An investigation of the neural basis of face individuation through spatiotemporal pattern analysis

摘要

What neural system is responsible for face individuation and what is its structure? Extensive research on the topic offers divergent responses to the first question and rather few clues to the second. Our work deals with these issues by appealing to a sequence of multivariate pattern analyses applied to functional magnetic resonance imaging (fMRI) data. Specifically, we combine information-based brain mapping and dynamic discrimination analysis to address the first question. The goal of this analysis is to locate spatiotemporal patterns cable of supporting face classification at the individual level. Our results reveal an a??individuation networka?? of anterior temporal and fusiform areas. Moreover, they provide the first demonstration that the bilateral fusiform face area (FFA) responds with distinct activation patterns to different face identities. The second part of our work examines the distribution of diagnostic information across this network using recursive feature elimination. Our results show that information is distributed evenly among anterior regions. Also, an information-based network analysis suggests that one region located in the right anterior fusiform gyrus plays the role of a hub within the neural system responsible for face individuation. This work explores the specifics of distributed processing in the context of face perception; however, more generally, it speaks to its informational basis irrespective of domain in the context of functionally-defined cortical networks. Finally, our research explores ways in which the analyses above can integrate functional connectivity in order to recover the dynamics of the information flow within the face individuation network.