Abstract
Attentional deployment during visual search is controlled by a distributed network of brain areas including the frontal eye field (FEF). When search is inefficient (e.g., searching for a T among rotated L's) visual activity in FEF selects the target location before the macaque homologue of the N2pc (m-N2pc), a component of the event-related potential (ERP) indexing attentional selection. However, during pop-out search tasks (e.g., a green target among red distractors) it has been proposed that bottom-up processing is sufficient to perform the task and the focusing of attention is unnecessary. While monkeys performed a color pop-out search task, we simultaneously recorded spikes from individual neurons in the FEF along with FEF local-field potentials (LFPs) and ERPs over lateral visual cortex. As expected, saccadic response time and the time of target selection for all neural signals did not vary with the set size of the array. Contrary to proposals that pop-out search tasks are performed in the absence of top-down attentional guidance, the target was selected earliest by the FEF neurons, next in the FEF LFPs, and latest in the m-N2pc. These results suggest that feedback from attentional control structures like FEF is involved in generating the macaque homologue of the human N2pc component and that the temporal cascade of selective activity is similar for both efficient and inefficient search tasks.
F32-EY019851, T32-EY007135, R01-EY08890, P30-EY08126, and P30-HD015052 and the E. Bronson Ingram Chair in Neuroscience.