Abstract
Reduction of brain activation with repetition of faces, so-called repetition suppression, has been used to investigate face processing in the human visual cortex (Grill-Spector et al., 1999). In the present study, we investigated effects of emotion on repetition suppression of faces. Previous studies provided contradictory evidence on this issue, such that emotional faces produced less (Rotshtein et al., 2001) and more (Ishai et al., 2004) repetition suppression than neutral ones. We conducted an event-related functional magnetic resonance imaging (fMRI) experiment designed to resolve differences from the previous studies, and compared repetition suppression between emotional (angry and happy) and neutral faces in the fusiform gyrus and in the amygdala. When the same face was presented consecutively, we found that the peak amplitude of the fMRI signal in the middle and posterior parts of the fusiform gyrus was smaller for the second presentation than for the first, and this repetition-suppression effect was the largest for neutral faces and the smallest for happy ones. Results also suggested repetition suppression in the amygdala for neutral faces, which was positively correlated with repetition suppression in the fusiform gyrus. It thus appears that emotional faces ensured more sustained activation than neutral ones both in the fusiform gyrus and in the amygdala, providing support for the notion that emotional information undergoes extensive processing through interaction between the two regions.