Interactions between phasic alerting and consciousness in the fronto-striatal network

Abstract

Only a small fraction of all the information reaching our senses can be the object of conscious report or voluntary action. Although some models propose that different attentional states (top-down amplification and vigilance) are necessary for conscious perception, few studies have explored how the brain activations associated with different attentional systems (such as top-down orienting and phasic alerting) lead to conscious perception of subsequent visual stimulation. The aim of the present study was to investigate the neural mechanisms associated with endogenous spatial attention and phasic alertness, and their interaction with the conscious perception of near-threshold stimuli. The only region demonstrating a neural interaction between endogenous attention and conscious perception was the thalamus, while a larger network of cortical and subcortical brain activations, typically associated with phasic alerting, was highly correlated with participants' conscious reports. Activation of the anterior cingulate cortex, supplementary motor area, frontal eye fields, thalamus, and caudate nucleus was related to perceptual consciousness. These data suggest that not all attentional systems are equally effective in enhancing conscious perception, highlighting the importance of thalamo-cortical circuits on the interactions between alerting and consciousness.

Figure 1. Sequence and timing of events in a given trial.

The bottom-right panel shows an example of left and right tilted Gabors with the minimum (2°) and maximum (10°) tilt used in the study.

Figure 2

(A) Mean RT in the objective task for consciously reported Gabors as a function of Alerting and Validity. (B) Mean Michelson contrast values used to achieve ~50% “seen” and “unseen” targets (in the subjective task) as a function of Alerting State and Validity. Error bars represent standard errors.

Figure 3

(A) Brain renderings showing activations for the Seen > Unseen whole-brain contrast (p < 0.05, FWE cluster-level corrected). (B) Percent (%) signal change from functionally identified ROIs as a function of Awareness, Alerting State, and Validity, showing a significant Alerting State x Awareness interaction (Table 2). All regions demonstrated increased % signal change for seen as compared to unseen reports, which was enhanced when the alerting tone was not presented. The center of mass of each ROI is indicated in parenthesis, and the volume in mm is indicated in squared brackets. The following ROIs were also analyzed and demonstrated no significant effects: Left IPL (−38, −48, 45) [6152], left SPL (−24, −67, 51) [3408], left thalamus [(−11, −18, 6) [7792] and right thalamus (11, −16, 7) [5032]].

Figure 4

(A) Representation of the brain regions demonstrating an Alerting State x Awareness x Time interaction. All these regions demonstrated larger signal intensity for seen targets than for unseen targets when no alerting tone was presented. The figure represents the time interval where the differences in signal intensity started to be significant in each region, with cue onset at time 0. (B) Averaged signal intensity time courses for the regions demonstrating a significant Alerting State x Awareness interaction. (C) Averaged signal intensity time courses of the region demonstrating a significant Validity x Awareness interaction (left and right Thalamus). Time 0 represents the moment of cue onset; the target was presented 2 s (an MR frame) later. Asterisks represent significantly larger signal intensity for seen versus unseen targets at different time points (p < 0.05, Bonferroni corrected).

Figure 5. Pairwise functional connectivity for regions demonstrating a significant interaction between Alerting State and Awareness in the ANOVA of the beta-correlation values.

Line thickness represents the coupling strength among regions.

Figure 6. Brain renderings showing whole-brain functional connectivity analysis using the right ACC as a seed for Seen > Null and Unseen > Null contrasts (p < 0.01, FWE voxel-level corrected).