The effects of attentional scope on voxel receptive fields and population codes for space

Abstract

Prior studies in both humans and macaques have investigated how spatial attention alters the response properties of spatial encoding units -- e.g. single neurons in macaques, or single voxels in humans – across different cortical areas. These studies have all shown that spatial receptive fields (RFs) change their center, shape, and gain with spatial attention (for review, see Anton-Erxleben & Carrasco 2013; in humans, see Klein et al. 2014; Sprague & Serences 2013; de Haas et al. 2014; Kay et al. 2015; Sheremata & Silver 2015; Vo et al. 2017). In these studies, subjects typically attend to a single stimulus of a fixed size and scope (but see Niebergall et al. 2011). It remains unknown how patterns of RF changes across the visual field might support spatial attention divided across many stimuli. We had human participants perform a task in which they focused their attention on 1 of 3 locations or divided their attention across all 3 locations. As a control, they also performed a contrast change discrimination task at fixation. We threshholded task difficulty to equate performance across conditions. They then performed the task in an fMRI scanner as we measured BOLD signals across the whole brain. We compare how (1) voxel RFs (vRFs) changed across attention condition, and (2) how population codes for space changed across condition by using an inverted encoding model. Preliminary data suggest that altering the spatial scope of attention causes widespread changes in the preferred center, size, and gain of vRFs. These modulations across the whole population of vRFs jointly influence the quality of information encoded about spatial location in each condition. Meeting abstract presented at VSS 2018