Large-scale information processing during spontaneous brain activity revealed by cross-frequency coupling

Abstract

Amplitude-amplitude (AAC), phase-phase (PPC), and phase-amplitude (PAC) coupling of brain activity are mechanisms that shape the information flow across multiple spatiotemporal scales; however, it is unclear how they are related. We used source-space projected resting-state magnetoencephalography data and empirical mode decomposition to obtain AAC-, PPC-, and PAC-based functional connectivity matrices. We found that specific PAC interactions are highly variable across subjects, but the global topological properties of the network are consistent. PPC and AAC were consistent at both the local and global scales. Additionally, the higher the spatial complexity of PAC is, the stronger its correlation with AAC and PPC will be. Finally, direct and indirect functional connections are differently correlated to the properties of the underlying anatomical scaffold. Together, our results suggest that PPC of high frequencies facilitates the integration of information, AAC of low frequencies facilitates the segregation of information, and PAC is the mechanism binding these two information streams.