Abstract
Motion is an important cue in elucidating 3D-structure of complex objects. Previous fMRI work comparing 3D rotating and 2D translating line patterns in passively viewing subjects indicated that an occipito-parietal network, including hMT/V5+, may be involved in extracting depth-from-motion (1). In the current fMRI-study, discrimination tasks were used to test to what degree these regions were involved in processing 3D shape rather than 3D motion contained in the displays. Stimuli consisted of rotating, randomly deformed, textured spheres and featural attention was directed to either shape-from-motion, to motion (judging rotation axis) or to the texture by means of a same-different task performed on each of these attributes. Two control conditions in which either central or peripheral luminance dimming (present in all conditions) had to be detected were also included to control for visuospatial attention effects. Random effects analysis on fMRI data from 12 human subjects revealed shape specific activation in a bilateral lateral occipital area, posterior and superior to hMT/V5+. A bilateral inferior satellite of hMT/V5 was also specifically activated in the shape task, while shape and motion tasks activated bilateral hMT/V5 proper equally. Shape also engaged right anterior fusiform cortex, which was also active in texture, but not motion. Activation in the dorsal parietal cortex common to motion and shape was observed in two sites: one posterior along the intraparietal sulcus was slightly more active in shape and one anterior and lateral from IPS was slightly more active in motion. These results show that attention to 3D shape extracted from motion activates a subset of the passive viewing network together with ventral regions.
1
OrbanGASunaertSToddJTHeckePVanMarchalG(1999) Human cortical regions involved in extracting depth from motion. Neuron, 24:929–940.
Supported by GOA 2000/11 to GOA;HP is supported by FWO Flanders.