Pre-stimulus cortical state predicts context-dependent expression of learned distractor suppression

Abstract

Suppressing visual distraction depends on learned attentional settings and the neural state preceding stimulus onset. We reanalyzed an fMRI visual-search dataset (N = 34) to test whether pre-stimulus global signal (GS), a broad state-sensitive BOLD measure, predicts subsequent performance and cortical responsivity across different learned distractor-suppression contexts. Participants searched for an orientation target while ignoring same-dimension orientation or different-dimension color distractors. Spatial regularity applied to the same-dimension distractor in one group and the different-dimension distractor in the other. Same-dimension distractors produced strong behavioral interference and recruited Dorsal Attention and Salience/Ventral Attention networks, whereas different-dimension distractors produced little interference. Spatial learning was also clearer when regularity applied to the target-relevant same dimension. Within these learning contexts, pre-stimulus GS predicted response speed in opposite directions across groups: higher GS preceded slower responses in the same-dimension spatial-bias group but faster responses in the different-dimension spatial-bias group. And pre-stimulus Default Mode Network activity showed a convergent context-dependent pattern. Higher pre-stimulus GS preceded reduced activity across Dorsal Attention, Salience/Ventral Attention, Control, Early Visual, and Default Mode regions. Although this neural attenuation was broadly expressed across conditions and groups, its association with the GS-RT relationship differed between learning contexts, most clearly on different-dimension trials. Together, these findings indicate that pre-stimulus GS reflects a broad modulation of subsequent cortical responsivity, whereas its behavioral consequences-and their relationship to regional neural attenuation-depend on the attentional-control context established by distractor dimension and spatial regularity.