Abstract
Anecdotal evidence from electrical stimulation of human fusiform cortex suggests that the visual perceptual event induced by stimulation of a given cortical position varies in amplitude depending on the visual input to the eyes. This is not compatible with the intuitive view, associating a singular phosphene-like event with stimulation of a given position on the visual cortex. Here we systematically test this question. In two macaque monkeys, we chronically implanted LED arrays over the central inferior temporal (IT) cortex where we had previously transduced excitatory opsin C1V1. We also implanted an array on the opposite IT cortex where no opsin was transduced. The animals were then trained to report the presence or absence of optogenetic stimulation while simultaneously presented with an image randomly drawn from a set of 40 object images. Each behavioral trial was composed of a 1s image presentation, and a 200ms illumination impulse delivered to the IT cortex in the middle of image presentation in half of the trails, randomly selected. Both animals learned this task significantly above chance level. The analysis of results revealed a strong effect of image on the animals’ ability to detect cortical stimulation (p<0.001) with their performance ranging from 88% to 52% for various objects. The animals were not able to detect cortical illumination over control areas with no virus transduction. In a separate experiment, we showed that detectability of optogenetic stimulation increases as the intensity of LED activation increases, and that the slope of this increase is different for various objects (p< 0.02). Our results show that detectability of optogenetic stimulation of IT cortex varies across different visual inputs. This is consistent with human observations and suggests that the perceptual events induced by artificial activation of the IT cortex highly depend on the content of vision.