Experimental Results of Intracortical Electrode Stimulation in Macaque V1

Abstract

Abstract: : Purpose: To investigate how a large array of V1 intracortical electrodes might be used to invoke artificial vision. Methods: 192, 35-micron diameter activated iridium oxide wire electrodes, of surface areas 200 and 500 sq. microns, were prepared for implantation in a Macaque V1. The electrodes were physically configured as sixteen 8-electrode arrays, and 64 single electrodes. Customized connector housing, fitted to the curvature of the animals skull contained 24 Omnetics nano-connectors. A standard eye-coil technique was used to measure eye position. In the 14 months, following surgery, the electrodes were physically and functionally mapped, and the receptive field location for each electrode was determined. The animal was trained to perform a memory saccade task in which a visual point-stimulus was presented while the animal maintained central fixation on a visual spot. Once the fixation point was extinguished, the animal performed a saccade to the memory-based location of the visual stimulus. After the visual saccade task was mastered, electrical stimulation of the electrodes replaced the visual stimuli. The expectation was that the animal would learn to use the artificially-induced percepts in a manner similar that of the visual-induced percepts. Repeated trials were stored on an automated data collection system and analyzed for statistical significance. Results: 114 of the implanted electrodes were electrically accessible. This number was expected based upon known defects in the hardware prior to, and during the, surgery. No electrodes were "lost" following surgery. Of these, we obtained spatial receptive-field maps for 60-70 electrodes using 1-degree visual stimuli that were measured using an averaging system over 15,000 trials. After appropriate training, the animal learned to used the electrically-induced percepts in a manner similar to that of the visual ones, with a high degree of statistical significance for the saccade end-point locations relative to the electrode receptive fields. Conclusions: It appears that an animal model for studying stimulation strategies for an intracortical visual prosthesis is feasible. Our ongoing studies are directed at understanding how an artificial interface to the cortex might be used to exploit the natural V1 tuning so that a cortical visual prosthesis could be implanted into a human.

Keywords: visual cortex • visual fields