Time-locked activities of resting-state networks that transiently occur in human resting-state brain activities

Spontaneous brain activities at rest are organized into resting-state networks, such as the default mode network. While the default mode network is assumed to interact with other networks to integrate distributed information, its fine-scale temporal dynamics remain unclear. Here, we demonstrate how the default mode network interacts with other networks at millisecond-scale temporal resolution. Using a large dataset of resting-state magnetoencephalography and our recently proposed method, we estimated the spatiotemporal patterns that recur in resting-state brain activity and then examined the interactions across the resting-state networks in these patterns. In the alpha/beta band component of the spatiotemporal patterns, the default mode network and the visual/sensorimotor network alternately appear with a lag of about 24/13 ms, corresponding to about π /2 rad in the alpha/beta oscillation. These alternating and orthogonal activities of the resting-state networks may reflect the brain’s strategy to achieve both the integration and separation of visual and sensorimotor information via the default mode network. Furthermore, we reveal how these activities consecutively move between the default mode network and the visual/sensorimotor network in tens of milliseconds. They move in different ways depending on their direction, suggesting that distinct mechanisms generate the movements in different directions.