Abstract
Introduction: While fMRI in normals can identify regions activated and thus potentially “involved” in a particular process, patient studies may be required to establish causal relationships between brain areas and performance. I used these two complementary methods to study biological motion perception.
Methods: In the fMRI study, 12 healthy participants viewed point-light biological motion or scrambled biological motion. In the neuropsychological study, a large group of brain-injured patients (47 left and 12 right hemisphere; lesion site initially nonselective) and 19 age-matched controls discriminated biological motion from scrambled biological motion in the presence of a variable number of noise dots.
Results: In the fMRI study, comparing biological motion to scrambled motion revealed bilateral activation in lateral temporal and in inferior frontal/premotor cortex. In the neuropsychological study, voxel-wise lesion analysis revealed that damage to very similar temporal and frontal regions was associated with deficits in biological motion processing. Therefore, the neural activity in normal adults agreed with the lesion data from the group of patients. Indeed, ROIs derived from the group lesion data via formal cluster analysis and thresholding showed significant neural activity for biological motion (but not for scrambled motion) in the normal controls.
Overall, compared with controls, both left and right hemisphere lesioned patients had significant deficits in biological motion perception. Patients' impairments did not correlate with lesion size or behavioural scores on other tasks (except face recognition).
Thus, specific regions in superior temporal and inferior frontal/premotor cortex appear to be both involved in and necessary for intact biological motion perception. The involvement of premotor cortex in biological motion perception may reflect the contribution of the mirror neuron system in the perception of these stimuli.
I would like to thank M.I. Sereno, N. Dronkers, E. Bates, D. Hagler, S. Wilson, F. Dick, J. Driver and G. Rees for help and comments, the UCSD Center for Functional MRI and VA Northern California Health Care System for providing access to imaging and patients. Support contributed by the National Science Foundation (BCS 0224321 to Sereno), the National Institutes of Health (RO1 DC00216 to Bates/Dronkers), and the European Commission (Marie Curie FP6 Mobility grant to Saygin).