Abstract
Congenital prosopagnosia (CP) is a specific deficit in identifying faces thought to be heritable and not the result of stroke or other brain injury. Functional brain imaging (fMRI) in CP has not reliably found differences between CPs and controls in ventral visual cortex, but rather reduced grey matter or connectivity between visual cortex and temporal/frontal lobe (Avidan & Behrmann, 2009). Here we reexamine the hypothesis that CP is associated with differences in face selective regions in visual cortex. We scanned 8 CPs who show poorer performance than controls on various tests of face recognition including recognition of famous faces, the Cambridge Face and Memory Test, but not in recognition of famous places, or recognition memory for scenes and objects. During fMRI, subjects fixated and performed a one-back task while viewing blocks of faces, places, and objects. Previously published data (Golarai et al., 2010) from 9 adults who participated in identical experiments served as control. In each subject, we drew anatomical ROIs in ventral temporal cortex (VTC), excluding early retinotopic areas and the anterior 1/3 of temporal cortex. In CP subjects, in the right VTC only, we found significantly fewer face-selective voxels (faces>objects) than in controls. There were no differences between groups in the spatial extent of object- or place-selective activations. To examine differences in signal change, we extracted signals from independent data. Face selective voxels in the right VTC of CPs showed a selective decrease in the response amplitude to faces relative to controls. However, groups did not differ in anatomical ROI size or the average model fit to the data in the ROI. These data suggest that congenital prosopagnosia may be associated with changes in both the extent and pattern of face selectivity in right ventral visual cortex.
Supported by NRSA 1F32EY018533-01 and NSF BCS 0920865.