The First Foundations of Artificial Eyes
I see that Popular Mechanics is running an article on the presently crude and early steps towards artificial vision. In comparison to work on artificial hearts and kidneys, development of artificial eyes lags far behind - it's a challenging problem and the eye is arguably a more complex system than anything else outside the brain. The present mainstream approach involves building a grid of electrodes in place of the retinal cells lost to forms of degenerative blindness; images captured by a worn camera are analyzed and the electrodes stimulated appropriately.
In the morning, a surgeon at NewYork-Presbyterian Hospital will make an incision in Barbara's left eye and lift the saran-wrap-like membrane that covers it, called the conjunctiva. He’ll then suture a small electronics package, about the size of a watch battery, to the outside wall of the eye and secure it with a piece of silicone rubber that wraps around the eye’s equator. Next, he'll thread a thin cable through an incision in the wall; the cable connects the electronics to an array of 60 electrodes. After removing the vitreous humor that fills the inside of the eye - a material that’s essentially Jell-O, minus the sugar and food coloring - the surgeon refills the eye with fluid so that he can manipulate the array onto the retina, tacking it in place with what is perhaps the world’s tiniest pushpin. The whole procedure will take 4 to 5 hours....
The Argus II implant that Barbara will be receiving is the second generation of the device; the first had only 16 electrodes. Information gleaned from this clinical trial will be used to improve the 60-electrode version, which will be commercialized, first in Europe, as early as December. But even as the trial continues, a much larger effort, involving six national labs, four universities and a commercial partner, Second Sight Medical Products, is developing technologies that will enable third- and fourth-generation models using as many as 1024 electrodes - which could provide enough detail to read 24-point font and recognize faces.
Progress in this model is at present a matter of making implantation safer and more reliable, greatly increasing the density of electrodes, and improving the ability to translate a camera's view into a helpful picture - a combination of medicine, electrical engineering, and computer vision research. The end result of this form of technology will never produce anything more than a detailed, glowing sketch of dots and lines for the patient: it is not true vision as experienced by those of us fortune enough to retain our sight.
Nonetheless it works - already providing a great improvement for patients over being blind - and it will serve as a foundation for later forms of artificial sight technology. An established research and development community doesn't stand still after the first products are commercialized, but rather moves onward to new breakthroughs.