Are the patches important? The effect of inversion on the responses of face-selective cells found throughout the monkey superior temporal sulcus.
35.22, Sunday, 18-May, 5:15 pm - 7:15 pm, Talk Room 2
Jessica Taubert1,2, Goedele Van Belle1, Wim Vanduffel2,3,4, Rufin Vogels2, Bruno Rossion1; 1Institute of Research in Psychology and Institute of Neuroscience, University of Louvain, 2Laboratorium voor Neuro- en Psychofysiologie, KU Leuven, Leuven, Belgium, 3MGH Martinos Ctr., Charlestown, MA, USA, 4Harvard Med. Sch., Boston, MA, USA
It is assumed that upright faces are processed by a set of functionally defined brain areas. In the monkey brain, face areas are comprised of a high proportion of face-selective cells, however, such cells can be found outside these fMRI defined face patches and throughout the Superior Temporal Sulcus (STS). It is not known whether the face-selective cells outside of the face patches behave the same way as the cells inside the face patches. In this study, we asked whether inversion has a differential effect on the responses of face-selective cells depending on anatomical position. We first localized, in two monkeys, several patches in STS that responded more to faces than other objects. Then, we recorded single units in two of these face patches (ML and AL) in addition to control positions between ML and AL. We searched for each cell using two categories of objects (faces and non-face objects) presented either in their canonical orientation or upside down. Face selectivity was defined using a Face-Selectivity Index (FSI = (mean responsefaces – mean responsenonface objects) / (mean responsefaces + mean responsenonface objects)). Face-selective cells that were recorded in ML responded more strongly to upright than inverted faces. The inversion of the preferred upright face had the same effect on the average response of the face-selective cells in the anterior face patch (AL). Additional recordings indicated that these effects in ML and AL were independent of retinal position. For face-selective cells that were found outside the face patches there was only evidence of a FIE when the search stimuli were upright. These findings imply that, while inversion has an impact on face representations in fMRI identified face patches, the same is not necessarily true for representations built elsewhere in monkey STS.