Estimating Criticality of Resting-State Phase Synchronization Network Based on EEG Source Signals

EEG phase synchrony is an important signature in estimating functional connectivity of brain network, in which criticality of phase-locking state has been viewed as the key factor in facilitating dynamic reorganization of functional network. Based on source trace of resting-state EEG signals recorded from 24 subjects, this study extracted phase-locking intervals (PLIs) between pairwise source signals and constructed PLI sets with size higher than \( 10^{5} \) for each subject, from frontal-parietal, frontal-temporal, and temporal-parietal cortical areas, respectively. Through further data fitting in power-law model, this study finds that θ- (4–8 Hz) and α-band (8–13 Hz) activities have longer phase-locking duration in a broader power-law distribution interval, compared to those in high frequency bands, indicating higher temporal stability of functional coupling between brain areas. In contrast, the probability density of PLIs oscillating in β (13–30 Hz) and γ (30–60 Hz) bands has less data fitting errors and bigger power-law exponent, suggesting higher criticality and flexibility of reorganization of phase synchronization networks. The findings are expected to provide effective neural signatures for comparison and recognition of neural correlations of cognition, emotion, disease etc. in the future.