Mining cross-frequency coupling microstates (CFCμstates) from EEG recordings during resting state and mental arithmetic tasks.

The functional brain connectivity has been studied by analyzing synchronization between dynamic oscillations of identical frequency or between different frequencies of distinct brain areas. It has been hypothesized that cross-frequency coupling (CFC) between different frequency bands is the carrier mechanism for the coordination of global and local neural processes and hence supports the distributed information processing in the brain. In the present study, we attempt to study the dynamic evolution of CFC at resting-state and during a mental task. The concept of CFC microstates (CFCμstates) is introduced as emerged short-lived patterns of CFC. We analyzed dynamic CFC (dCFC) at resting-state and during a comparison task by adopting a phase-amplitude coupling (PAC) estimator for [δ phase-γ-amplitude] coupling at every sensor. Modifying a well-established framework for mining brain dynamics, we show that a small sized repertoire of CFCμstates can be derived so as to encapsulate connectivity variations and further provide novel insights into network's functional reorganization. By analyzing the transition dynamics among CFCμstates, in both tasks, we provided a clear evidence about intrinsic networks that may play a crucial role in information integration.