Researchers race to produce bionic vision

By Daithí Ó hAnluain
THE race to develop a bionic eye has caught the public imagination, and fed the urgent hopes of millions of potential patients. While researchers have made enormous progress over the last 15 years - already 22 patients in the US and Europe are using prototype prostheses involving three different technologies - fundamental problems remain.
There are more than 10 major groups studying four fundamental approaches that could, in the long term, cure most forms of blindness. At present, the most promising approaches apply to retinitis pigmentosa (RP), 'dry' age-related macular degeneration (AMD) and Usher's Syndrome, conditions where the photoreceptors degenerate but the optical 'wiring' is still in place.
The four approaches include subretinal implants, epiretinal implants, optic nerve and cortical prosthetics. Within each approach there are competing teams, and even within teams, variations and novel solutions exist. Subretinal and epiretinal approaches are the furthest advanced, with clinical trials already underway.
Two brothers, Drs. Alan Chow and Vincent Chow, one an electrical engineer and the other a paediatric ophthalmic surgeon, have pioneered the development of the sub-retinal chip. Their implant is just 2 mm across and contains 5,000 microphotodiodes, each with its own stimulating electrode. They are powered by incident light, requiring no added current.
Subretinal approach
These chips, produced by the Optobionics company (founded by the Chow brothers), are designed to directly stimulate the remaining overlying cells of the retina. The chip is surgically implanted in the subretinal space and it is designed to produce visual signals similar to those produced by the photoreceptor layer.
Optobionics currently leads the research pack, with 10 patients having received the implants as part of a clinical trial begun three years ago. Initial reports indicate that all patients have expressed moderate to substantial improvement of visual function. Some patients could read letters on an eye chart who could not do so before surgery. Others have indicated improved ability to function in new environments, to see facial features and to recognise people.
But the Chow brothers have come in for considerable criticism. Although their results are good, critics argue that the chip has nowhere near enough power to generate the effects recorded. Critics speculate that the improvements are caused by neurotrophic factors, released as a biological response to injury.
"In the long term the neurotrophic effect might be more important. One could speculate that if you just put in a buzzer in there, and give the retina and brain a little excitation, you could have the same effect - at least temporarily. The Chow Brothers are well aware of this neurotrophic effect, however, and they know that more work needs to be done, with rigorous, regular testing," said Gerald Chader MD, Chief Scientific Officer of the Foundation Fighting Blindness, a US non-profit organisation that supports vision research
The Chow brothers are not alone in the pursuit of a subretinal prosthesis. The German Southern Consortium, headquartered in the University of Tuebingen, led by Professor Eberhart Zrenner, has taken a similar approach, while the Japan Hybrid Retinal Implant team has added a neat twist. That chip relies on living neural cells to transmit the signals from the microphotodiodes.
Epiretinal electrodes
With epiretinal technology, on the other hand, the implant is placed atop the retina where an array of electrodes transmit light signals through the retina. The most advanced project is led by Mark Humayun MD, professor of ophthalmology at the University of Southern California:
"We have successfully completed enrolment and implantation of three patients in the trial. And we have found that the devices are indeed electrically conducting, and can be used by the patients to detect light or even to distinguish between objects such as a cup or plate in forced choice tests conducted with one patient so far," Dr Humayun said.
The epiretinal device is a 4 x 5 mm sliver of silicone and platinum with 16 electrodes in a
4 x 4 array. It receives signals from a video camera, held on a pair of glasses. The epiretinal programme has received the imprimatur of the US Department of Energy (DOE), which gave Dr Humayun's group the largest grant it has awarded for implant research.
"The group we started with was Mark Humayun's group because even before we got into it they put a device into one patient. There's been a validation of the epiretinal device. This is not written in stone, but we decided that the epiretinal device looked the most promising. The 16-electrode device showed results in the first patient that were even better than anticipated. We decided that we would jump in on that and focus on a 1000-electrode device," said Michael Viola MD, director of the medical science division of the US DOE.
With an array of 32 x 32 (1024 electrodes) patients could expect vision of 20/26, and even at 500+ electrodes (say a 25 x 25 array, or 625 electrodes), a patient could read large print, one of the internal goals set for the DOE's retinal project.
"The two key research questions that remain are what happens at the electrode/tissue interface and how the brain perceives/adapts to the signals that are presented to it (the psycho-physics of phosphene vision). Humayun's group is the most advanced in terms of being able to provide insights into the latter question," said Professor Nigel Lovell of the Graduate School of Biomedical Engineering, University of New South Wales and one of the leaders of the Australian Bionic Eye project.
Other research teams working on epiretinal implants include a team at Harvard University, headed by Professor John Rizzo, and the German Northern Consortium, headquartered at the University of Bonn, led by Professor Rolf Eckmiller.
Getting on patients' nerves
The optic nerve implant takes an entirely different approach. In 1998, a spiral cuff nerve electrode was implanted intracranially and wrapped around the optic nerve of a RP volunteer. The cuff stimulates the nerve fibres via four electrodes, though the team now hopes to double that number. Very low current gives rise to the perception of phosphenes. While there was no way of knowing in advance how stimulation would affect the optical nerve fibres, trial-and-error taught the team how to induce limited pattern recognition in the patient.
"If you're going to do artificial vision, you don't have to use the eyeball, do you? There is a problem, however. As someone said, cynically, to underline just how difficult a task it could be: 'The fibres are not colour coded.' You don't know what fibres will provoke what images," notes microelectronics engineer John Alderman of Ireland's National Microelectronics Research Centre.
Other problems dog the research, too, including large, external, processing equipment and a transcutaneous power and data RF link, as with the epiretinal device. But a new project by the team hopes to improve the system. The team plans to implant three more patients by 2005 and expects large-scale human trials to follow.
Yet another approach, the cortical implant, plugs directly into the visual cortex bypassing the eye, the optic nerve and most of the rest of the brain to stimulate the nervous system where the majority of visual processing takes place.
"Depending on your point of view, the cortical implant is either the Lowest Common Denominator or the Highest Common Multiple," commented Dr Chader.
"Cortical approaches offer the possibility of treatment for a wider variety of vision problems but are considerably more difficult due to the mapping of stimulation sites on the brain," adds Professor Lovell.
Several teams are following the cortical implant route, including the US National Institutes of Health, and the University of Utah headed by Professor Richard Normann.
The most successful, and most controversial of the cortical research projects, is the Dobelle Group, based in Lisbon, Portugal, and headed by William Dobelle. Dr Dobelle's research offers the first commercial artificial vision system. One of his patients reportedly even drove a car. But his secrecy and avoidance of standard US clinical trials have prompted many to dismiss his work.
"Dobelle's group has virtually done an animal experiment on a human. I was fortunate enough to attend a lecture by Dr. Dobelle and its very interesting technology and hopefully it might work. One thing that's very encouraging is that it looks like the safety issues might have been solved by Dr. Dobelle, because one chip has been in since the 1970s. That's a very long time," notes Dr. Chader.
Only time will tell which of the approaches now being studied will produce functional artificial vision. Each of the four approaches has advantages and disadvantages, and each faces its own technical obstacles. Huge technical challenges remain in the areas of materials sciences, microelectronics and computational modelling. There are also unresolved issues with electrochemistry and neurochemistry. Finally, there is the issue of the implantation surgery itself, which has already shown to present a steep learning curve.
Nonetheless there is reason enough for optimism, since research has already progressed further and faster than many would have thought possible. The next few years should see even faster progress, fuelled by the combined forces of scientific inquiry and commercial competition.

Gerald Chader MD,
The Foundation Fighting Blindness
ALaban-Baker@blindness.org

Mark Humayun MD
Keck School of Medicine
humayun@hsc.usc.edu

John Alderman
Ireland's National Microelectronics Research Centre,
+353-21.490.4396
john.alderman@nmrc.ie, alderman@nmrc.ie

Nigel Lovell PhD
Graduate School of Biomedical Engineering,
University of New South Wales
N.Lovell@unsw.edu.au

Michael Viola MD
US Department of Energy
michael.viola@oer.doe.gov

Top