Responses to interocular disparity correlation in the human cerebral cortex

Purpose: Perceiving binocular depth relies on the ability of our visual system to precisely match corresponding features in the left and right eyes. Yet how the human brain extracts interocular disparity correlation is poorly understood. Methods: We used functional magnetic resonance imaging (fMRI) to characterize brain regions involved in processing interocular disparity correlation. By varying the amount of interocular correlation of a disparity-defined random-dot-stereogram, we concomitantly controlled the perception of binocular depth and measured the percent Blood-Oxygenation-Level-Dependent (%BOLD)-signal in multiple regions-of-interest in the human occipital cortex and along the intraparietal sulcus. Results: A linear support vector machine classification analysis applied to cortical responses showed patterns of activation that represented different disparity correlation levels within regions-of-interest in the visual cortex. These also revealed a positive trend between the difference in disparity correlation and classification accuracy in V1, V3 and lateral occipital cortex. Classifier performance was significantly related to behavioural performance in dorsal visual area V3. Cortical responses to random-dot-stereogram stimuli were greater in the right compared to the left hemisphere. Conclusions: Our results show that multiple regions in the cerebral cortex are sensitive to changes in interocular disparity correlation, and that dorsal area V3 may play an important role in the early transformation of binocular disparity to depth perception.