Melisa Menceloglu¹ (), Nora Indenberga¹, Sydney T. George¹, Seeun Oh¹, Søren K. Andersen², Taosheng Liu¹; ¹Michigan State University, ²University of Southern Denmark

Attention to visual features not only facilitates neural responses to task-relevant stimuli but also to task-irrelevant stimuli at other locations that share the attended feature. This global attentional facilitation of features across the visual field has been reported during active selection (e.g., attending to blue dots among other colored dots). Here, we investigated whether this global spread also occurs during preparatory attention (e.g., preparing to select blue among other colored dots, before the presentation of such dots). We presented two overlapping fields of blue and orange dots to the left and right of a central fixation cross. We recorded EEG while observers (N=20) were cued to attend to a specific dot field (e.g., blue dots on the left) to detect coherent motion at threshold. In the immediate-onset condition, dot fields on both sides were presented simultaneously following the cue; whereas in the delayed-onset condition, attended-side dot fields were presented after a ~1.5-2.7s delay, extending the preparatory period. We used the frequency tagging method to measure the effects of attention on the visual cortical responses by flickering the dot fields at four different frequencies and comparing their inter-trial phase coherence (ITPC). We observed a local feature-based attention effect on the attended side: ITPC elicited by the attended-color dots was stronger than unattended-color dots. We also observed a global feature-based attention effect on the unattended side, in the immediate-onset condition: ITPC elicited by the dots matching the attended color was stronger than unattended color. Notably, we observed a robust, albeit reduced, global effect in the delayed-onset condition during the preparatory period. These results show that preparatory attention to a feature can bias the analysis of that feature in the absence of a stimulus in the spatially attended region, suggesting that the global spread of feature-based attention can be a purely endogenous process.

Acknowledgements: This work was supported in part by grants from NIH (R01EY032071) and NSF (2019995).