The EEG multi-scale brain network and high-order synergy changes in disorders of consciousness during general...

Brain network alterations during general anesthesia (GA) have been extensively investigated in individuals with intact consciousness. However, network dynamics in patients with disorders of consciousness (DoC) under anesthetic perturbation remain poorly understood. We recorded electroencephalogram (EEG) signals from 33 DoC patients (16 UWS and 17 MCS) during GA across three states (DoC-Resting, DoC-MOSSA, and DoC-Recovery). A multi-scale brain network framework was proposed. Low-order network characteristics were assessed using unbiased minimum spanning tree (MST) and directed phase lag index (dPLI), while high-order organization was evaluated using topological data analysis (TDA) and high-order mutual information (HO-MI). UWS patients exhibited a fragmented, line-like topological backbone with lower leaf fraction (p < 0.05) and tree hierarchy (p < 0.05) than MCS patients. dPLI analysis revealed impaired prefrontal top-down directional communication in UWS patients, whereas MCS patients showed transient disruption and recovery of frontoparietal directional organization during anesthesia. Under anesthesia, TDA revealed reduced H1 maximum Betti number in UWS patients (p < 0.01), indicating impaired topological resilience, while MCS patients showed relatively preserved network organization. Furthermore, HO-MI analysis showed increased total correlation with redundant information processing in UWS patients during anesthesia, whereas MCS patients restored synergistic information integration within the central-parietal network during recovery. GA reveals distinct scale-dependent network alterations between MCS and UWS patients. Persistent impairment of directed communication and high-order synergistic integration characterizes disrupted network organization in UWS. The proposed multi-scale EEG framework integrating MST, dPLI, TDA, and HO-MI provides candidate network biomarkers for perioperative consciousness monitoring and assessment in DoC patients.

The online version contains supplementary material available at https://doi.org/10.1007/s11571-026-10567-4.