Microstimulation in LGN Produces Focal Visual Percepts J. S. Pezaris and R. C. Reid Department of Neurobiology Harvard Medical School Boston, MA 02115, U.S.A. Abstract Existing efforts to develop a visual prosthesis have concentrated on stimulation of the retina or the primary visual cortex. We are pursuing a third approach which targets the dorsal lateral geniculate nucleus of the thalamus (LGN). We hypothesized that highly localized electrical stimulation of the LGN would generate focal percepts corresponding to the receptive fields of cells at the electrode tip. To test this we compared visually-guided saccades made to targets presented through the optic apparatus of the eye against those made to targets presented through electrical stimulation in the LGN. Daily single tetrode penetrations were made in the LGN of one alert behaving monkey. Receptive field locations were mapped for single units and the local field potential for each electrode position using 2D white noise stimuli. Eye positions were then recorded while a center-out saccade task was performed by the animal where, in interleaved fashion, saccade targets were presented optically as points flashed on a computer monitor or electrically through brief LGN stimulation. Our primary observation is that the animal could readily saccade to electrical targets consistent with electrical stimulation generating small, focal percepts. Correspondence between electrical saccade endpoints and receptive-field centers was 1.6±0.7° (n=5, with eccentricities of 6-13°), versus optical saccade accuracy of 1.6±0.5°. Electrical and optical saccades were comparable: endpoint scatter was 0.5 vs 0.6°, latency was 150 vs 172 ms. At times, however, the animal's reaction in electrical trials was to hold it's gaze steady, as if continuing to foveate the extinguished fixation point. Our observations demonstrate the possibility of creating artificial visual percepts through electrical stimulation in the LGN. Such percepts are visual in nature insofar as an animal immediately generalized to them in a visual task. Further, these percepts are focal as saccades to them are tightly clustered. Supported by: Kirsch Foundation, Dana/Mahoney Foundation, NIH R01 EY12815 and P30 EY12196. Fetch PDF of abstract (0.02 MB) Fetch PDF of slides (0.7 MB) John Pezaris, Harvard Medical School, 220 Longwood Ave, GB-203, Boston, MA 02115 pz [at] hms [dot] harvard [dot] edu, 10 November 2004.