Temporal fluctuations in vigilance and neural networks after sleep deprivation

Abstract Vigilance instability in the sleep‐deprived state was deemed to result from the imbalance in thalamic‐FPN‐DMN circuits (FPN: frontoparietal network; DMN: default mode network), but the behavioural correlation of this neural hypothesis is still unclear. To address this issue, we applied dynamic functional connectivity (DFC) analysis on the task‐based fMRI data and detected high arousal state (HAS) and low arousal state (LAS). Relative to HAS, LAS demonstrated higher positive connectivity within task‐positive networks (TPN), attenuated TPN‐DMN anti‐correlation, and greater anti‐correlation between cerebral and subcortico‐cerebellar networks. Critically, DFC differences between HAS and LAS were correlated with the ongoing vigilance performance in the sleep‐deprived state. The current findings confirmed a direct link between vigilance instability and DFC in the thalamic‐FPN‐DMN circuits. In particular, we postulated that the integration within task‐related system and segregation between task‐related system and the subcortico‐cerebellar system might be the critical neural markers underlying vigilance instability in the sleep‐deprived state.