Retinal and Cortical Chip implants: Artificial Vision (Vision Prostheses)

The following summary comes from notes taken at the 2002 Eye and Chip World Conference on Artificial Vision held at Lawrence Tech University in Southfield Michigan, June 2002. The Eye and Chip conferences are organized by the Detroit Institute of Ophthalmology (DIO). Video tapes of the three day conference can be purchased from DIO.

For those visitors who want a comprehensive understanding of the entire field of wayfinding and vision substitution technologies, we recommend two websites. Dr. Peter Meijer is a researcher for Phillips Electronics in the Netherlands. Dr. Meijer has tremendous energy and works tirelessly on his own invention for vision substitution called The vOICe. But Dr. Meijer's major contribution to the entire field (in my opinion) is his comprehensive (and highly technical in places) web site; which he updates constantly. The second comprehensive website is our own, at the Institute for Innovative Blind Navigation (IIBN). We update our electronic textbook on wayfinding technology as time permits. IIBN has the advantage of a rehabilitation, consumer, and special education perspective. Also, review The Retinal Transplant Newsletter from Columbia University, Department of Ophthalmology.

Placing computer chips (and components) into the human head is an extremely complex undertaking. Media coverage of developments in the field of artificial vision (or any technology that addresses alternatives to blindness and vision impairment) are commonly over-hyped and misleading. "Sound bite" reporting promotes false hopes and over-simplification.

On the other hand, for the first time in history we do have hope that the machine/body interface will be solved (to various degrees), and that we will have ever more sophisticated medical procedures that address blindness and severe vision impairment. At the moment (June, 2002), we are at the pioneering stage. Important and exciting developments are in the early phases of creation.

The important idea to keep in mind is that this is a very complex problem requiring very careful and sophisticated approaches. To give you an idea of the complexity, I have divided the discussion into the following areas:

These are not complete; I work on this when I get a chance.

01. History; some comments.

02. Research teams in 2002 and their approach to the problem.

03. Issues and variables, includes: demographics; etiology (type and progression of disease states), the complexity of vision (cyborgs; training and rehab issues); measurement; surgery; bio-compatibility, upgrades, tech support, monitoring, and repair.

05. Legal issues.

06. The future: A convergence of technologies: Implants, plus biotech, plus wearable computers, plus robotics, plus the networks (internet, object net, internal area net, space net, personal area networks).

07. Ethical issues.

08. "Fresh" Links. As I become aware of articles and comments concerniong vision implants, I'll post them here.

History

New vision chips (Cortical and retinal) are often compared to the cochlear implant for the deaf. This is misleading and naive for several reasons. First, the media has over-hyped the cochlear implant. As wonderful and "successful" as this "hearing chip" is, many challenges remain (the discussion below of the issues facing implants are relevant to both auditory and vision implants).

More importantly, the challenges facing vision chip designers are way more complex than the challenges faced by cochlear researchers. The retina is a multi-cellular, multi-stage neuro-processing system. It is actually a part of the brain. The processing that takes place at the level of the retina is massively complicated and not entirely understood. Disease processes target regions, or cell types, of the retina, and have complicated evolutions. On a cortical level, there are over 35 centers for vision processing spread over the entire human brain. These centers have complex efferent and afferent networks; meshworks of crisscrossing and redundant neuro "pathways". Foveal signals (those responsible for very high pattern recognition) mapped in the primary cortical vision center (V-1), are located in a fold inside a fold of the occipital cortex. This means that to attain higher resolution it will be necessary to "invade" the brain rather than place the chip on the surface of the cortex.

Therefore, although we have recent historical evidence that sensory chip implants can be placed in the human body (the cochlear implant), the conclusion that vision chips can be as easily created is premature. We are at the pioneering stage.

In the 1960s, Giles S. Brindley of the University of Cambridge in England, attached 80 electrodes to miniature radio receivers and implanted them into the brain of a blind patient. The patient reported seeing phosphenes (flashes of light). William Dobelle began experiments with cortical stimulation in human volunteers during the 1970's. His early implant system was (understandably) bulky by today's standards. Large wires protruded from a hole drilled in the subject's skull and ran to a "large" computer processing system. The patient saw light blobs (phosphenes) in black and white. Some of Dr. Dobelle's first implanted patients have used the system (without infection or severe complication) for over twenty years. In 1992 a blind volunteer at the U.S. National Institute of Neurological Disorders and Stroke learned to recognized phosphene letters (using an implant).

Research Teams

Two teams are on the verge of marketing their implant (probably as you read this, they will have appeared on the market). The Dobelle Institute has a cortical implant, and Optobionics has a retinal chip implant.

There is debate among the teams about the best approach to creating artificial vision. Dobelle belongs to the sector that believes cortical (brain level) implants provide the best answer. Higher level vision processing occurs at various centers in the brain. Placing the chips in selected regions of the brain holds the promise of higher resolution. Placing these chips (eventually) in targeted areas of the head (visual processing centers) might enable improved color, motion, and shape discrimination. At the moment, Dr. Dobelle's team is placing the chips on the surface of the visual cortex (not deep inside the brain, and not near specialized vision centers). The implants were placed so that they stimulated both visual hemispheres of the brain.

The retinal implant teams fall into two camps. One group believes the best approach is to place the chip on the surface of the back of the eye, onto the ganglion cell layer. These teams use chips that are called epi-retinal implants.The second group believes that it is best to place the chips directly in the photocell (rod and cone cell) layer of the retina. These chips are called sub-retinal implants. Teams working on epi-retinal strategies include: The Harvard/MIT group, The Doheny team at the University of Southern California, and Professor Rolf Eckmillers team at the University of Bonn (all the research teams are discussed below). Sub-retinal teams include: Optobionics, The University of Tubingen team in Germany, and the Wayne State University of Medicine (Ligon Center) in Michigan. Cortical implant teams include: The Dobelle Institute, Terry Hambrecht's group at NIH, and Professor Richard Normann's team at the University of Utah. The Ligon group, under the leadership of Dr. Patrick McAlister, is also experimenting with cortical vision chips (in animals so far in 2002).

It is possible that eventually combinations of cortical and retinal chips networked together might be used to provide even higher degrees of visual processing than can be attained using either approach alone. (this is pure speculation on my part).

For the layman, it is necessary to understand the unique (and beautiful) architecture of the human retina. There are five layers to the retina (as determined by looking at slices through a light microscope). In a strange way, the light coming into the eye must pass through four of the layers before it reaches the section of the retina where the photocells receive and react to the light. When the photocells fire, they send signals back toward the surface of the retina, through the four layers above. Various levels of image processing take place as the signals leave the photocells and transverse the other layers. The ganglion cells are the final layer for retinal processing. Their axons (nerve fibers leaving the cell body) are actually the nerves (about a million of them) that leave the retina and travel back into the brain. The first epi-retinal team believes that resting the chip on the ganglion cells (and their axons) is easier to do surgically and avoids the damaged (diseased) tissue below. The second sub-retinal group feels that it is best to surgically "cut" beneath the four layers and place the chip directly in the (damaged) photoreceptor layer.

Dobelle Institute

In June 2002, at the 48th meeting of the American Society for Artificial Organs, Dr. William Dobelle announced that his team had created a commercially available "artificial eye." The cost for the system is $98,000.

The Dobelle artificial eye uses a miniature digital video camera mounted on the lens of sunglasses. Images from the camera are sent to a computer processing unit on the belt. The cortical chip implant is embedded in the brain through a hole drilled in the skull. Cables run out of the head down to a stimulating control system (also on the belt). Inside the head, the chip contains electrodes that penetrate the brain in the area of the primary visual cortex. Images from the camera go to the microcomputer on the belt where the processor determines which electrodes are to be stimulated. Signals are sent to the stimulator unit which activates selected electrodes in the implant. The implants hold sixteen electrode arrays (placed on the mesial surface of the occipital lobe).

Eight patients from six different countries (one each from Canada, England, Italy, Argentina, and Germany, and three from the United States) received the Dobelle implant. The ages ranged from 27 to 77. All the patients had suffered traumatic brain injury that resulted in blindness. The patients had been blind from a low of two years to a high of 57 years (this is the range, the details are on the Dobelle Institute web site). The surgical procedures had no complications and no infections were reported as of April 2002. "The original surgery in 1978 was performed under local anesthesia, and implants in future patients can probably be performed on an outpatient basis by most neurosurgeons."(quotations come from the Dobelle institute website)

Results of the implants varied depending on the unique situation of each patient. Ocular motility is reduced (absent?) in patients with long standing blindness. This reduces the ability to direct the eyes, and reduces the ability to successfully use the implant. Visual fields also varied. Four patients reported color "perception" (Dr. Dobelle speaks of phosphenes, flashes or perceptions of light). The ability to use the cortex to see may be related to when blindness occurred. In one patient (earlier study), where vision had been lost at age five and the chip implanted at age 68, the patient was unable to see phosphenes at all. Dr. Dobelle reports that "the time course of blindness, at least if blinded as an adult, has little effect on the ability to see phosphenes". Patients (using Sobel edge detection algorithms) reported they could see "people in light colored clothing..., could detect straight edges (including a cane and a pencil), and one patient could even drive a car (slowly, on private property).

The system can be disconnected from the camera and plugged directly into a television or a computer, allowing direct access to these technologies. Low resolution does not allow for "normal" viewing of these screens but it does demonstrate how future chip technologies could be directly linked to output devises. A video monitor can also be patched into the system so that observers can monitor what the patient is viewing.

Optobionics Corporation

The optobionics team uses a sub-retinal chip implant. The corporation was founded by Vincent Chow and his brother Dr. Alan Chow (an ophthalmologist; retinal surgeon). The ASR microchip (Artificial Silicon Retina) was collaboratively developed in association with Hines Veterans Administration Medical Center, the Louisiana State University Eye Center, the University of Illinois Eye Center in Chicago, Stanford University's Nano-Fabrication Facility, and Tulane University Medical Center.

The ASR chip is designed to stimulate damaged retinal cells, specifically in patients with retinitis pigmentosa (RP). There is potential for the chip to also be used with patients who have age-related macular degeneration. Any of the retinal diseases that damage the photocell layer of the eye, but leave the other four layers more or less "minimally" damaged may also benefit from the chip. These include the Rod and Cone Dystrophys, Stargardt's Disease, Leber's Congenital Amaurosis, Usher's Syndrome, Best's Disease, long-standing retinal detachments, Choroideremia, and Gyrate Atrophy. In January, 2000, the FDA authorized Optobionics to implant the ASR in patients with RP. Six patients successfully received the implant over a two year period.

The chip is about the size of Abraham Lincoln's nose on a US penny. The chip is 25 microns thick (less than a human hair), 2mm in diameter, and contains about 5,000 independently functioning microphotodiodes. The ASR is powered solely by incident light and does not use wires or batteries. In all the cases, the chips were implanted about 20 degrees superior and temporal to the macula. The chip was tested in animal models before the human trials began.

The FDA study was "to determine the safety, feasibility, and efficacy of implanting microphotodiode-based silicon chip retinal prostheses into the subretinal space of patients with RP to restore visual function." Patients signed legal consent documents that stated that the procedure was not guaranteed to improve their vision in anyway. The implantation of the chip was to determine if it was safe to place the object in human eyes, and whether or not the whole process was feasible as a medical procedure.

Five men and a woman were selected for the study, all with RP, and all with severe vision loss, ranging from bare light perception to "count fingers" ("count fingers" is a uniquely ophthalmological functional and quick "test" to get a raw understanding of gross visual perception. It is not a clinically valid test. See the discussion below on measurement). The patients also were tested before and after implantation using electrophysiology, fundus angiography and photography, automated visual field perimetry, and ETDRS. They also gave subjective responses during interviews.

Although there was no legal expectation of success for the ASR, patients subjectively reported increased ability to see. Some kind of (as of yet) unexplained stimulation caused cells near to the implant to be "rescued." The reason is unclear why the patients seemed to have enlarged fields (beyond the size of the implant's fields) and improved acuity (why these regions were rescued). As a part of the brain, it could be anticipated that plasticity would be a characteristic of the retinal system. We know that binocularity can be lost for a period of time and then restored. We know that people coming out of comas can wake up and see "normally." So, it does seem plausible that retinal regions that had gone "dormant" from disuse, could begin to operate (could be rescued) when signals began to flow anew.

Surgery was successful. The chips worked after implantation, and resulted in no infection, inflammation, rejection, migration, erosion, retinal detachment, nor implant breakdown (over a two year period). Both subjective and objective improvements in visual function were documented. The surgical procedure, which takes about two hours on an outpatient basis, is described on the optobionics website under the clinical trials section.

University of Bonn

University of Tubingen

This team, lead by Dr. Eberhart Zrenner at The University of Tubingen in Germany, uses a chip to replace the photoreceptors (sub-retinal approach). The chip responds to light that then activates the healthy bipolar and ganglion layers of the retina (the processing layers). Dr Zrenner describes several advantages to the sub-retinal implant: degenerated photoreceptors are replaced by microphotodiode arrays (MPDA's) that can activate the retinal processing layers; positioning and fixation of the MPDA's in the subretinal space is relatively easy; no external camera or external image processing is required; eye movements can be used for localization of objects; and MPDA's are well tolerated by the inner retina. The subretinal approach only works when the optics of the eye are functional, and where optic nerve transmission is normal. This strategy targets RP and macular degeneration.

The Tubingen team lists several challenges that remain unanswered in 2002:

01. Although it has been consistently shown that the chips do result in stimulation of visual cortical regions, it is not clear how well they activate regions responsible for orientation and movement ( so that people with the chip could navigate).

02. Implant studies have shown some breakdown of the silicon chips over time. It is not clear what very long term affects would have on the implants.

03. Although retinal neurons do respond to the chip, it is not clear how well these neurons will tolerate long term electrical stimulation without morphological and/or functional alteration.

04. We can only speculate about what kind of image a patient would have using the MPDA's.

The Tubingen team is a combination of expert staff from the University Eye Hospital at Tubingen, the University Eye Hospital at Regensburg, the Natural and Medical Sciences Institute at Reutlingen (NMI), the Institute for Microelectronics in Stuttgart (IMS), and the Institute for Physical Electronics at Stuttgart (IPE). Dr. Zrenner is Director and is a Professor of Ophthalmology at the Tubingen Eye Hospital. The team has been collaborating since 1995 and has produced several prototypes of subretinal chips.

Doheny Retina Institute: Department of Ophthalmology; Keck School of Medicine, University of Southern California

Early research on implantable artificial vision systems was done at John Hopkins University. Several members of that pioneering team have moved to the Doheny Institute to carry on their research. This team is developing an epi-retinal system. Funding for this team is extensive and includes The Alfred E. Mann Fund at the Applied Physics laboratory (for Drs. Mark Humayun, and James Weiland), DARPA (Tissue based biosensors program; Drs. Humayun, Weiland, and Gislin Dagnelie), Foundation Fighting Blindness (microelectrode stimulation and recording from the mammalian retina), NIH and NEI (retinal electrical stimulation in photoreceptor loss), NSF (the realization of a retinal prosthesis for the totally blind, and Implantable multiple unit visual prosthesis, 2nd and 3rd generation).

Wayne State University Ligon Research Center of Vision

The Michigan team is working on both retinal and cortical vision chips. The retinal chip has a unique activation and delivery system, using chemical activation at the photocell layer employing a neurotransmitter. This is actually a micro-drug-delivery system. Dr. Gregory Auner, Director of WSU's Smart Sensor lab, describes their invention as " a microfluidic drug delivery system with an advanced high frequency microarray chip for retinal stimulation."

Dr Gary Abrams is the Director of the Ligon Center. Other members of the team include: Dr. Auner, Dr. Fernando Diaz (chair of the WSU Department of neurosurgery), Dr. Linda Hazlett (chair of the WSU Department of Anatomy), Dr. Raymond Iezzi (Director of the retinal implant project), Dr. Robert Johnson (neurosurgeon at the Detroit Medical Center working on cortical implants), Dr. Patrick McAllister (head of research in the WSU Department of Neurosurgery and director of the cortical implant project), Dr. Loren Schwiebert (Assistant Professor and researcher Department of Electrical and Computer Engineering), and Dr. Pepe Siy (Associate Professor and researcher, Department of Electrical and Computer Engineering).

Dr. Lezzi feels that several major challenges face retinal implant teams. They need a better understanding of the anatomy of the normal and of the diseased retina. They need a solid understanding of the materials that will be used in the chips; the biocompatibility issue. To be effective over the long term, these implants need to do no harm to the retinal tissue and they need to avoid erosion and breakdown. The chips will have to become very sophisticated and will have to learn to speak the language of the retinal level vision processing systems. Like all of the teams working on the artificial vision chips, there is a strong belief that this undertaking is entirely achievable over a reasonable length of time. The Michigan team is hoping to test their evolving chips in humans within a few years.

Concerning cortical implantation, Dr. McAllister feels that chip design will have to respond to the brains plasticity. Over time, the brain adapts and responds different to stimuli. The chips will have to evolve (in place; be re-programmable) to adjust to this plasticity. As patients become skilled in the use of the implants, they will need a way to "tune" the technology.

At the engineering level, the Michigan team is exploring three approaches: a sub-retinal chip similar to the Optobionics system, an epi-retinal chip, and a cortical chip. Dr. Auner feels that the sub-retinal chip holds the greatest promise at the moment.

A UPI news release in September, 2002 reports on the Wayne State approach and it's connection to work being done at Stanford. These news reports tend to be removed from the web and end up as bad web links. I've copied the report below:

Docs outline artificial vision progress

By Ed Susman
UPI Science News
From the Science & Technology Desk
Published 9/23/2002 11:59 PM
WASHINGTON, Sept. 23 (UPI) -- Artificial devices that may allow the blind to see could be available for human use within a decade, scientists said Monday.

Already researchers said they have begun experiments in animals with devices that involve use of electronic microchips, which allow cells to grow onto their surface, mimicking the natural molecular process in which light reaches the retina of the eye and produces signals that are turned into vision by the brain.

"Our goal is to develop a high resolution, neural chip that connects a signal from a digital camera to individual nerve cells in the retina of patients," explained Harvey Fishman, director of ophthalmic tissue engineering at Stanford University School of Medicine in Palo Alto, Calif.

Fishman said the devices would target individuals blinded by age-related macular degeneration or those with retinitis pigmentosa, two diseases that blind tens of thousands of people in the United States each year.

Fishman's artificial synapse chip would be likely be coupled with a system, similar to that being developed by Raymond Iezzi, assistant professor of ophthalmology at the Kresge Eye Institute at Wayne State University in Detroit.

The scientists outlined their devices at a symposium sponsored by Research to Prevent Blindness.

"We are developing a new type of visual prosthesis implant that will mimic normal chemical signaling between neurons in the retina and brain," Iezzi said. "Our overall hypothesis is that digital images may be transposed into neurochemical signals through an implanted microchip, that provides chronic, meaningful visual information to the retina and brain."

Artificial vision implants have been a staple of science fiction novels, television and movies for decades, but Karl Csaky, senior investigator at the National Eye Institute in Bethesda, Md., said rather than science fiction, the ideas presented at the symposium by Fishman and Iezzi represent the infancy of the science of restorative vision.

"I think that within a decade crude implantable devices along the lines that were discussed today will be available for people," he told United Press International. "These prostheses may give people who have no vision the ability to see movement, to make out large forms."

Csaky said with time and refinement the devices might make it possible for blind people to read or to carry on normal activities of daily living, but that would be much further down the road.

Iezzi is developing "caged" neurotransmitters, essentially a drug delivery system that would unleash molecules activated by light. When exposed to light, the molecules would respond in less than a nanosecond, releasing chemicals that transmit signals to the brain.

Iezzi said he expects to be able to begin animal studies of his camera-brain-computer neurotransmitter system in the next 12 months.

Although Fishman conceded his work still requires answers to some questions, he is enthusiastic about the project.

"The overall potential of these devices is unlimited," Fishman said. If the concept which is now just moving into testing in rabbits proves successful, Fishman suggested that in addition to a digital camera device that gives general local vision, the patients with the artificial prostheses could be hooked up to binoculars or even telescopes that would give them super vision.

Different types of interfaces might even be able to give these people infrared vision as well, Fishman said.

Csaky said such developments were, of course, many more years along the scientific path.

Pritzker Institute of Medical Engineering, Illinois Institute of Technology

Research and development at the Pritzker Institute is focused on a cortical visual prosthesis that uses penetrating microelectronic arrays. The research team does not feel at this time that we have sufficient understanding to confidently place these emerging technologies into the human brain. Quoting Dr. Philip Troyk, Director of the Neuro-engineering Program at IIT (From the Eye and Chip documents): "Previous studies are insufficient as a basis for considering whether proceeding to chronic human implantation is warranted. The overall objectives of our multi-institutional team-based project are to advance the technology sufficiently to provide a reasonable expectation of reliability and safety for implantable hardware, and to develop an animal model to perform crucial pyschophysical and electrical stimulation studies, so that a multi-model decision process about proceeding to human volunteers can be defined."

Dr. Troyk identifies three areas where fundamental questions remain:

1. Questions of physiology: How do we maximize the information transferred from the artificial system to the brain? What's the best way to deliver a stimulus through the electrode array? How can stimulation through multiple channels be patterned to optimally control perception?

2.Questions about electrode technology: Can we design, fabricate, and implant an electrode array(s) in the human brain over a long period of time? Can we supply continual current to the neurons through this large collection of electrodes? Can it be done safely and efficiently over the long time frame required?

3.Questions about the stimulation units: "Can reliable modular implantable electronic packages, capable of driving large numbers of electrodes, via transcutaneous RF power and bi-directional data links, and suitable for surgical implantation, be designed and fabricated?"

Harvard/MIT: The Retinal Implant Project

This project began in 1988 as a joint venture between the Massachusetts Eye and Ear Infirmary and MIT. This team uses an epi-retinal approach. As with all the epi-retinal systems, an external mini-camera (usually attached to spectacles) is used to transmit digital images to the retinal chip. The Harvard/MIT group uses two microchips in their unique design. One of the chips is a small solar battery unit which receives light from a tiny laser mounted on the glasses (to power the chip). This chip also provides the digitally coded visual picture that is received from the miniature video camera on the glasses. The second chip is implanted on the ganglion cell layer of the retina. It decodes the digital imput from the camera and selectively stimulates the retinal surface layer.

Since 1988, the Harvard/MIT team has done research on the response of ganglion cells to electrical stimulation. They have explored surgical avenues for safe implantation, measured visual cortical responses to retinal implant stimulation, and experimented with a variety of systems that make up the epi-retinal unit (laser transmitters, mini-cameras, solar batteries, chip designs, stimulator components, etc.). All intial studies have been done with animal models. Recent research is being conducted using human volunteers. Implants are not left in an eye after experimentation because the team does not feel that the bio-compatibility issue has been adequately resolved.

Seoul National University, Korea

The Seoul Artificial Eye Center was established in 2001 by two ophthalmology professors at the National University College of Medicine in Seoul, Dr. Jin-Hak Lee, and Dr. Hum Chung. Their work on the artificial retina takes place at the Nano Bioelectronics and Systems Research Center, which is funded by the Korea Science and Engineering Foundation. The team is developing a polymide electrode array (PEA) and exploring methods of surgical implantation. They are also measuring the biocompatibility of PEA. Retinal tacks have been an issue for epi-retinal attachment for all the research teams. The Korean group is experimenting with silicon based retinal tacks. A vertical cavity surface emitting laser (VCSEL) was selected to transfer power and signals to the artificial retina. This is a wireless system. The group is testing the feasibility of using the VCSEL. Testing has occurred in a rabbit animal model.

Early test results in the rabbit model show that PEA had no harmful effects on the RP retina after 12 weeks of implantation. The silicon retina tack also worked successfully over the 3 month period of testing. VCSEL was shown to be a promising candidate for use in a retinal prosthesis".

Australian Vision Prostheses Group

Oak Ridge National Labs; Tennessee

University of Utah

NIH neuroprosthesis

Nidek Technologies, Japan

University of Houston

Issues and Variables

Demographics

Children can process upside down images of faces until about age 8 or 9 at which time this ability is lost. Protective and developmental reflexes appear and then disappear as the human child grows. The age at which we implant a chip is critical. Implants are not placed in a static system. They are placed in a dynamic, ever changing, ever evolving, matrix. When a chip is placed in an eye or visual cortical center it will impact in some way with human growth and development.

Approximately 48 million people are blind in the world according to the World Health Organization. Forty million of these people are needlessly blind. Most of them have cataracts which could be removed (and vision restored) with a fifteen minute outpatient operation. More people are becoming blind from cataracts than people getting sight back from operations, so the number of blind people in the world is increasing. We have a pandemic on our hands. Blindness will only be eradicated when the injustice and the poverty that causes unequal access to health care is overcome. Hundreds of millions more people cannot even get an adequate pair of glasses. Approximately one hundred and thirty million people have low vision (glasses will not correct their eye problem). This group, like the forty million needlessly blind individuals have no access to special education or rehabilitation. It cannot be said therefore that artificial vision in any form (surgically implanted or worn on the body) will eradicate blindness.

Patients with retinal disease are a relatively rare group (on a global scale). These individuals do not (by nature of their retinal disease) have damage to other parts of their brains. Most of them have normal intelligence, normal motivation, and a full capacity to functionally adjust to their very restricted disease. They tend to be (as a group) very good at navigating blind and at using assistive technologies to circumvent their impairments. They tend to be users of high technology. The RP patients also tend to be a young group.

When we subtract the 40 million needlessly blind individuals, and examine the demographics of the group that is left, we find that blindness is (in the developed nations, where there is a high level of health coverage) a disease of the multiply handicapped. A large number of the remaining people who are blind are senior citizens who have associated orthopedic, auditory, and cognitive impairments. Another large category of blind individuals in the industrial nations have diabetic retinopathy. Diabetes affects the entire body; it is a systemic multiple impairment. Diabetes affects the nerves in the eye and brain and therefore cannot be addressed with the current state of artificial vision implants. We cannot just treat the eyes of diabetics without potential severe complications.

The final large category of blind individuals in the developed world is the severely multiply handicapped population. The brain is a tough, resilient set of organs. Vision centers and/or connecting neuropathways are found in every region of the brain. To become totally blind, the brain must undergo severe trauma that is wide in extent. This means that blindness is often associated with severe cognitive and/or orthopedic damage. Many visual processing regions are usually affected. No retinal or cortical chip will solve this problem.

Age related macular degeneration impacts 700,000 Americans every year. It is the leading cause of legal blindness (central vision loss) in the developed nations. As the number of older individuals increases proportionally, so will cases of macular degeneration. If the implants could be used successfully with this population, it would be a major medical accomplishment. So far, the implants have only targeted patients with RP. Retinitis Pigmentosa affects over a million and a half individuals worldwide. It is the leading cause of inherited blindness.

Retinal and cortical vision chips are part of a larger revolution. The implantation of smart machines into the human body will continue at a blistering pace. The interface between damaged tissue and smart machines will eventually be successful. Smart machines will communicate with each other, monitor their own operation, self repair, or call for repair. The merging of man and machine is the future.

Etiology (type and progression of disease states)

The design and success of an implant depends on the situation in the eye or cortex at the time of surgery. Most of the retinal implants are (currently) targeted at RP patients. There are many kinds of RP depending on the location and extent of genetic damage, some appear sooner in life, some later. Some forms are severe and acute, other forms are more chronic and play out over a longer time frame.

The success of a retinal implant depends on how much retina is left (that is "healthy" enough to receive a chip). There are stages of disease progression in RP. A remodeling (of the five stage retinal cell layer architecture) and a rewiring of the neurons occurs as the disease progresses. Blood vessels also reposition and relocate as the PR unfolds. The astrocytes and meuller cells (especially) proliferate as the disease advances. No two eyes are ever alike. The success or failure of an implant depends on individual circumstances.

A very chemically powerful neurotransmitter called glutamate is associated with retinal photocells. Glutamate works fine as long as it is contained inside it's normal cycle of operation. When disease processes begin in the retina, glutamate is released into tissues that it was not designed to come in contact with. In effect glutamate is "dynamite;" it becomes a cellular toxin, poison to everything around it. Glutamate is somehow involved in the breakdown of the photocell layer in RP. But what makes RP patients eligible for retinal implants is that, although the rods and cones deteriorate (from glutamate poisoning?), 50% to 75% of the ganglion cells survive, and about 80% of the bipolar and amacrine cells survive. In other words, RP is a disease that targets one layer of the retina severely (only about 5% of the photoreceptor "survive") but more or less spares the other (processing and transmission) layers; ie. only the sensors are (initially) damaged severely in PR. Other complicating factors with RP include a marked decrease in blood flow (circulation drops 80%), a breakdown in the layers (so that it is difficult to determine them), and a progression in some cases to a retinal soup that leaves no viable retina for implantation (end stage of some varieties of the disease). For an implant to work, there must be preservation of some retinal layers.

The progression (etiology) of many retinal, visual tract, and visual cortex diseases (and damage) cannot currently be addressed using these pioneering (retinal) implants. These ("untreatable by implant") disorders include: retinopathy of prematurity (ROP), glaucoma, diabetic eye disease, optic nerve damage, and impairments caused by vascular damage. The potential for implant success is greatest where the vision loss is discovered early and where specific cell layers or regions are involved (RP, macular degeneration from age, and Leber's Amaurosis). Cortical implants by-pass the retina and could conceivably address vision loss caused by damage to systems anterior to the occipital lobe.

The complexity of vision

Cyborgs

Eagles have 20/5 vision. Iguana's can see 360 degrees. Butterflies see in ultraviolet light. Bumble bees can freeze frame moving objects with 300 pulses per second (humans are capable of 50 frames per second movement resolution). Cats can see 8 times better than humans at night. At the University of Toronto's Personal Imaging Lab Dr. Steven Mann experiments (and creates) cyborg systems for human beings. Dr. Mann uses a technology called Eyetap to feed computer modified images to the eye. These images can be adjusted to feed ultraviolet or infrared images to the eye, 360 degree fields (or vision out of any part of the body), altered frame processing speeds, digitally enhanced image resolution, and more. Dr. Mann's work is with wearable computers that can be networked. This creates a personal area computer network (PAN), turning the human body into a computer processing system. What will happen (what could happen) when we get good at implanting computer chips inside the body? Using Dr. Mann's work we can begin to envision logical scenarios.

If we create complex retinal and cortical chips that are sophisticated enough to process signals at various levels of complexity, we should also be able to create (or modify) internal chips to enhance human capability. We can offer insect vision, and iguana vision, and improved acuity and night vision. We can network the implants together (an internal area network; IAN). We can link the IAN with a PAN and with external networks like the internet and object nets.

Training and Rehabilitation Issues; Consumer Responses

In these pioneering stages of chip implantation, researchers are attaining very low acuity. Researchers, from a rehabilitation perspective, are taking people who are blind or nearly blind and making them severely visually impaired. That is the reason why the researchers are claiming that their goal is to improve the mobility of these patients. The claim is that the recipients of the (early) implants will not read or see faces well (if at all), but will be able to use shadows and gross form perception to avoid objects in their path. From a disability perspective, the rehabilitation strategies used with these implanted patients is no different from that used with blind individuals. They will still need to learn blind travel skills, and still need assistive technologies.

It is not well understood that blind individuals (those with no secondary impairments, like most RP patients), usually have excellent mobility using non-visual travel strategies. Indoor mobility is easily mastered by blind individuals. With training and practice, blind adults could travel all over the world with minimal assistance (many do). Quite often, the totally blind individual travels better than the person with low vision. This is because there is a powerful (natural?) impulse to use the vision system for navigation. A visually impaired individual will struggle with seeing using a poor resolution visual perceptual system when the could much more readily use the other senses to navigate fluidly (if they block out the misleading visual signals). So, giving blind individuals severely low vision (at this pioneering stage) is a dubious accomplishment (ie. we should be careful about proclaiming the benefits).

Also, many blind individuals (in my experience; especially blind teenagers) will not even carry a cane because it draws attention to their blindness. These individuals refuse to use available technologies that are unusual looking, bulky, heavy, or expensive. This group will not be lining up to receive any kind of implant that is obvious, ugly, or unusual looking.

What the implant researchers and developers can claim is that they are pioneering a new source of hope. Implants are another set of developing tools to offer consumers.

Measurement

Measurement is a complex issue reflecting the complicated design of the human vision system. The best way to attack the problem is to use the United Nations definitions that were established during the decade of the brain. The UN differentiated the terms "impairment, "disability," and "handicap." Each of these categories are the domain of specific professions. Each has a unique set of measurement strategies.

An impairment is damage to a body part. Measurement is restricted to the body part. Medicine is the domain of impairment assessment and remediation. Blindness is an impairment to both eyes resulting in total loss of vision. Doctors do all they can to preserve vision, using medications, surgery, and whatever technologies they can provide (like the emerging artificial vision implants). When doctors examine the eye, they are measuring it's ability to correctly refract light, it's ability to move through an accepted range of motion, it's ability to focus and for the two eyes to converge and work binocularly (etc.). The doctors can speculate with some accuracy about the functional ability of patients, but they do not enter the realm of the disability specialist where actual functional skills are considered.

A disability is an inability to accomplish an everyday task, like being able to read print visually, seeing a traffic light change from the corner of an intersection, reading a chalk board in a classroom, identifying a face from various distances under various environmental conditions. Disability measurements take place in the real world in the various atmospheres of that real world. They measure how well the introduction of technologies impact real world vision; how well a telescope helps a visually impaired individual see a house number or real a store logo; how well various kinds of lighting improves the ability to perceive real world objects like faces, letters for reading books, letters for reading signs, clock faces, etc. Disability measurement is the domain of orientation and mobility specialists, special education teachers and rehabilitation specialists.

A handicap is a social role that cannot be attained because of one or a set of disabilities. The inability to get certain kinds of jobs (airline pilot) is a social role that cannot be accomplished by blind individuals. The ability to find a mate and raise a family is complicated by blindness, as is the ability to serve in the military or to serve in various leadership roles. Social stigmas, insensitivity by sighted individuals, and cultural ignorance lead to handicapping conditions for people who are blind. Measurements of handicap are quality of life, subjective responses. These are every bit as valid as the finest psychophysical measurements or the functional vision examines of the disability specialist. Social workers, rehabilitation consultants, and social service agency specialists administer these tests.

To determine the value of any new technology, it is important to test responses in all three domains using the experts from these professions.

Surgery

The eye is only about one inch in diameter. Retina implants have to be very small and be surgically placed precisely into a thin strip of tissue (the retina is sometimes compared to a wet slice of Kleenex). To get to the retina, surgeons have to remove much of the internal part of the eye. The lens may have to be removed and replaced, and the vitreous is removed and later a salient solution is injected as a replacement. The skill of the surgeon is very important in these delicate operations

At the Eye and Chip conference, Dr. McAllister (Wayne State University) showed slides from animal experiments in which cortical tissue was distorted by the weight and placement of implants (these were implanted deeper in the brain, not on the surface). This raises concerns about long term effects of implants on tissue morphology, especially for deep brain implants.

Bio-compatibility

Upgrades, Tech support, Repair

Monitoring

Legal Issues

The future: A convergence of technologies

Ethical Issues

Many magazines and newspapers are watching this story unfold. Take a look at Wired Magazine's comprehensive article in September, 2002 about artificial vision, called Vision Quest. As news releases come out I'll list links below to try and stay current with this rapidly evolving technology.

*************

"Fresh" links:

Sandia Labs News Release: Enabling the blind to see, a task once thought the province of miracles, is the goal of a technical team that includes Sandia National Laboratories, four other national labs, a private company, and two universities.

The idea, funded by a $9 million, three-year grant from the Department of Energys Office of Biological and Environmental Research, is to create 1,000 points of light through 1,000 tiny MEMs [microelectromechanical systems] electrodes. The electrodes will be positioned on the retinas of those blinded by diseases such as age-related macular degeneration and retinitis pigmentosa. These diseases damage rods and cones in the eye that normally convert light to electrical impulses, but leave intact the neural paths to the brain that transport electrical signals. Eventually the input from rods and cones ceases, but 70 to 90 percent of nerve structures set up to receive those inputs remain intact.

*************

Below is an article that appeared in September, 2002 in the Los Angeles Times under the title "Vision of the Future". It is about the Dobelle brain implant as well as retinal implants.

Vision of the Future

Researchers are on the right track to produce artificial sight for the blind.

By THOMAS H. MAUGH II. TIMES STAFF WRITER. September 16 2002.

Jens was a 17-year-old nailing down railroad ties when a splinter broke off and punctured his left retina, blinding him in that eye. Three years later, while he was fixing a snowmobile, a splinter of metal broke off from the clutch, destroying his other eye.

Now 39, Jens lives in rural Canada, where he splits and sells firewood, sometimes loading as much as 12,000 pounds a day. He and his wife have eight children, and when he is not working in the forests, he programs computers, tunes pianos and gives an occasional concert.

But he wanted to see again so badly that Jens - he prefers not to have his last name known - recently paid an estimated $115,000 to have surgeons in Portugal drill a hole in his skull and place an array of electrodes on the surface of his brain. The electrodes are connected to a miniature television camera and a sophisticated computer that, together, have given Jens a rudimentary form of vision.

This summer he demonstrated his new sight at a New York meeting by navigating through rooms, finding doors and even driving a car briefly around a parking lot, avoiding a trash bin and other obstacles placed in his path.

"This is pretty crude vision right now, if you want to compare it to what I had before," Jens said. "But fortunately, there is a good improvement compared to being blind totally."

Researchers have been struggling for more than two decades to produce some kind of artificial vision for the blind, and they are now on the right track. For now, Jens is one of only a handful of people who have received newly developed treatments to restore vision, and few experts expect the number of recipients to grow rapidly any time soon. But the fact that anyone has been treated at all represents a major technological breakthrough.

Reflecting that newfound optimism, nearly a dozen labs throughout the world are racing furiously toward a common goal - bypassing damaged components of the visual system to restore sight. Some are developing artificial retinas, some are using electrodes to stimulate neural pathways in the eye, and still others are trying to stimulate the brain directly. Few, however, have done much in humans, yet.

"We're still at a very primitive stage, in the sense of how much work will be required to ... produce enough vision for mobility," said Dr. William Heetderks of the National Institutes of Health. "It's very easy to underestimate the technical problems involved." Although the prosthesis used by Jens is now being marketed to consumers, experts predict that it will be a decade before other devices see much use.

The need for technological breakthroughs has never been greater. According to a recent report from the National Eye Institute, more than 1 million Americans over 40 are blind, and an additional 2.4 million are visually impaired as a result of diabetic retinopathy, age-related macular degeneration, cataracts or glaucoma. Those numbers are expected to double over the next 30 years as the baby-boom generation ages. New drugs and other therapies are delaying the onset of blindness for many victims, but once blindness sets in, there has traditionally been nothing more that could be done.

Now that situation is changing, as is illustrated by work in three labs that seem to be ahead of the field.

Miniature Electrode Array

At USC's Doheny Retina Institute, Dr. Mark Humayun and Dr. Eugene de Juan Jr. have developed a miniature electrode array that can be implanted in the eye to replace a damaged retina. The array is attached by thin wires buried under the skin to a radio receiver that is implanted behind the ear.b

Visual signals from a video camera are processed through a microcomputer worn on a belt, then transmitted to the receiver. The retinal array stimulates optical nerves, which then carry a signal to the brain. The signal is perceived as phosphenes, bright points of light. With correct stimulation, patterns of phosphenes can draw a picture in the mind similar to that seen on a stadium scoreboard, where letters and pictures are produced by arrays of individual lightbulbs.

The preliminary results have been "encouraging," Humayun said. "The brain can make a lot of sense out of crude inputs."

The team has implanted the devices temporarily in 17 patients to date. When they used as few as four electrodes, De Juan said, patients could tell if an object was in front of them, moving from left to right or vice-versa, and so forth. With a four-by-four array of 16 electrodes, patients can see shapes and outlines and pour a liquid from one cup to another.

"To read, we may need 1,000 electrodes," Humayun said.

In February, the USC researchers performed the first permanent implant of a four-by-four electrode array in a patient, and they plan to do a total of three this year. The device is manufactured by Second Sight LLC of Valencia. Sandia Laboratories is developing second-generation versions of the devices.

The first patient "is doing much better than the patients with the short-term implants - far better than we expected," Humayun said. The brain, he added, is constantly learning how to interpret the visual signal. "It meets us halfway."

If the first three implants are successful, De Juan and Humayun have permission from the Food and Drug Administration to implant another seven. But completing the test will take a long time, Humayun cautions, because "it's a Class III device - the highest-risk device according to the FDA. That's because it's an implant that will be left in for the rest of the person's life."

Eventually, De Juan said, they would like to shrink the device so that everything would fit into the eyeball. Similar technology is being developed at the Massachusetts Eye and Ear Infirmary in Boston and at the Catholic University of Louvain in Belgium.

Tiny Artificial Retina

Dr. Alan Chow of the University of Illinois at Chicago Medical Center has produced a completely implantable artificial retina, but there are substantial questions about how well it works. Chow and his brother Vincent, president of Optobionics Corp. of Wheaton, Ill., have developed a silicon chip that is 2 millimeters in diameter - smaller than the head of a pin - and half the thickness of a sheet of paper.

The chip contains about 5,000 small solar cells, each attached to a miniature electrode on the back of the chip. The idea is that light falling on the chip will activate the electrodes, stimulating the optical nerves behind the retina. Most critics, such as Heetderks, do not think the solar cells can generate sufficient electrical power to activate the nerves.

Nonetheless, Chow has implanted the devices in six patients. "All of the patients have improved visual function, sometimes quite dramatic," Chow said.

One patient, for example, had no light perception previously but "can now see people standing in front of her." Another patient, "who could read no letters on an eye chart, is now reading about 25" letters on the chart.

"We're pretty excited," he added. "We hope these improvements persist."

But Chow concedes that the implant may not be working the way it was designed to. Retina cells that are physically separated from the implant "seem to have improved in function in all the patients," he said. That suggests that the surgery may have triggered the release of a growth factor or some other chemical in the eye that led to regeneration of retinal cells.

Whatever is happening, Chow plans to implant four more of the devices and monitor the progress of all 10 recipients. "If [the implants] seem to be consistently effective and safe, the study will be expanded to an undetermined number of patients."

A significant number of the blind, like Jens, have badly damaged optic nerves, however. For them, a simple retinal implant will be of no benefit, and a more invasive procedure is required.

Skull Implant

Electrical engineer William H. Dobelle of the Dobelle Institute in Commack, N.Y., has been attempting to produce such a device for nearly three decades. He has developed a thin array, containing 64 electrodes, that is implanted inside the skull on the surface of the occipital lobe of the brain. There the electrodes directly stimulate the visual cortex. The device is connected to an electrical socket that passes through the skull and skin.

Outside the skull, the device is similar to the retinal stimulators, with a television camera mounted on glasses and a small computer to process the signal. This is the device that Jens wears.

Dobelle's team has now implanted the devices in eight patients from six countries. One patient, blind from birth in one eye, lost the other at age 45. Another, who was 77 at the time of the surgery, lost both eyes in a mortar attack during World War II. For some of the patients, insurance covered all or part of the cost of the procedure; others paid the full $115,000 themselves.

Because of U.S. Food and Drug Administration restrictions on implanting medical devices in the brain, the surgical procedures were performed at the University of Lisbon Medical School in Portugal. Arrays were planted on each side of the brain.

The implants produced "excellent displays of phosphenes" in all eight of the patients, Dobelle told the New York meeting of the American Society for Artificial Internal Organs, and four of them reported that the phosphenes were brilliantly colored. The devices are primarily designed for mobility rather than reading, Dobelle said. But "rapid advances provide the possibility that the patients will be able to scan the Internet and watch television," he added.

Implanting the arrays on the brain's surface creates many problems, however.

A moderately high voltage is required to stimulate the target cells inside the brain, and it is not possible to stimulate small groups of cells specifically. The voltages applied could, in some cases, provoke epileptic seizures.

The National Institutes of Health has thus been sponsoring studies by researchers such as Dr. Richard Normann of the University of Utah and Dr. Phil Troyk of the Illinois Institute of Technology in which the electrode arrays will be implanted in the interior of the brain, directly in contact with the cells to be stimulated.

One key difference is that the voltages required to stimulate the cells reduced by a factor of hundreds to thousands. "As a rule of thumb, when something changes by a factor of 10, you are in a different ballpark, and here we are talking about hundreds or thousands," said NIH's Heetderks.

"We need to put in structures that are comparable in size to neurons," Normann said. "And the materials themselves have to be regarded by the brain as not overtly hostile. It's a challenging problem," but when they succeed, surgeons will have a tool that can be used to treat not only blindness but also deafness, incontinence, spinal injuries and other problems, he said.

Researchers in all these areas still have a long way to go, but the progress is heartening. In the long run, many researchers believe that electrodes implanted deep within the brain will provide the best results, but the other alternatives may provide benefits to blind people who cannot wait that long.

"I believe that I am safe in saying that Braille, the long cane and the guide dog are doomed to obsolescence," Dobelle said. "By the end of this century, they will be as obsolete as the airplane made the steamship."

Los Angeles Times source URL:
http://www.latimes.com/features/health/la-he-eye16sep16002051.story?null

Similar article in the Toronto Star, via the (very long) source URL:
http://www.thestar.com/NASApp/cs/ContentServer?pagename=thestar/Layout/Article_Type1&c=Article&cid=1026145714801&call_page=TS_Life&call_pageid=991479973472&call_pagepath=Life/News&col=991929131147

*************

Lateral Geniculate Body Implant

R. Clay Reid, associate professor of neurobiology at HMS, is one of three new recipients of a Kirsch Investigator Award, which recognizes innovative research with a potential for significant breakthrough. The program was started in 2000 by the Kirsch Foundation.

"Each investigator works in an area that has demonstrated the potential for significant breakthroughs in the near term," said Kathleen Gwynn, president and CEO of the foundation.

Reid will receive $200,000 over the next two years to study electrical stimulation of the visual thalamus as a means toward a visual prosthesis. Jonathan Tilly, HMS associate professor of obstetrics, gynecology and reproductive biology at Massachusetts General Hospital, will receive a second year of funding. This will be the third year of funding for Ronald DePinho, HMS professor of medicine (genetics) at Dana-Farber.

Source URL: http://focus.hms.harvard.edu/2002/Aug30_2002/bulletin.html

Abstract of Research Plan. Provided by Dr. Reid.

For those who have lost vision due to accident or to degenerative retinal diseases, restoration of visual function would mean a tremendous increase in the quality of life. Spurred by the success of cochlear implants - which have provided limited hearing to patients with several forms of deafness - there has been a resurgence of work on devices to stimulate the visual system electrically. Most recent work on visual prostheses has been aimed at developing intra-ocular devices for retinal stimulation. We plan instead to stimulate the lateral geniculate nucleus (LGN), the part of the thalamus that relays signals from the retina to primary visual cortex.

Our approach will be to implant bundles of microscopic electrodes in the LGN of animals that can later be studied in the alert state. These electrodes can be used both to record the normal activity of LGN neurons and to stimulate the neurons electrically. This project combines two lines of research from our laboratory: multi-electrode recordings from neural ensembles in the LGN, and studies of visual cortical responses to electrical stimulation of the LGN. The advantage of working in alert animals is that we can examine the physiological responses of visual cortical neurons at the same time that we explore the perceptual correlates of electrical stimulation. The long-term goal is to restore low-resolution but usable visual percepts in otherwise blind or low-vision patients. Given the current state of brain-stimulation technologies and our strong understanding of the early visual system, this goal should now be within reach.

*************