A self-powering retinal implant that can turn light into sight has been successfully tested in rats.
The device, the brain child of scientists at Stanford University led by Yossi Mandel, is intended to treat disorders caused by the loss from the eye of the rods and cones that convert light waves into electrical signals that the brain can understand.
Critically, in these conditions, the underlying retinal tissue, including the cells that give rise to the optic nerves that convey the sight-information to the brain, remain intact. So artificial implants can feed electrical signals into these cells, which respond just as if the rods and cones were still working normally.
But existing retinal implant prototypes have a number of problems, not least that they needed an external power supply, and running wires across the retina can lead to abberant signals being fed into the wrong cells, reducing the quality of the resulting image.
The solution offered by the Stanford team, published in Nature Communications, is to use a self-powering device that obtains the energy it needs from the light that lands on it.
It comprises a series of miniature photovoltaic cells that react to incoming light waves by producing small electrical currents that are then fed directly into the underlying retina.
Using rats with degenerative retinal disease as well as normal healthy animals, the researchers successfully tested versions of their implants containing arrays of up to 186 photovoltaic cells - or pixels - in the animals.
The devices were placed inside the eye directly onto the retina where the rods and cones would normally reside. Then, spots of near-infrared (NIR) light, to which the implants are designed to respond, were beamed into the eye.
When the implants were illuminated, the electrical activity of the visual areas of the animals' brains correspondingly altered, indicating that the implant was registering the near-infrared light and correctly stimulating the retinal circuitry.
The implants also faithfully reproduced flashes of light, which were presented at rates varying from twice to twenty times per second, producing similar brain responses to a healthy retina, and the system worked for at least 6 months following implantation.
It's early days yet, but the thin dimension and relative simplicity of the Stanford groups prosthesis, and the fact that it can be scaled up to make larger arrays of light-detecting pixels, means that it could be used to produce a viable device for human trials relatively soon.
The disadvantage is that the present design relies upon near-infrared light, which would need to be beamed into the eye using a miniature projector coupled to a camera and mounted, most probably, on a pair of glasses or a visor.
This device was reported as only working in Infra-Red light, which is the bandgap of Silicon. This would be quite useful to a previously blind person who wanted to find their way in daylight (which has lots of IR).