Abstract
It has been suggested the ventral face areas, including occipital face area (OFA) and fusiform face area (FFA), process invariant aspects of faces such as identity while the dorsal areas, including the face-selective areas in posterior superior temporal sulcus (pSTS-FA) and in inferior frontal gyrus (IFG-FA), represent changeable aspects such as viewpoints and expressions. To investigate this division of labor, we manipulated which face aspect participants attended to while they were scanned. In addition, by testing developmental prosopagnosia participants (DPs) with the same task, we examined whether deficits in attentional modulation contribute to their face impairments. We first used a dynamic localizer to identify category-selective areas in 16 participants with normal face processing and 12 DPs. Next they performed a one-back task in which they selectively attended to either face identity, expression, or viewpoint. In normal participants, accuracy for identity and expression was comparable while accuracy for viewpoint was significantly better. Areas in the left and right hemisphere showed similar responses. In the OFA, responses in the three conditions were comparable. In the FFA, responses to identity and expression were stronger than to viewpoint. In the pSTS-FA, expression elicited the strongest responses and face view elicited stronger responses than identity. In IFG-FA, response to expression was the strongest, followed by identity, and then face view. DPs and controls showed comparable accuracy for expression and viewpoint judgments, but DPs were significantly worse than controls for identity judgments. Neurally, DPs showed comparable task modulation as controls across all the areas. In summary, face-selective areas show different response patterns to attentional modulations. Some of these patterns are consistent with the ventral/dorsal and invariant/changeable framework mentioned above, but some are not. In addition, normal attentional modulation effects in DPs indicate their impaired face processing does not result from deficits with face attentional modulation.
Meeting abstract presented at VSS 2017