Brain connectivity analysis from EEG signals using stable phase-synchronized states during face perception tasks

• Quasi-stable phase synchronized patterns are observed in ββ band of multichannel EEG.
• Synchrostates characterize the temporal brain dynamics found during face perception.
• Inter-synchrostate transitions are semi-deterministic and depends on the stimulus.
• Phase synchronization index for EEG synchrostates yields functional brain networks.
• Complex network measures can be used to understand information processing in brain.

Degree of phase synchronization between different Electroencephalogram (EEG) channels is known to be the manifestation of the underlying mechanism of information coupling between different brain regions. In this paper, we apply a continuous wavelet transform (CWT) based analysis technique on EEG data, captured during face perception tasks, to explore the temporal evolution of phase synchronization, from the onset of a stimulus. Our explorations show that there exists a small set (typically 3–5) of unique synchronized patterns or synchrostates  , each of which are stable of the order of milliseconds. Particularly, in the beta (ββ) band, which has been reported to be associated with visual processing task, the number of such stable states has been found to be three consistently. During processing of the stimulus, the switching between these states occurs abruptly but the switching characteristic follows a well-behaved and repeatable sequence. This is observed in a single subject analysis as well as a multiple-subject group-analysis in adults during face perception. We also show that although these patterns remain topographically similar for the general category of face perception task, the sequence of their occurrence and their temporal stability varies markedly between different face perception scenarios (stimuli) indicating toward different dynamical characteristics for information processing, which is stimulus-specific in nature. Subsequently, we translated these stable states into brain complex networks and derived informative network measures for characterizing the degree of segregated processing and information integration in those synchrostates, leading to a new methodology for characterizing information processing in human brain. The proposed methodology of modeling the functional brain connectivity through the synchrostates may be viewed as a new way of quantitative characterization of the cognitive ability of the subject, stimuli and information integration/segregation capability.