Retinal implant offers spatial vision to blind patients

Key Points

Sylmar, CA-In a series of experiments, investigators have demonstrated that blind subjects who receive a retinal prosthesis implant (Argus 16, Second Sight Medical Products) have predictable spatial vision and that 16 electrodes provide better spatial resolution than four.

The first experiment showed that spots of light generated by the individual electrodes are well localized in the visual field, the second showed that visual acuity is determined by the electrode spacing, and the third showed that performance is poor when pixels are scrambled, said Matthew J. McMahon, PhD, senior principal scientist at Second Sight in Sylmar, CA.

Researchers developed a long-term, 16-electrode retinal prosthesis that produces perceptions of light, which lead to image perception and provide rudimentary sight to patients blinded from outer retinal degenerations such as retinitis pigmentosa and macular degeneration. Dr. McMahon and colleagues are conducting experiments with this prosthesis as part of a long-term collaboration with Doheny Eye Institute at the University of Southern California, where an ongoing investigational device exemption trial of the 16-electrode retinal implant is being conducted. In January, Second Sight received FDA approval to conduct a clinical study of a second generation of its electronic retinal implant (Argus II), which has 60 independently controllable electrodes.

"The point of this project is to map out the precision with which those phosphenes generated by individual electrodes are located in the world and also to determine if subjects can use this array of phosphenes to perform spatial tasks," he said.

The prosthesis consists of a tiny camera and transmitter mounted in eyeglasses, an implanted receiver, and an electrode-studded array secured to the retina with a microtack. A microprocessor and battery pack worn on the belt receives the video image from the camera and powers the entire device. The transmitter sends power and data wirelessly from the processing unit to the implanted receiver. The implanted receiver then sends electrical stimulation pulses to the individual electrodes in the array; each electrode points to a specific location in the visual field.

Experimentation

In the first experiment, subjects were asked to place a pointer in the perceived location of the phosphenes; a tracking system was used to record their spatial position. The second experiment consisted of a four-alternative, forced-choice orientation discrimination task that measured visual performance for square-wave gratings and examined the relationship between the number of active electrodes and visual performance.

The system can be operated in two modes, said Dr. McMahon, depending on the experiment being performed. For phosphene-mapping experiments, the individual electrodes are controlled by a personal computer.

"The personal computer allows us to access the electrodes in the array individually and send stimulation profiles to them directly," Dr. McMahon said. "In the second series of experiments, the visual acuity experiments, the system is used in its normal mode of operation. In the normal mode, the subject wears a pair of glasses that have a video camera embedded in the bridge. That video image is sent to the processing unit, which takes the section of the image that corresponds with the field of view of the array and converts that image into a series of electrical pulses that go to the individual electrodes."

The spatial mapping experiment was conducted by first stimulating a reference electrode for 2 seconds, followed by stimulating a test electrode for 2 seconds. The subject was instructed to fixate straight ahead and place a reflective ball in the position of the two phosphenes. The position of the test electrode phosphene was plotted against that of the reference electrode for each experiment. A stereo camera system recorded the position of the ball in 3-D space.

The investigators also conducted a control experiment in sighted observers by presenting visual targets that mimicked the spatial and temporal characteristics of the stimulations given to the blind subjects. The subjects viewed the experiment through a video camera that was set to record mode or playback mode. In record mode, subjects saw their hand in front of the camera while positioning the ball, which provided visual feedback. Because blind subjects cannot see their hands while positioning the ball, however, the resulting data did not provide a good baseline for judging the blind subjects' performance. Therefore, a second phase of the experiment with sighted subjects was performed while the video camera played a recording of the stimulation sequence. Because the subjects could not see their hands through the viewfinder, the conditions of the experiment were similar to those of the blind subjects.