The activation of interactive attentional networks

Abstract

Attention can be conceptualized as comprising the functions of alerting, orienting, and executive control. Although the independence of these functions has been demonstrated, the neural mechanisms underlying their interactions remain unclear. Using the revised attention network test and functional magnetic resonance imaging, we examined cortical and subcortical activity related to these attentional functions and their interactions. Results showed that areas in the extended frontoparietal network (FPN), including dorsolateral prefrontal cortex, frontal eye fields (FEF), areas near and along the intraparietal sulcus, anterior cingulate and anterior insular cortices, basal ganglia, and thalamus were activated across multiple attentional functions. Specifically, the alerting function was associated with activation in the locus coeruleus (LC) in addition to regions in the FPN. The orienting functions were associated with activation in the superior colliculus (SC) and the FEF. The executive control function was mainly associated with activation of the FPN and cerebellum. The interaction effect of alerting by executive control was also associated with activation of the FPN, while the interaction effect of orienting validity by executive control was mainly associated with the activation in the pulvinar. The current findings demonstrate that cortical and specific subcortical areas play a pivotal role in the implementation of attentional functions and underlie their dynamic interactions.

Keywords: Alerting; Attentional networks; Executive control; Orienting; fMRI.

Figure 1

Schematic of revised Attention Network Test (ANT-R). In each trial, depending on the cue condition (none, double, and valid or invalid cues), a cue box flashes for 100 ms. After a variable duration (0, 400, or 800 ms), the target (the center arrow) and two flanker arrows on both the left and right side (congruent or incongruent) are presented for 500 ms. Participants must indicate the target’s direction. Before the target appears, a cue in the form of a box flashing on one or both sides is displayed. The cue can be valid, which predicts the target position correctly, or invalid, which predicts the opposite position. There is also a double cue condition, in which both boxes flash, to provide temporal but not spatial information, while in the no cue condition no cue is presented. The post-target fixation period varies between 2000 and 12,000 ms. Note: The location congruency manipulation was not treated as a manipulation in data analysis in this study.

Figure 2

Attentional effects and interactions in terms of (A) reaction time (RT in ms), and (B) error rate (%). Error bars represent the standard error of the mean.

Figure 3

Brain regions showing increased activation associated with the alerting effect.

Figure 4

Brain regions showing increased activation associated with the (A) disengaging, (B) moving + engaging, and (C) validity effects of orienting.

Figure 5

Brain regions showing increased activation associated with the flanker conflict effect.

Figure 6

Brain regions showing increased activation associated with (A) the alerting by flanker conflict interaction effect, and (B) the conjunction of alerting and flanker conflict effects.

Figure 7

Brain regions showing increased activation associated with (A) the validity by flanker conflict interaction effect, and (B) the conjunction of validity and flanker conflict effects.