Author:

Keywords:

FMRI, TEMPORAL LOBE, MACAQUE

Abstract:

Stimulus repetition decreases the response of neurons in the monkey inferior temporal (IT) cortex (Miller et al., 1991; Ringo, 1996) - an area important for processing objects, including faces. A number of functional magnetic resonance imaging (fMRI) studies in humans have shown a decrease of neural activity in many areas of the occipito-temporal cortex when identical faces are repeated as compared to different faces (Grill-Spector & Malach, 2001). This ‘repetition suppression’, ‘habituation’, or ‘adaptation’ effect (Grill-Spector et al., 2006) has been used to infer that these areas are sensitive to differences between individual faces. More recently, an organized network of face patches have been identified in monkey IT using fMRI (Tsao et al., 2006). The first purpose of this study was to test whether these areas also show a decrease of activation when an individual face is repeated. The second purpose was to determine the optimal stimulation frequency rate at which such an effect would be observed. First, a whole-brain face localizer was run on two male monkeys, that were conditioned to fixate for long periods of time, using the contrast agent enhanced fMRI technique. In agreement with previous studies, a comparison between faces and other objects yielded multiple patches of activation, mainly in the posterior, middle and anterior superior temporal sulcus (STS). In the subsequent fMR-adaptation experiment, there were 24 unique conditions (12 with an identical face; 12 with different faces). In each of these conditions, a human face appeared and disappeared through a uniform gray screen (sinusoidal contrast modulation, as in Rossion & Boremanse, 2011), at a variable rate of stimulation (1-12 Hz, equally spaced by 1Hz). In the identical face condition, the same face was presented repeatedly (although size of the face changed in every presentation). In the different faces condition, there were 10 individual faces that appeared in a random order (presentation size was varied throughout each different face condition). Each condition was presented for 10 seconds in a standard block design. All of the functionally defined face patches showed a larger response to blocks of different faces than when the same face was repeated. These effects were the largest when faces were presented at a rate of 5 Hz to 7 Hz. This observation suggests that the face processing network was most sensitive to differences between faces at these middle frequency rates (between 140-200 ms SOA between faces). Future electrophysiological experiments have been designed to clarify the response properties of the neurons in these fMRI defined patches, during this repeated face stimulation task.