Retinal degenerative diseases lead to blindness due to the loss of photoreceptors even though the inner retinal neurons remain largely intact. Visual percepts, also called “phosphenes,” can be produced by electrical activation of the inner retinal neurons. This alternate route to visual information has the potential for restoring sight to the blind. Current retinal prosthesis designs, with electrode arrays implanted in the retina facing either the ganglion cells or the inner nuclear layer, rely on serial telemetry to deliver stimulation signals to the electrodes, requiring bulky receiving and processing electronics and a trans-scleral cable. Surgery is complex and the design is difficult to scale up to attain higher visual acuity. In addition, patients cannot use natural eye movements to scan the visual scene because retinal stimulation patterns are transmitted from an external camera to the retinal implant, independent of eye orientation. These limitations can be overcome by devices that use photosensitive pixels but they depend on an external power source. Recently, however, researchers from the Palanker group at the Hansen Experimental Physics Laboratory and the Department of Ophthalmology at Stanford University designed a photovoltaic retinal prosthesis where video goggles were used to deliver both power and visual information through pulsed NIR illumination, preserving the natural link between image perception and eye movement without complex electronics and wiring.

In an article published in the June issue of Nature Photonics (DOI: 10.1038/nphoton.2012.104; p. 391), Keith Mathieson, James Loudin, and co-researchers from Stanford University and the University of California–Santa Cruz, describe their prosthesis design in which video images captured by a head-mounted camera are processed by a portable computer. The video goggles use a liquid-crystal display (LCD) illuminated by pulsed near-infrared light (880–915 nm) to project the images onto a subretinal photodiode array (consisting of 70 µm pixels, each with ~20 µm stimulating electrodes), which converts the light to local currents that stimulate the nearby neurons in the inner nuclear layer of the retina.

The researchers fabricated silicon photodiode arrays consisting of pixels with single diodes as well as those consisting of pixels with three diodes connected in series. These triple-diode pixels can produce 1.5 V, which triples the charge injection on the sputtered iridium oxide film electrodes (from 0.5 mC cm‒2 for a single-diode pixel to 1.5 mC cm‒2). The triple-diode pixels require light intensities three times higher than single-diode pixels because the photosensitive pixel area is divided into three subunits. However, the researchers found that their single-and triple-diode devices had very similar thresholds for eliciting retinal responses.

The researchers tested their design concept by stimulating healthy and degenerate rat retinas in vitro with NIR light intensities at least two orders of magnitude below the ocular safety limit. They showed that the elicited retinal responses can be modulated by both light intensity and pulse width, although their current optical design allows only for intensity modulation within each video frame. However, if the retinal response is modulated by varying the pulse width, the researchers said that digital light processing technology can also be used, adding, “Such a device would allow both the duration and timing of exposure to be precisely controlled on the scale of individual pixels. In addition to higher throughput compared to an LCD, this high-speed control would allow the sequential activation of nearby pixels to further reduce pixel crosstalk—interference of currents from nearby pixels.”

figure 1

Subretinal photodiode array with triple-diode pixels arranged in a hexagonal pattern. Pixels of 70 μm and 140 μm in size were made. Left inset: Central electrodes are surrounded by three diodes connected in series, and by the common return electrode. Right inset: The subretinal implant.