Lasyapriya Pidaparthi¹ (), Frank Tong¹; ¹Vanderbilt University

In our daily lives, we are often confronted with salient distractors that we must effectively ignore in order to focus on our task. Research has shown that object-based attentional modulation can selectively boost the neural responses in both low- and high-level visual areas to prioritize task-relevant information, even in the presence of a competing, overlapping distractor (Cohen & Tong, 2015; O’Craven et al., 1997). However, we do not fully understand how this attentional modulation is affected by distractor saliency, particularly when the distractor is overwhelmingly more salient than the target. Here, we investigated this question by using an fMRI decoding-based approach (Kamitani & Tong, 2005) to examine top-down attentional biasing of neural responses along the ventral visual pathway, including early visual areas (V1-hV4) and category-selective regions (fusiform face area/FFA; parahippocampal place area/PPA). In our experiment, participants were presented with overlapping face-house images that varied in their relative contrast. On each trial, participants attended to either face or house and indicated whether the attended object across sequential pairs of images was the same or different. On some trials, only a single face or house was presented (“attend-single”). To measure object-based attentional selection, we trained an SVM classifier on the normalized multivariate responses during attend-single trials and tested how well it could decode the attended object from responses during attend-face and attend-house trials. Notably, modulatory effects of object-based attention were found throughout visual areas V1-hV4 and FFA/PPA, evidenced by greater than 50% decoding bias for the target, even when the distractor was twice as strong as the target in stimulus contrast. Further, we observed a trend of strengthening attentional modulation as one ascends the visual hierarchy. Overall, our novel findings showcase the efficacy of object-based attention in overcoming stimulus-driven effects to strengthen the neural representation of an attended object.

Acknowledgements: This research was supported by NIH grants R01EY035157 and R01CA240274 to FT, T32EY007135 to LP, and P30EY008126 to the Vanderbilt Vision Research Center.