Increasing implant resolution by a factor of four is significant, says E. J. Chichilnisky, a neuroscientist at the Salk Institute for Biological Studies, in La Jolla, CA. But compared with the human eye, the resolution is still very limited. "Imagine a camera with 60 pixels," Chichilnisky says. "You can't really see a face in an eight-by-eight image, or even a word. In the long run, we'll need hundreds or thousands of electrodes to get something interesting. So there is a lot more to be done." Both Chichilnisky and the USC researchers are working with Second Sight Medical Products, the company based in Sylmar, CA, that is manufacturing the devices, on the next version of the implant. The third-generation device will have 500 electrodes, boosting resolution by a factor of almost 10.

But increasing the number of electrodes won't be the only hurdle in developing implants that can give blind people truly useful vision. Scientists also need to figure out how to electrically stimulate the retina in a way that the brain can interpret with high spatial resolution, says Joseph Rizzo, an ophthalmologist at the Massachusetts Eye and Ear Infirmary and codirector of the Boston Retinal Implant Project. A ray of light, for example, stimulates retinal cells in a more precise and refined way than does the electric current coming from an electrode. "It doesn't matter if you have 10 or 1,000 electrodes," he says. "If you don't know how to use them, it doesn't matter."