Abstract
When one visual object moves behind another, it provides a compelling sense of which object is closer to the viewer and which object is farther in depth. This percept is referred to as depth-order from motion (DFM). The neural mechanisms of DFM are largely unclear, including the relative roles of the first-order (i.e., luminance-based) vs. the second-order (i.e., non-luminance-based) motion processing mechanisms, and the relative contributions of the two known types of DFM cue, the accretion-deletion (AD) cue and the common motion (CM) cue. To help elucidate the neural mechanisms of DFM perception, we performed a human whole-brain fMRI scan using a mixed (i.e., events-within-blocks) design, which allowed us to compare the responses across blocks as well as across individual trials. Depending on the stimulus block, subjects viewed random dot stimuli that either elicited strong depth-order percepts (DFM stimuli) or did not (null stimuli). Subjects reported the perceived depth-order using a button press. The DFM stimuli contained both types of DFM cue, whereas the null stimuli contained neither cue. The two sets of stimuli were otherwise identical, including in terms of average luminance, contrast, and motion energy. The only brain region that showed a significant activation in all subjects (N = 21; p <0.05, corrected for multiple comparisons) was a bilateral occipital region that partially overlapped with the kinetic occipital region (KO), previously known to be involved in the processing of motion borders. Similar results were obtained when real-world 2-D photographs were used as DFM stimuli, indicating that the responses were not stimulus-dependent. Importantly, the responses of this region varied systematically with the strength of DFM information in the stimuli, and the DFM percept they elicited, on a trial-to-trial basis. Together, these results indicate that this occipital region plays a key role in DFM cue processing and DFM perception.
Meeting abstract presented at VSS 2012