Abstract
Different posterior brain regions are consistently activated when viewing body movements or static body images (pSTS, FBA, and EBA), yet their distinct functional roles including how they code information remains elusive. Our results from five patients with ventral visual lesions and control groups (including n>50 brain damaged patients without ventral cortex damage (Saygin 2007)) indicate that ventral visual cortex is not critical for the perception of and sensitivity to biological motion, as evident from the patients' effortless recognition of point light displays and their normal perceptual thresholds. Lesion delineation indicates that EBA or FBA damage does not impair biological motion perception. In contrast, these patients have form perception deficits and cannot recognize people from full-body static images. pSTS is rather spared in these patients. Following these and previous findings I propose a model that outlines the functional contributions of pSTS to biological motion recognition, and of EBA and FBA to human body recognition: While the integrity of pSTS is critical for biological motion recognition, the integrity of EBA and FBA is critical for recognizing human bodies. More generally, pSTS processes the motions/kinematics of self-moving objects, partially by relying on low-resolution static body-in-motion snapshots within it. The EBA engages in visual representation of biologically-moving objects, and perhaps of other self-moving objects. Fusiform regions engage in high-resolution visual representation of all object types, with enhanced representation of self-moving objects due to the varying appearances caused by their self-motion. The model posits that these representations are based on and modulated by experience/familiarity, explaining the sensitivities to biological motion [human body] in pSTS [EBA/FBA], the human biological motion inversion effect (absence of exposure to inverted stimuli leads to absence of representation), and findings in clinical populations. Furthermore, the model provides testable predictions for future research.
Meeting abstract presented at VSS 2014