Abstract
It is reasonable to assume that change of activity of a neuron in inferior temporal (IT) cortex is typically interpreted by the rest of the brain as a visual event in the external world, which is why artificial stimulation of IT cortex alters visual perception. Accordingly, we have demonstrated that monkeys’ ability to detect optogenetic cortical stimulation interacts with the concurrent visual input. Here we asked whether behavioral detectability of a cortical stimulation event varies with the plausibility of a corresponding visual event in the external world. To test this we carried out optogenetic stimulation in monkeys performing a stimulation-detection task while viewing images whose structure and visibility were systematically varied using two methods: grinding and blurring. Grinding was achieved by texturising natural images at different spatial scales using a texture synthesis model (Deep Texture Network, Bethge Lab). Blurring was done by low-pass filtering. Two monkeys were implanted with LED-arrays over a region of their IT cortex transduced with the depolarizing opsin C1V1. In each trial, following fixation an image was displayed for 1s. In half of the trials, randomly selected, a 200ms illumination impulse was delivered halfway through image presentation, and the animal was rewarded for correctly identifying whether the trial did or did not contain cortical stimulation. We found that blurring, which diminishes image structure and visibility, systematically reduced stimulation-detectability (ANOVA, p < 0.001 for both animals), whereas grinding, which reduces image structure while preserving image visibility, did not (ANOVA, p > 0.1 for both animals). The results suggest that stimulation of IT cortex is easily perceived when the retinal input suggests the existence of an external visual event, regardless of the degree of structure it contains. The same stimulation impulse however, is harder to perceive when the retinal input suggests the absence of an external visual event.