Item

Abstract

The visual system is sensitive to multiple visual cues indicating the shape of objects; these cues are combined by the brain to result in unified shape perception. The goal of this study was to examine whether different visual areas are involved in the analysis of individual cues versus the representation of the perceived shape based on combined cues. We manipulated horizontal binocular disparity and geometric perspective cues to the shape of test stimuli (hinged planes). Disparity and perspective cues could indicate the same shape (consistent cues) or different shapes (inconsistent cues). Psychophysical judgments about the angle between the hinged planes of two sequentially-presented stimuli were collected concurrently with fMRI responses. A sequential-presentation fMRI adaptation paradigm was employed, in which stimulus changes result in increased fMRI responses (rebound effect) compared with repeated presentation of the same stimuli. We used conditions in which either individual cues changed, or both cues changed. We tested for fMRI responses in early retinotopic visual areas, higher object-related (Lateral Occipital Complex-LOC) and motion-related (hMT+/V5) areas. We observed significant fMRI rebound effects in early and higher visual areas. These rebound effects followed similar patterns to psychophysical discrminiability (d prime) of the perceived shape in areas LOC and hMT+/V5. However, rebound effects in areas V3 and V3a were consistent with changes in each of the individual cues rather than the perceived shape per se. These results suggest that early visual areas are involved in cue-based representations, while representations based on a combination of cues appear to be encoded in higher object-related (LOC) and motion-related (hMT+/V5) areas.