Abstract
While animal and psychophysical studies suggest that attention can enhance visual processing by multiplicatively increasing the gain of an attended item, human fMRI studies instead report that attention modulates overall responsivity, and does not seem to interact with stimulus intensity. A potential reason for this disparity in results between the different methods is that population-based measurements using fMRI find predominantly linear contrast response functions, lacking response saturation. Recent work from our lab has illustrated that leveraging adaptation allows for reliable measurements of saturating contrast response functions. Here, we sought out to test how attention modulates the population contrast response function (pCRF), when we are able to capture the nonlinear relationship between stimulus contrast and BOLD response. We used an fMRI adaptation paradigm to measure BOLD responses in early visual cortex (V1-V3). Participants viewed full-field stimuli displays composed of radially oriented, cortically magnified gratings varying in contrast throughout a scan (9 contrast levels, spaced between 0-88% Michelson contrast). Participants were cued before the onset of each contrast presentation to either covertly attend to the stimulus, performing a color probe detection task, or to perform a demanding task at fixation, drawing attention away from the stimulus. Using deconvolution analyses to estimate the BOLD response for each contrast level, we were able to capture nonlinear pCRFs of individual voxels within each visual area. Importantly, we found that the influence of attention was best summarized by a combination of an additive and contrast gain modulation. In sum, our results demonstrate that attentional modulation of the pCRF as measured with fMRI is not purely additive, but additionally exhibits multiplicative gain increases.