Neural Gain Dysregulation in Major Depression: Multidomain EEG Signal Analysis from Spectral Power to Network Topology

Abstract Electroencephalography (EEG) offers a cost-effective window into the neural dynamics of Major Depressive Disorder (MDD), yet most studies focus on isolated domains of analysis. We present a multidomain computational framework spanning spectral, complexity, temporal, and network levels to test the hypothesis that MDD reflects dysregulation of neural gain. Using a publicly available resting-state EEG dataset, we quantified relative band power, nonlinear complexity (Higuchi fractal dimension, multiscale entropy), microstate dynamics, and graph-theoretic topology. MDD patients displayed significantly elevated beta power alongside higher short-scale entropy and fractal dimension values, indicating increased fast-frequency noise. Microstate analysis revealed reduced temporal stability and more frequent transitions, while graph-theoretic measures showed reduced small-worldness, particularly in the theta band, consistent with a shift toward random connectivity. Together, these results suggest that MDD is characterized by amplified fast-frequency activity with degraded signal-to-noise structure across temporal and network scales. This multidomain framework demonstrates how gain dysregulation manifests simultaneously in oscillatory, dynamical, and topological properties of EEG, offering a computational profile of depression that may support biomarker development and treatment monitoring.