Prefrontal Control of Visual Distraction

Abstract

Avoiding distraction by conspicuous but irrelevant stimuli is critical to accomplishing daily tasks. Regions of prefrontal cortex control attention by enhancing the representation of task-relevant information in sensory cortex, which can be measured in modulation of both single neurons and event-related electrical potentials (ERPs) on the cranial surface [1, 2]. When irrelevant information is particularly conspicuous, it can distract attention and interfere with the selection of behaviorally relevant information. Such distraction can be minimized via top-down control [3-5], but the cognitive and neural mechanisms giving rise to this control over distraction remain uncertain and debated [6-9]. Bridging neurophysiology to electrophysiology, we simultaneously recorded neurons in prefrontal cortex and ERPs over extrastriate visual cortex to track the processing of salient distractors during a visual search task. Critically, when the salient distractor was successfully ignored, but not otherwise, we observed robust suppression of salient distractor representations. Like target selection, the distractor suppression was observed in prefrontal cortex before it appeared over extrastriate cortical areas. Furthermore, all prefrontal neurons that showed suppression of the task-irrelevant distractor also contributed to selecting the target. This suggests a common prefrontal mechanism is responsible for both selecting task-relevant and suppressing task-irrelevant information in sensory cortex. Taken together, our results resolve a long-standing debate over the mechanisms that prevent distraction, and provide the first evidence directly linking suppressed neural firing in prefrontal cortex with surface ERP measures of distractor suppression.

Keywords: attention capture; executive control; extrastriate cortex; frontal eye field; inhibition; prefrontal cortex; suppression.

Figure 1

Task and Electrode Penetration Maps. (A) Visual search task. Monkeys fixated for a variable duration (500–1000 ms), at which point the fixation point extinguished and the search array appeared. Monkeys were trained to covertly search for the target, and were rewarded for a first saccade to the target item. A salient, irrelevant distractor appeared unpredictably on half of all trials, which monkey were trained to ignore. (B) Penetration maps for recordings in each monkey including the total number of units isolated at each location, regardless of task-related modulation. AS = Arcuate Sulcus; PS = Principal Sulcus

Figure 2

Salient distractor processing in behavior and prefrontal neurons. (A) Effects of salient distractor on visual search accuracy when color singleton distractor was present (red) relative to absent (black). Monkeys Ga and He were exposed to the color singleton distractor early in training, and learned to avoid distraction. Monkey Da performed the form visual search task for many months before exposure to the color singleton distractor, and subsequently suffered distraction when the distractor was introduced. As training progressed, behavioral distraction decreased but never disappeared. The salient distractor exerted a significant influence on behavior during early and late periods of neurophysiological data collection for this monkey. * P < 0.05, ** P < 0.01. (B) Mean (± SEM error bars) population responses of FEF neurons when the target (thick), non-salient distractor (black), or salient distractor (red) appeared within the receptive field for each monkey. The target was selected through elevated discharge rates relative to the non-salient distractor in all monkeys. The salient distractor was suppressed by reduced discharge rates relative to the non-salient distractor only in the monkeys that did not exhibit behavioral distraction. (C) Distribution of target selection and distractor suppression for all neurons with a significant visual response. The response ratio was calculated by dividing the magnitude of responses to targets or salient distractors by responses to nonsalient distractors in the interval 50–150 ms following presentation of the search array. Values greater than 1.0 indicate enhancement and values less than one indicate suppression. Both target enhancement and distractor suppression were consistent features in the FEF of the monkeys that were not distracted. Only target enhancement was observed in the FEF of the monkey that was distracted by the color singleton.

Figure 3

Salient distractor processing in prefrontal neurons and extrastriate ERP. (A) Configurations of search array used for analyses. Dashed line indicates receptive field of FEF neurons in the visual hemifield contralateral to cranial electrode. (B) Mean (± SEM error bars) discharge rates (top) and voltage (bottom) combined for monkeys Ga and He. Neural signals are aligned on presentation of the search array, and responses were truncated 10 ms prior to the saccade. The response to the target (thick) becomes elevated relative to the response to a non-salient distractor (thin black) whether the salient distractor was present (left) or absent (right). The response to the salient distractor (red) becomes suppressed relative to the response to a non-salient distractor. Vertical lines indicate the time at which responses deviated significantly from one another. Both target selection and distractor suppression emerged earlier in FEF and later in extrastriate ERP responses.