Spatiotemporal dynamics of high-gamma activities during a 3-stimulus visual oddball task

Abstract

Although many studies have investigated the neural basis of top-down and bottom-up attention, it still requires refinement in both temporal and spatial terms. We used magnetoencephalography to investigate the spatiotemporal dynamics of high-gamma (52-100 Hz) activities during top-down and bottom-up visual attentional processes, aiming to extend the findings from functional magnetic resonance imaging and event-related potential studies. Fourteen participants performed a 3-stimulus visual oddball task, in which both infrequent non-target and target stimuli were presented. We identified high-gamma event-related synchronization in the left middle frontal gyrus, the left intraparietal sulcus, the left thalamus, and the visual areas in different time windows for the target and non-target conditions. We also found elevated imaginary coherence between the left intraparietal sulcus and the right middle frontal gyrus in the high-gamma band from 300 to 400 ms in the target condition, and between the left thalamus and the left middle frontal gyrus in theta band from 150 to 450 ms. In addition, the strength of high-gamma imaginary coherence between the left middle frontal gyrus and left intraparietal sulcus, between the left middle frontal gyrus and the right middle frontal gyrus, and the high-gamma power in the left thalamus predicted inter-subject variation in target detection response time. This source-level electrophysiological evidence enriches our understanding of bi-directional attention processes: stimulus-driven bottom-up attention orientation to a salient, but irrelevant stimulus; and top-down allocation of attentional resources to stimulus evaluation.

Figure 1. Experimental paradigm of 3-stimulus oddball task.

Figure 2. High-gamma activity relevant to a 3-stimulus oddball task (p<0.05, FWE corrected).

Figure 3. Time course of high-gamma power changes in each region of interest.

Blue bars indicate significant differences between the non-target condition and the standard condition, while red bars indicate significant differences between the target condition and the standard condition (p<0.05, FDR corrected).

Figure 4. Time-frequency representation of high-gamma power changes in each region of interest.

Warm colors indicate synchronization, while cold colors indicate desynchronization.

Figure 5. Time course of high-gamma power changes in response-locked analysis.

Figure 6. Significant correlation between reaction time and the strength of high-gamma imaginary coherence or regional high-gamma power.

(a) Reaction time vs. the strength of high-gamma imaginary coherence between the left IPS and the left MFG during the time window of 300 to 400 ms and (b) during the time window of 400 to 500 ms. (c) Reaction time vs. the strength of high-gamma imaginary coherence between the left IPS and the left MFG during the time window of 400 to 500 ms and (d) during the time window of 500 to 600 ms. (e) Reaction time vs. the high-gamma power in the left thalamus during the time window of 300 to 400 ms.