Abstract
Dynamic faces convey a wealth of social information. Both macaques and humans depend on facial motion to facilitate the recognition of facial expression in their social interactions. However, the extent to which the specialization for facial motion is represented in the visual systems across the two species remains unclear. Here, we used fMRI to investigate this issue. Four male macaque monkeys were scanned at 4.7T. During scanning, they viewed blocks of dynamic neutral monkey faces, dynamic common objects, static neutral monkey faces, and static common objects. Sixteen human subjects participated in a similar experiment at 7T. Participants viewed blocks of identical stimulus categories, except they viewed human faces. For both monkeys and humans, the static face/object stimuli were extracted from the corresponding dynamic video stimuli. Motion energy in the dynamic face blocks was equated to that in the dynamic object blocks. General linear model analyses were performed on whole brain fMRI data to evaluate the fMRI responses evoked by the four stimulus categories. The contrast of the fMRI response to motion caused by faces (dynamic faces versus static faces) relative to the responses to motion caused by objects (dynamic objects versus static objects) defined the brain areas selective for facial motion. Our results showed that, in all monkeys, significant activations evoked by facial motion were found in the anterior superior temporal sulcus (STS), within the anterior fundus face patch bilaterally (p< 0.001). In humans, facial motion activated three separate foci in the right STS (p< 0.001): anterior STS, middle STS and posterior STS, with the anterior STS focus showing the most significant selectivity for facial motion. Our results suggest that monkeys and humans share similar neural substrates within the anterior STS for the processing of facial motion.
Meeting abstract presented at VSS 2016