On Thursday, scientists at the University of Southern California (USC) announced their plans to test an improved retinal implant in blind patients. The new implant, which scientists hope will improve patients' vision even more, has four times the resolution of the previous version. 

"My expectation, without really knowing what is going to happen, is that this will be useful for people in allowing them to find a lit doorway or the edge of an object when going into a room," says James Weiland, a scientist at USC involved in the project.

People with retinal-degeneration diseases, such as retinitis pigmentosa and macular degeneration, lose their sight as the cells in the eye that normally sense light deteriorate. Retinal implants can take over for these lost cells, converting light into neural signals that are then interpreted by the brain. Simpler versions of these devices, developed by researchers at USC and other institutions, have already been tested in humans, giving patients rudimentary vision, such as the ability to detect light and to occasionally distinguish between simple objects. One patient, for example, wears the device to her grandson's soccer games and reports that she perceives the sensation of the players' movement as they run by, says Weiland.

The device, developed by Mark Humayun and colleagues at USC, consists of a tiny chip dotted with hair-thin electrodes. When implanted in the retina, the electrodes transmit electrical signals from the chip to neural cells in the eye, which then send the message to the brain. A wireless camera mounted on glasses and a video processing unit worn on the belt capture and process visual information from the wearer's surroundings and wirelessly transmit those signals to the chip.

The new version of the implant, which the researchers have been working on for the past eight years, has nearly quadrupled the number of electrodes--from 16 to 60--and is about half the size of the previous model. The researchers recently received permission from the Food and Drug Administration to start human tests, which they plan to begin in the next few months.

Once the device is implanted, researchers will need to do extensive tests to figure out how to optimize it. "A camera gets at least tens of thousands of pixel information, and we need to transmit that to just 60 stimulating channels," says Weiland. "We have to figure out what is the most important information to keep."