Abstract

Visual attention enables the selection of relevant information in the environment for further processing. Spatial attention results in changes in spatial receptive fields (RFs) of single neurons and fMRI voxels (vRFs), including warping, shifting, and rescaling (e.g., Womelsdorf et al, 2006; Klein et al, 2014; Vo et al, 2017). In many of these studies, attention was directed either to the RF mapping stimulus or the fixation point, which typically results in an increase in measured vRF size in extrastriate cortex when attending the mapping stimulus, or attention was directed to a fixed location while vRFs were measured with an ignored mapping stimulus, which typically results in vRF position shifts. However, in all these studies, the entire mapping stimulus was either attended or ignored. Thus, it remains unknown how attentional selection of a subpart of the stimulus impacts vRFs. Here, we acquired fMRI data while participants viewed a traversing bar stimulus broken into 3 segments. Each segment contained moving black and white dots. Across scanning runs, we manipulated which segment(s) of the bar participants attended using a central cue. Importantly, the stimulus sequence was fixed across scanning runs, so changes in observed signals necessarily result from a change in the locus of attention. Direct comparison of activation timeseries revealed similar spatial selectivity, but a dynamic reweighting of activation levels when the bar passed each voxel’s preferred position based on which bar segment was attended. This reweighting was more pronounced in extrastriate visual cortex as compared to V1, consistent with previous reports that attentional modulation of vRF properties is smallest in V1. In ongoing work, we are comparing nonlinear vRF models that jointly incorporate the stimulus drive, attentional locus, and a combination of these factors (Kay et al, 2013; Klein et al, 2014; Mackey et al, 2017; van Es et al, 2018).