Vision for the blind

The two scientists at Weill Cornell Medical College in New York, who published their findings in the journal Proceedings of the National Academy of Sciences, have also deciphered the retinal neural code of monkeys - which is essentially identical to that of humans.
"It's an exciting time," said computational neuroscientist Sheila Nirenberg, in a statement from the college. "We can make blind mouse retinas see, and we're moving as fast as we can to do the same in humans."
Novel prosthetic vision system
The novel prosthetic vision system consists of two parts: an 'encoder' and a 'transducer'. The encoder mimics the role of the retina by converting light patterns into streams of electrical pulses using the same 'code' by which a healthy retina communicates to the brain. The transducer, or 'stimulator' - a light-sensitive protein - then 'fires' the appropriate retinal neuron cells, called ganglion cells, as specified by the code.
"What these findings show is that the critical ingredients for building a highly-effective retinal prosthetic - the retina's code and a high-resolution stimulating method - are now, to a large extent, in place," said Nirenberg.
At present, most bionic vision systems use a camera that converts visual input into electrical signals communicated via electrodes implanted either on the retina or the brain. This usually results in simple light dots and outlines with very low resolution that assist the user with basic navigation.
Blind can discern facial features, movement and landscape
The neural code behind this latest system is so accurate, the researchers say, that it can allow facial features, movement and landscapes to be discerned.
"The reason [other bionic vision researchers] have not reached their goal is not just because of a resolution problem," the authors wrote in the paper, "but also because of a coding problem."
Wai Ho Li, a lead researcher at the Monash Vision Group in Melbourne, who is working towards a direct-to-brain bionic vision system, said this new research is "an important step towards better visual prostheses".
"The research demonstrates a new approach for improving the quality of bionic vision generated using retinal optogenetics," he said. "Instead of increasing the visual resolution of bionic vision, they argue that encoding information for the visual system so that it doesn't get 'lost in translation' is just as important."