There has been a wide interest in neuroprosthetics that replace primary sensory organs. Perhaps the oldest and most successful example of a sensory neuroprosthetic is the cochlear implant. While not a perfect replacement, it has had a tremendous impact in restoring basic auditory function to patients. Research is also underway to develop a visual prosthetic device. However, limitations in current technology and our understanding of the visual pathway present roadblocks in the development of an adequate device.
Seminal work in developing an auditory prosthesis was conducted by French researchers in 1957, and involved stimulation of the auditory nerve with a single channel electrode. Even with this simple setup, the patient was able to distinguish monosyllabic words. More recent technologies informed by our understanding of the auditory system have dramatically improved this performance. For example, the portion of the auditory nerve that runs inside the organ of Corti is tonotopically organized; tones are spatially arranged by frequency. Cochlear implants take advantage of this by breaking incoming sound up into multiple frequency channels. Each channel is fed along its own electrode to the corresponding portion of the auditory nerve replicating the natural function of the cochlea. Just four frequency channels are sufficient to understand speech, which is remarkable considering that roughly 30,000 hair follicles normally stimulate the nerve.
Current implants have up to 22 channels, more than adequate for understanding speech, but still inadequate for auditory tasks such as music perception. A 2001 study in Nature demonstrated that improvements in music percep-tion will require improvements in sound encoding. Not only does the auditory system decompose sound by frequency, it also separates sound into two components: a slowly varying frequency envelope suitable for speech recognition, and a rapidly varying fine structure used in pitch perception and sound localization.
The development of a visual prosthetic has been much more difficult. Firstly, the amount of visual information entering the eyes is more complicated and dense compared to the auditory system. While four frequency channels were adequate for functional hearing, having only four discriminate regions for vision is inconceivable. Researchers have suggested that a 10x10 array of pixels can restore just enough vision for navigation and object avoid-ance, while as many as 1000 pixels are need to restore face recognition and reading. The retina is also an incredibly compact organ with a receptor cell density that cannot be easily replicated. One group demonstrated that a silicon microarray for nerve stimulation can be machined with electrodes that are 0.4mm apart, although this is still a far cry from the maximal cone density found in the fovea.
Despite the technological challenges of creating an artificial replacement retina, it still remains the primary site of visual prosthetic research for a number of reasons. Firstly, just as the cochlea is tonotopically organized, the retina has a natural spatial organization which simplifies prosthesis design. Secondly, a retinal approach takes advantage of natural visual processing that begins at that level, including colour perception and edge accentuation.
Address: 5636 Lemon Ave.
Dallas TX 75209
Phone: +1 214 5203694