hide
Free keywords:
-
Abstract:
Introduction:
A classical model of attention holds that independent neural networks realize stimulus-driven orienting and top-down control of attention (Petersen and Posner, 2012). Questioning full independence, the two functions do, however, engage partially overlapping networks with activation increases in cingulo-opercular regions such as anterior insula (Nelson et al., 2010) and a more complex pattern of activation (stimulus-driven orienting), and deactivation (top-down control) in temporoparietal junction (Corbetta et al., 2008). Therefore, concurrent demand of both functions may result in additional response costs (Fan et al., 2009). To characterize such overlap and interference on a neural level and thus determine whether a common bottleneck underlies stimulus-driven and top-down control of attention, we used an fMRI task that isolates individual from concurrent demands in both functions.
Methods:
From 332 subjects of a large scale longitudinal study (Singer et al., 2015), 282 clean data-sets were available. We recorded T2*-weighted echo-planar images during the 15-minute task and during 5 minutes of eyes open resting-state, as well as T1-weighted structural scans on a 3T Siemens Verio scanner. The task (Fig. 1A) involved a 2x2 design, engaging top-down control through flanker-target conflict (50% of trials); and inducing stimulus-driven orienting through spatial cueing of the invalid (20%) vs. valid (80%) target location. Additionally, intermittent thought probes assessed the extent of task-unrelated thought (TUT). We used SPM 8 to model task induced activations, task dependent, and task-unrelated resting-state functional connectivity, with FWE corrected threshold at p < 0.05.
Results:
Behavioral data showed interactions with super-additively increased response times (F(1,281) = 47.47; p < 0.001) and error rates (F(1,281) = 286.55; p < 0.001) during concurrent demands in stimulus-driven orienting and top-down control (Fig. 1B). Furthermore, for error rates, inter-individual differences in both capacities were significantly correlated (rs = .17; p < .01). Neural activations for stimulus-driven orienting and top-down control overlapped in the cingulo-opercular network and in fronto-parietal regions (Fig. 2A). Importantly, activity in left AI increased super-additively when both functions were demanded at the same time (Fig. 2B). These increases were mirrored by non-additive decreases of activity in the default mode network (DMN), including posterior TPJ, regions where activity increased with TUT (Fig. 2B and
Fig. 2C). The deactivations in posterior TPJ were spatially separate from stimulus-driven orienting related activation increases in anterior TPJ, a differentiation that we replicated in task-free resting state (Fig. 3). Furthermore, analysis of task-induced changes in functional connectivity revealed enhanced negative coupling between posterior TPJ and AI during concurrent orienting and top-down control of attention, suggesting an inhibitory relationship (Fig. 2D).
Conclusions:
Results demonstrate shared mechanisms between stimulus-driven orienting and top-down control of attention, two functions that have previously been thought to be independent. Furthermore, they suggest a central role of AI in supporting both functions by down-regulation of internally directed processes in DMN. These findings are consistent with recent network-based approaches to human brain function, demonstrating involvement of domain-general brain networks in attention and a central role of anterior insula in their regulation.
