Greater fMRI activation in the “fusiform face area” for the processing of expression than the processing of identity: Implications for face-recognition models

Abstract

According to several neuroanatomical face-recognition models that have recently been put forward, the processing of facial identity is mediated by a core system that includes the fusiform-face area (FFA), whereas the processing of facial expression is mediated by an extended system in dissociable brain regions, such as the superior temporal sulcus (STS) and the amygdala. We designed an fMRI study to test whether or not these two systems are indeed independent. To this end, we used a modified selective-attention paradigm that has been rigorously used behaviorally to study the relationship between the processing of identity and expression. Participants were asked to attend either to the identity or to the expression of the same set of faces. If the processing of identity is neuroanatomically dissociable from that of expression, then one might expect the “identity selective” system to show higher activation when processing identity as opposed to expression, and the “expression selective” system to show the opposite pattern of activation. Contrary to this prediction, we found that both systems were modulated in the same manner by the type of processing required. Both systems, not only the STS and the amygdala but also the FFA, showed higher activation when processing expression than when processing identity. The only regions that showed higher activation for identity judgments were located in early visual areas that are not specific to faces, presumably due to top-down allocation of attention to basic visual features. Overall, these findings suggest an interactive and hierarchical network for the processing of facial expression and identity, in which information about expression is computed from the unique structure of individual faces.

Footnotes

 Supported by a grant to MAG from the Canadian Institutes of Health Research