Datensatz

Zusammenfassung

Using fMRI in anesthetized monkeys and a variety of computer-generated 3-D objects defined by shading, random dots, texture elements, or silhouettes and presented either statically or dynamically (rotating), we have previously identified 3-D shape-specific areas in occipital (areas VP and V3) and temporal (areas MT and FST; mid-to-anterior STS; and the AMTS) cortices (Sereno et al., Soc. Neurosci. Abstr., 26, 498.11, 2000). The present study investigates representation of 3-D shape from motion parallax using dynamic random dots in brain regions beyond the occipital and temporal lobes. Control stimuli consist of constructed objects with scrambled motion gradients. Such stimuli contain the same local motion information as the original objects, but the disruption of the cue gradient across the image diminishes an impression of depth. Spatially resolved BOLD contrast-based functional images of monkey visual cortex were obtained using a high-field (4.7 T) scanner and multi-shot, multi-slice, gradient-recalled, echo-planar imaging (EPI) sequences (voxel size, 1 x1 x 2 mm). Results showed significant activation in previously identified shape-specific regions of occipital and temporal lobes but also several areas in the intraparietal sulcus and two frontal lobe regions (the FEF and ventrolateral prefrontal cortex). This distributed network of areas cuts across both ventral and dorsal processing streams, reflecting multiple uses for 3-D shape representation in perception, recognition, and action.