Abstract
INTRODUCTION. Binocular disparity signals are small spatial differences between the left and right eyes and create the perception of three-dimensional depth. Electrophysiology and optical imaging in primates have associated disparity processing with the thick stripes of the secondary visual cortex (V2). However, imaging at this scale in the human depth system has not yet been demonstrated. Using very high resolution functional magnetic resonance imaging at 7 Tesla (0.75mm isotropic voxels), this study investigates the organisation for binocular disparity in the human visual cortical areas. METHODS. Cortical blood oxygen-level dependent (BOLD) responses were acquired from 4 subjects with good binocular vision whilst viewing binocular stimuli using anaglyph glasses. In the ‘depth’ condition, the stimulus was divided into a grid of 3 x 3 squares. The central square contained a fixation cross, while the disparity of the surrounding squares changed every second to a random disparity between ±0.3°. Subjects viewed the ‘depth’ condition for 16 seconds, alternating with 16 seconds of uncorrelated or random dots in depth. Each subject performed 8 repeats of the 3 minute scan. Data were acquired on 2 separate days for 2 subjects to test reproducibility. Standard retinotopic mapping data obtained at 3 Tesla were used to define the primary visual cortex (V1) and extrastriate areas (V2, V3 and V3a) for each subject. RESULTS. Discrete clusters of voxels responding to the ‘depth’ condition were predominantly found in extrastriate cortex, including small regions of V1. Reproducibility was evident both across runs within sessions, and across the repeated sessions. CONCLUSION. High-resolution fMRI suggests a functional organisation for disparity processing in the human early visual areas, indicated by a repeatable clustering of responses broadly consistent across individual subjects. Further investigation will examine whether there is fine grain organisation within these clusters of disparity sensitivity.
Meeting abstract presented at VSS 2012