Whole-brain network dynamics underlying transitions between sleep and wake
Laura Lewis, Harvard University
Cognition and behavior vary dramatically over the course of the day, as the brain transitions between alert, inattentive, drowsy, and sleep states. How brain networks dynamically reorganize to create these diverse arousal states and modulate neural computation is not well understood. Studies of large-scale network dynamics have been limited by the fact that conventional neuroimaging methods cannot capture whole-brain activity at subsecond timescales. We developed a new approach to noninvasively measuring local dynamics throughout the whole brain simultaneously using ‘fast fMRI’, enabling direct imaging of cortical and subcortical oscillations in humans. We integrated fast fMRI with simultaneous EEG to image human subjects as they fell asleep, and identified transitory shifts in patterns of global thalamocortical dynamics that predict moment-to-moment arousal state. These results demonstrate that rapid changes in large-scale network function can be detected through new techniques for fast whole-brain neuroimaging, identify distinct neural activity patterns that signal transitions into and out of sleep, and suggest broad future potential for these approaches to identify the neural mechanisms that flexibly control brain states and cognition.