The processor would send those impulses to the implanted electrodes. And the activity on the electrodes would enable patients to see images. It would not result in a full range of sight -- primarily because it is impossible to replicate the function of the retina with a digital camera -- but it might enable patients to identify objects and recognize faces. But the ability to deliver high-resolution images to the brain should improve over time. The inventor, John Pezaris of Massachusetts General Hospital in Boston, says the best candidates for the device are those who are totally blind, whether from birth or from an illness or traumatic head injury. For those who are partially blind, the risks of the surgery may not outweigh the benefits they could reap.