Seeing the Future

The renewed interest and competition in the international effort to construct a retinal prosthesis [see article, p1] is a positive development for both the neurotechnology industry and the visually impaired community. In our view, the presence of multiple parallel efforts to pursue a worthwhile engineering goal is almost always worthwhile if it leads to a commonality in purpose and adequate funding and public support.

In many ways, the market for retinal implants that emerges down the road may resemble the current competitive landscape in the cochlear implant market, where Australia-based Cochlear Ltd., Al Mann spinoff Advanced Bionics, and European vendor MedEl each offer viable products. We suspect that the s-curves of increases in product capabilities and number of individuals implanted over time will resemble the maturation of the cochlear implant business, where devices developed from single-channel systems to 22-channel systems relatively quickly after years of fundamental research in multiple laboratory settings. We also expect that bilateral retinal implants and implants for blind children will enhance the market potential, as they did for cochlear implants.

But the development of the retinal implant market will likely differ from the cochlear market in several respects, not least of which is the sheer complexity of processing and volume of information needed for a functional visual prosthesis. Here, a 22-channel device will not suffice; many experts believe that an electrode array of 1000 or more elements will be required for a useful device. But even with fewer electrodes than that, the retinal implant will benefit from factors that do not correspond perfectly to the cochlear implant. For instance, manufacturers of passive imaging devices like scanners, monitors, and digital printers have learned that they can greatly enhance perceived image quality using techniques such as dithering, anti-aliasing, and pixel depth so that a resolution of 300 dots per inch, for example, has the effective resolution of 2400 dpi or more. Unlike a scanner or printer, a retinal implant provides a steady stream of pixels over time so that a user can assimilate this data to build a more robust mental picture. As an example, Second Sight president Bob Greenberg relates the story of one of his early users who was able to refine her basketball skill by scanning the scene to identify the backboard.

And this points up the greatest promise awaiting users of retinal implants: the cortical processing and built-in optimization that comes along for free with the device. So long as we provide some minimal visual input, there’s no limit to what a retinal implant user’s brain can build with this information. So long as we provide some minimal incentive and support, there’s no limit to what the global neural engineering community can build to help people with visual impairments.

James Cavuoto
Editor and Publisher