGUCKEN
Dr. Elliot McGucken was born and raised in Akron, Ohio, and he has studied
and taught physics ever since he left Akron to attend Princeton University
as an undergraduate. He recently received his Ph.D. in physics from the
University of North Carolina at Chapel Hill (1998), where his research on
the Multiple Unit Artificial Retina Chipset To Aid The Visually Impaired
often led him down the road to North Carolina State University. He is
currently continuing his involvement with the retinal prosthesis's
prototype development at NCSU, while also teaching physics and astronomy
as an assistant professor of physics at the neighboring Elon College.
His favorite hobbies are celestial navigation, sailing and windsurfing,
reading the classics, and writing poetry. Dr. McGucken received the Tanner
Award for Excellence in Undergraduate Teaching while at the University of
North Carolina at Chapel Hill, where he also received an honorary
membership in the the American Society of Physics Teachers.
Multiple Unit Artificial Retina Chipset (MARC) to Aid the Visually
Impaired By Elliot McGucken
1. Summary of the dissertation Engineering progress relating to the
development of the multiple-unit artificial retina chipset (MARC)
prosthesis to benefit the visually impaired is presented in my
dissertation, "Multiple Unit Artificial Retina Chipset to Aid the Visually
Impaired and Enhanced CMOS Phototransistors." The design, fabrication, and
testing of the first generation MARC VLSI chips are reported on. A
synthesis of the engineering, biological, medical, and physical research
is offered within the presentation of methods and means for the overall
design engineering, powering, bonding, packaging, and hermetic sealing of
the MARC retinal prosthesis. The retinal prosthesis is based on the
fundamental concept of replacing photoreceptor function with an electronic
device1, which was initiated by2 and has been extensively developed3,4 by
MARC team-members Dr. Humayun et al. The use of an inductive link for
power and telemetric communications is explored, and an experimental study
of RF coil configurations, showing their feasibility for this retinal
implant, is offered. An enhanced CMOS phototransistor with a holed emitter
(HEP), used in the first generation MARC, is presented, along with a
numerical model which also predicts its enhanced quantum efficiency. Due
to the small size of the intraocular cavity, the extreme delicacy of the
retina, and the fact that the eye is mobile, an artificial retinal implant
poses difficult engineering challenges. Over the past several years all of
these factors and contrasts have been taken into consideration in the
engineering research of an implantable retinal device. Initial steps3
towards fabricating a commercially available, implantable MARC device have
been taken by our team of engineers, physicists, and doctors.
2. Description of the most significant commercial aspect
A multiple-unit artificial retina chipset (MARC) would create a new
marketplace by offering a cure for forms of blindness including retinal
pigmentosa (RP) and age-related macular degeneration (AMD), which afflict
over 10,000,000 people worldwide. Clinical studies4 have shown that
controlled electrical signals applied to a small area of a dysfunctional
retina with a microelectrode can be used to initiate a local neural
response in the remaining retinal cells. The neural response, or
phosphene, is perceived by otherwise completely blind patients as a small
spot of light, about the size of a match-head held out at arm's length.
When multiple electrodes are activated in a two-dimensional electrode
array, an image may be stimulated upon the retina. The MARC system
consists of an extraocular means for acquiring and processing visual
information, a means for power and signal transceiving via RF telemetry,
and a multiple-until artificial retina chipset. The stimulating electrode
array is mounted on the retina with metal-alloy retinal tacks while the
power and signal transceiver is mounted in close proximity to the cornea.
An external miniature low-power CMOS camera worn in an eyeglass frame
captures an image and transfers the visual information and power to the
intraocular components via RF telemetry. The intraocular prosthesis will
decode the signal and electrically stimulate the retinal neurons through
the electrodes in a manner that corresponds to the original image
perceived by the CMOS Camera.
3. Description of the market for the proposed product and the competition
The multiple-unit artificial retina chipset (MARC) is designed to provide
useful vision to over 10,000,000 people blind because of photoreceptor
loss due to partial retinal degeneration from diseases such as Age Related
Macular Degeneration (AMD) and Retinitis Pigmentosa (RP). People who are
completely blind will initially gain the ability to discern shapes and
pictures, and even to read, with limited resolutions of 15x15 pixels.
Future MARC generations will provide greater resolutions, and the device
will chart a brand new marketplace a s a prosthetic device to aid the
3.1 Unique value derived by the customer
Before embarking on the MARC chip design, it was necessary to assess how
useful a limited-resolution view would b to the blind. Simple visual
feasibility experiments have been conducted at NCSU so as to determine how
well sight could be restored with a 15x15 array of pixels, each of which
would be capable of four-bit stimulation, or sixteen gray levels. A
picture from a video camera was projected onto a television screen at the
low resolution of 15x15 pixels. When subjects who wore glasses removed
their glasses, or when those with good sight intentionally blurred their
vision, the natural spatial-temporal processing of the brain allowed them
to actually distinguish features and recognize people. When the subject
focused on the screen, it appeared as a 15x15 array of gray blocks, but
when the subject "trained" themselves to unfocus their vision, they were
able to "learn" to see definitive edges and details such as beard, teeth,
and opened or closed eyes. These results are reminiscent of the
experiences with the artificial cochlear implant. When the artificial
cochlear was originally being designed, it was believed that over 2,400
electrodes would be needed to stimulate the nerves in a manner that would
be conducive to hearing. Today, however, within a few weeks of receiving
an implant, a patient can understand phone conversations with an
artificial cochlear that has only six electrodes. One of the advantages of
this project is that the MARC device will be interfaced with the world's
greatest computer - the brain. The MARC won't be duplicating the exact
functioning of the retina, but rather the device will be an entity that
the brain will "learn" to use. A good analogy to think of is that in
attempting flight, the Wright brothers did not attempt to imitate nature
by building a plane which flapped its wings, but rather they did it in a
way that had not yet appeared in the natural world. Thus we believe that a
15x15 pixel array will facilitate a level of sight which will be of
significant value to the patient. And after the initial prototype is
developed, there will be few barriers to stepping up the resolution.
3.2 Prior art, competition, and MARC advantages
The current design of the MARC clears several hurdles that exist is prior
inventions and research. Much of the prior art has relied upon structures
so complicated or biologically intrusive as to make their implementation
impractical, and thus, to date, an operating implantable artificial retina
has not been achieved. Several international teams are actively pursuing a
prosthetic device, including formidable competitors from MIT, a German
team of over 20 scientists and engineers who have received over
$14,000,000 for the German government and a team from Japan who have
recently received government funding. To date, members of the MARC team
Dr. Humayun et al. have been the only ones to electrically stimulate1,2,4
controlled visual percepts human patients. Chapter 2 of my dissertation
provides a treatment of the papers, patents, and prior art embodied by the
various teams' progress, but due to space limitations, only the advantages
of the MARC are presented here.
MARC Component Size: The novel multiple-until intraocular transceiver
processing and electrode array-processing visual prosthesis allows for
larger processing chips (6x6 mm), and thus more complex circuitry. Also,
by splitting the chips up into smaller components, and utilizing
techniques such as solder bumping to connect the chips with kapton
substrates, we shall keep the sizes to a minimum.
MARC Heat Dissipation: The power transfer and rectification, primary
sources of heat generation, occur near the corneal surface, or at least
remotely from the retina, rather than in close proximity to the more
MARC Powering: The novel multiple-until intraocular transceiver-processing
and electrode array-processing visual prosthesis provides a more direct
means for power and signal transfer, as the transceiver microprocessing
unit is placed in close proximity to the cornea, making it more accessible
to electromagnetic radiation in either the visible wavelength range or
radio waves. Solar powering and especially RF powering are made more
MARC Diagnostic Capability: The transceiver unit is positioned close to
the cornea, and thus it can send and receive radio waves, granting it the
capability of being programmed to perform different functions as well as
giving diagnostic feedback to an external control system. Diagnostic
feedback would be much more difficult with the solar powering.
MARC Physiological Functionality: Our device was designed in conformance
with the physiological data gained during tests on blind patients. We are
the only group who has yet created a visual percept (with electrical
stimulation) in a patient. Therefore, we have the unique advantage of
designing around parameters which are guaranteed to work.
Reduction of Stress Upon The Retina: Our device would reduce the stress
upon the retina, as it would only necessitate the mounting of the
electrode array upon the delicate surface, while the signal processing and
power transfer could be performed off the retina. Also, buoyancy could be
added to the electrode array, to give it the same average density as the
surrounding fluid. Approximately 10,000,000 people worldwide are severely
visually handicapped due to photoreceptor degeneration5 experienced in
end-stage age-related macular retinal degeneration and retinitis
pigmentosa. In addition to benefiting the visually impaired, restoring
vision to a large subset of blind patients promises to have a positive
impact on government spending.
4. Description of the five most important technical steps
The honing and development of several aspects of the MARC system must yet
be fully realized so as to optimize the final device's functionality and
performance. Concurrent engineering tasks which are both touched upon and
elaborated in chapters of my dissertation include the following:
The design, fabrication, and testing of the signal-processing and
stimulus-driving MARC2, MARC3 and MARC46 VLSI chips and the
video-processing chip. These are VLSI chips endowed with microprocessing
circuitry to encode and decode visual information, and drive the
The enhancement of the CMOS photodetectors and the Holed Emitter
Phototransistor. These are the fundamental building blocks of silicon
The final designs and optimization of the kapton/polyimide or silicon
stimulating electrode array. Kapton polyimide flexible polymer which would
allow for the fabrication of an electrode array which could conform to the
curvature of the retinal surface. So far it has proven to be
The design and refinement of the RF telemetry system and video protocol.
RF Telemetry is utilized to transmit both power and signal without the
presence of physical wires. Thus the device is entirely self-contained
within the eye.
The bonding, packaging, and hermetic sealing of the CMOS signal-processing
chips with the kapton electrode array. The hermetic packaging of a chronic
device with over 100 electrical feedthroughs is a challenge. The
integration of microelectronics with damaged or degenerated biological
systems in order to provide some of the lost function is a rapidly
emerging field, and we have been and will continue to share technologies
with other groups also working on biological prosthesises.
5. Description of how best to test prototypes
Extensive laboratory and clinical testing will be conducted before
functioning MARC is realized. The doctors on our team are conducting the
biocompatability and threshold-stimulation experiments within both humans
and animals, while the engineers at NCSU-ECE are concentrating on the
testing of the functionality of the computer chips, and the performance of
the RF telemetry transfer of power and signal. Hermeticity may be tested
by submerging device in saline baths for extended periods.
In order to test MARC1, which was endowed with HEP photosensors, the image
of a while paper E mounted on black paper was focused onto the MARC chip.
An adjustable incandescent light was shone onto both black and white
paper, and the difference in reflected power was measured, and found to be
around a factor of ten. This order of magnitude difference is easily
recognized by Mead's logarithmic photodetector circuit. Even though the
image of E was focused down to about 20% of its original size, so as to
fit upon the chip, the difference between the intensities of the
neighboring light and dark areas remained the same, as they were both
multiplied by the same factor.
All the pixels which were subject to the light of the E's image fired,
while those beyond the border remained off. The output from the "on"
pixels, which resulted in 250 mA, 2ms pulses at a 50 Hz clock rate, were
sufficient for retinal stimulation. The photosensing and
current-generating partition of the artificial retina chip has been
tested, and it ahs been demonstrated to work. These results suggest that
the chip would facilitate the perception of outlines where sharp contrast
existed, such as for windows or illuminated text. The Doctors have
demonstrated that the 5x5 electrode array functions, and the next step
towards an artificial retinal prosthesis is to connect the dual unit
visual prosthesis to the 5x5 electrode array, and implant the dual unit
device in an animal, so as to test biocompatibility.
6. Description of the limitations and challenges in the MARC project
The MARC project spreads itself across a diverse array of scientific,
engineering, and medical disciplines. Perhaps one of the greatest
challenges associated with this project is the interdisciplinary nature of
the device's design, which requires the devotion from members of a large,
unified team from a wide array of disciplines and distant institutions.
One of the goals of my dissertation was to aid the project by providing an
overview or synthesis of the wide-ranging research, within the
presentation of the complete system engineering of the MARC implantable
prosthesis. The inter-disciplinary challenge involves the fabrication of
the processing chips, the acquisition and transmission of visual data in a
way that is meaningful to the device and to the patient, a wireless power
source, and a form of biocompatible, hermetically-sealed packaging. The
MARC designs presented throughout my dissertation attempt to integrate the
multifaceted technologies in a final device that will be beneficial to a
As we approach a functioning MARC prosthesis, the design will continue to
evolve, as the refinement of any one parameter affects all the rest. For
instance, should the main intraocular chip be subdivided into smaller
individually-sealed chips so as to reduce the risk of realizing a complete
system failure if one chip should malfunction, the basic chip design, as
well as the hermetic packaging, will have to be altered. An alteration in
the hermetic packaging will affect where the chip may be mounted. A
different chip design will require a different power source and thus
telemetry configuration. And a different telemetry configuration may alter
the coil designs, which would affect the size of the external battery.
Thus an alteration in any one aspect of the design resounds throughout the
entire system. The purpose of this dissertation was to offer an overview
of all the parameters affecting the design of the MARC, elaborate on all
the engineering progress that has been made, anticipate design and
engineering hurdles, and suggest approaches for future research.
The photosensing/current-generating component of the artificial retina
chip has been tested, and it has been demonstrated to work. Investigations
into the feasibility of RF powering have so far been positive. The
electrode design is being honed, and the Doctors have demonstrated that a
5x5 electrode array can stimulate simple pictures upon a patient's retina.
The doctors are currently investigating ways of stimulating the retina
with lower currents, which will have a positive impact on the design of
the chip and RF powering system.
The next step towards an artificial retinal prosthesis will be to develop
the second and third generation MARCs which will be capable of driving a
15x15 electrode array and 25x25 electrode arrays, and testing the devices
for short periods within a human. The implications of this research may
extend beyond this immediate project, as contributions to the overall
field of implantable prosthetic devices and hermetic packaging. The
observations and clinical and engineering experiments performed should
lend insight into the actual functioning of the human retina. The feedback
gained by these studies should provide a vehicle for further understanding
of the retinal/vision/perception process.
In addition, a CMOS phototransistor which exhibits an enhanced quantum
efficiency was also developed, and a numerical model was presented which
also predicts its enhanced efficiency. The enhanced performance is
accounted for via the physics of transistor operation. The CMOS
phototransistor may find an application in the emerging field of CMOS
photodetectors, wherein researchers are attempting to create low-powered
References: 1 E.D. Juan, Jr. Mark S. Humayun, Howard D. Phillips; "Retinal
Microstimulation," US Patent #5109844, 1993
2 M. Humayun, "Is Surface Electrical Stimulation of the Retina a Feasible
Approach Towards The Development of a Visual Prosthesis?" Ph.D.
Dissertation UNCCH BME 1994
3 W. Liu, E. McGucken, K. Vichiechom, M. Clements, E. De Juan, and M.
Humayun, "Dual Unit Retinal Prosthesis," IEEE EMBS97
4 M.S. Humayun, E.D. Juan Jr, G. Dagnelie, R.J. Greenberg, R.H. Propst and
H. Phillips, "Visual Preception Elicited by Electrical Stimulation of
Retina in Blind Humans by Electrical Stimulation of Retina in Blind
Humans," Arch. Ophthalmol, pp. 40-46, vol. 114, Jan. 1996.
5 Research to Prevent Blindness, Progress Report 1993.
6 K. Vichiechom, M. Clemments, E. McGucken, C. Demarco, C. Hughes, W. Liu,
MARC2 and MARC3 (Retina2 and Retina3), Technical Report, February, 1998