Neural mechanisms of selective attention: fMRI studies of the Stroop task

We used fMRI to investigate the manner in which top-down attention modulates processing in posterior brain areas during the Stroop task. We hypothesized that attention could enhance color processing, suppress word processing, or both. We functionally defined color and word areas and examined how they were modulated by attention during the Stroop task. Contrary to previous Stroop imaging experiments but consistent with evidence from other tasks, we found enhancement of task-relevant color areas in Experiment 1.;There has been considerable disagreement in the literature regarding whether word areas are enhanced or suppressed during the Stroop task. In Experiment 2 we found direct evidence for suppression of word areas. Further, since we used abstract control stimuli, this result cannot be attributed to concreteness. We also confirmed the enhancement hypothesis in both hemispheres and extended the results of the previous experiment via group analyses showing activation in frontal and parietal lobes early in the experiment, as well as increased enhancement of color areas with practice over time.;In previous work, preparatory attentional modulations were hypothesized as a mechanism for biasing processing in task-relevant areas by top-down control. In Experiment 3 we used an event-related fMRI design to test the hypothesis that preparatory attentional modulation biases processing in the Stroop task. We found deactivation of task-irrelevant color areas preceding incongruent word identification trials.;Our results are most consistent with an inhibitory mechanism for endogenous attention and an excitatory mechanism for exogenous attention. Together these mechanisms may account for the enhanced processing of attended features and the suppressed processing of ignored features we observed.