In addition, in all six monkeys several regions were reproducibly more active to Shapes (both Learned symbols and Untrained shapes) than to Faces (conjunction of L > F AND U > F contrast maps) (Figure 4, Figure 5 and Figure 6, green click here patches). Three Shape-selective regions (s1, s2, s3, posterior to anterior) were consistent between the two hemispheres for each monkey, so we again
averaged the two hemispheres together to project each monkey’s Shape selectivity maps onto a common hemisphere (Figure 6, green patches). Again, by inspection of Figure 6, several regions are commonly Shape selective. The maximally selective voxels in each of the three largest Shape selective regions for each monkey are listed in Table S1. The posterior-most Shape patch (s1) was consistently localized ventral and slightly posterior PLX3397 chemical structure to Face patch f1 in posterior area TEO or in anterior V4, at the anterior tip of IOS, with maximal overlap at A2. The middle Shape patch (s2) extended from the bank of the STS near the anterior tip of PMTS out onto the inferotermporal gyrus, maximal overlap at A4 mostly within area TEpd or area TEO. The anterior most Shape patch (s3) was less consistent between monkeys; it was located in TEa/TEm, varying in position from A12 to A16. Shape selective regions that are distinct
from Face selective patches have also been previously described (Denys et al., 2004 and Sawamura et al., 2005). In all six monkeys, the relative category-selective regions formed three pairs of regions more responsive to Faces than to Symbols (Learned and Untrained) or the reverse, distributed along the inferotemporal gyrus (Figure 6A). The locations of the two posterior pairs of patches roughly correspond to the borders between the major subdivisions of the ventral temporal lobe (Boussaoud et al.,
1991, Desimone and Ungerleider, 1989 and Saleem and Logothetis, 2007)—V4/TEO and TEO/TE (Figure 6A). The anterior patches may be located at the TE/TG border, but their position was too variable to really say. Because Fossariinae our stimuli covered only the central visual field, the patches may correspond to foveal confluences between areas (Kolster et al., 2009). Alternating face, body, and object selective regions have been described previously in macaque temporal lobe (Bell et al., 2009, Denys et al., 2004 and Op de Beeck et al., 2008) and have been proposed to represent alternating regions selective for animate versus inanimate categories (Bell et al., 2009 and Op de Beeck et al., 2008). Our results are consistent with this hypothesis, and in one of our monkeys we confirmed that the regions activated by Shapes > Faces were also selectively activated by images of inanimate objects (data not shown).