Nested spectral analysis of brain functional networks in depression based on inner composition alignment

To address the limitations of traditional phase-locked value methods in electroencephalography (EEG) functional connectivity analysis, such as loss of frequency-domain information and insufficient capture of nonlinear dynamics, this study proposes a time-domain permutation-based inner composition alignment network nested-spectral analysis framework for analyzing depression-related brain functional networks. The results demonstrate that, regardless of depressive status, EEG networks in the theta and alpha bands predominantly operate in an integrated state globally. However, with increased depressive symptoms, the segregated state becomes more dominant. This shift leads to higher state transition frequencies in the alpha band for depression patients, while depressive conditions are also associated with reduced dynamic switching in the beta and gamma bands. We identified a “low-frequency integration decline and high-frequency segregation dysregulation” dual-modal imbalance phenomenon in the brain functional networks of depression patients, providing a novel paradigm to support the clinical diagnosis of depression.