Dynamic Resting-State EEG Alpha Connectivity: Quantifying Brain Network State Evolution in Individuals with Psychosis
This study introduces a novel pipeline for investigating brain activity in psychosis using dynamic connectivity from resting-state EEG. Seventy-eight individuals with psychosis and sixty control subjects were analyzed. Source estimation was performed using eLORETA, and connectivity in the alpha band was assessed with wPLI. A modified k-means algorithm was employed to cluster connectivity matrices into distinct brain network states (BNS), from which several metrics were extracted. The segmentation revealed five distinct BNS. The control group exhibited significantly higher connectivity power, while the psychosis group showed greater duration dispersion. In the control group, correlations were found between BNS metrics and cognitive scales. Negative correlations were identified between BNS metrics and the SANS scales in the psychosis group. These results underscore the variability of neural dynamics in individuals with psychosis, as demonstrated by this novel methodology.