Abstract
The stimulus evoked fMRI signal of single voxels in human visual cortex can be accurately predicted using a population receptive field (pRF) model (Dumoulin, 2008). We have adapted this approach to explore the effects of visual attention on single voxels. While fixating at a center marker, subjects (4) were cued to covertly attend to a specific target location within a dartboard-like array of stimulus segments that slowly rotated about fixation at 0.75 rpm. In attention runs, subjects continuously tracked a single target at 8[sup]o[/sup] eccentricity within the array. In sensory runs, the previously attended segment was presented in isolation while subjects performed an attention task at fixation. In the attention condition, voxels having pRFs positioned along the trajectory of the target segment were differentially activated when the focus of attention passed over their pRF. These voxels were also activated in the sensory condition when the isolated stimulus segment passed over the pRF. Thus, the duration of fMRI activation was proportional to either (1) the width of the attentional focus or (2) the width of the stimulus segment, thereby allowing a direct comparison of the relative sizes of the two. To compute the relative widths precisely for each voxel, we fit the empirical data with a timecourse generated by a model composed of the stimulus sequence, the attentional field, the voxel's estimated pRF, and the temporal hemodynamic response function. The resulting model parameters were compared to estimate the relative sizes of the attentional field and the sensory stimulus. Recent models of visual attention indicate that the size of the attentional field relative to a visual stimulus and sensory pRF are key factors in determining the behavioral effects of attention (Reynolds, 2009). The method described here permits empirical measurement of the attentional field size thereby providing key information for models of attention.
Meeting abstract presented at VSS 2013