Blind patient reads words stimulated directly onto the retina

Neuroprosthetic device uses implant to project visual braille
November 23, 2012

(Credit: Second Sight)

Researchers have projected braille patterns directly into a blind patient’s retina, allowing him to read four-letter words accurately and quickly with an ocular neuroprosthetic device.
The device, Second Sight‘s Argus II, has been implanted in over 50 patients, many of who can now see color, movement and objects.

It uses a small camera mounted on a pair of glasses, a portable processor to translate the signal from the camera into electrical stimulation, and a microchip with a 10 by 6 electrode array implanted directly on the retina.

“In this clinical test with a single blind patient, we bypassed the camera that is the usual input for the implant and directly stimulated the retina. Instead of feeling the braille on the tips of his fingers, the patient could see the patterns we projected and then read individual letters in less than a second with up to 89% accuracy,” explains researcher Thomas Lauritzen, lead author of the paper.

For this study, the researchers at Second Sight used a computer to stimulate six of the 60 points on the electode array to project the braille letters. A series of tests were conducted with single letters as well as words ranging in length from two letters up to four.

“There was no input except the electrode stimulation and the patient recognized the braille letters easily. This proves that the patient has good spatial resolution because he could easily distinguish between signals on different, individual electrodes,” says Lauritzen.

Primarily for sufferers of the genetic disease Retinitis Pigmentosa (RP), the implant Argus II has been shown to restore limited reading capability of large conventional letters and short words when used with the camera, Second Sight says. But the new study shows how the Argus II could be adapted to provide an alternative and potentially faster method of text reading with the addition of letter recognition software.