Abstract
Visual attention can be directed voluntarily to spatial locations and to different visual features such as shape and color. We used functional magnetic resonance imaging (fMRI) to investigate the neural correlates of endogenous attention to those three visual dimensions: shape, color, and position. Simple figures of different shapes and colors were presented in each of the four visual quadrants. Every two seconds, the stimuli changed their shape and color. The outline of each stimulus was notched, and the notch randomly changed position every two seconds. Observers attended to a specific shape, color, or position in the display as instructed by letters presented at fixation and reported the location of the notch in the attended stimulus. This allowed us to assess the deployment of attention to the correct stimulus. In a control condition, observers ignored the shapes and reported the location of a notch in the fixation square. High accuracy rates for every condition suggest observers attended equally to each visual dimension. Eye-tracking in the scanner confirmed that observers kept good fixation during the experiment. The analysis of fMRI data revealed distinct but overlapping cortical networks engaged in the different experimental conditions compared to the control condition. While there was a core posterior parietal region active in all cases of endogenous attention, attention to colors resulted in strong ventral-occipital activation, attention to positions involved strong posterior-parietal activation, and attention to shapes resulted in strong ventral-occipital and posterior-parietal activation. Our study demonstrates that endogenous attention recruits distinct but overlapping cortical networks depending on the attended visual dimension. The results argue for a modular view of voluntary visual attention.