Investigation of the role of context in the processing of low-contrast visual information using 7T (V1)

The brain uses contextual information from the visual environment to generate predictions about current and upcoming sensory information. Here, we are interested in how predictive mechanisms of perceptual inference map on to microcircuits in human visual cortex. Specifically, if contextual information amplifies low contrast feedforward input in layers of primary visual cortex. We used a partial visual occlusion paradigm in which we presented natural scenes with the lower right quadrant at low contrast (defined from an equivalent 3T study, using the group average for 75% accuracy in detecting the target image). The content of the lower right image either matched or did not match the contextual surrounding information present in the remaining image, manipulating the consistency (or predictability) of the contextual information. We used retinotopic mapping approaches to locate the cortical region responding to the lower right image quadrant only in V1. We are interested in how contextual information driven by the image surround, modulates BOLD activity responding to the low contrast portion of the scene in cortical layers.