General anesthesia with propofol reconfigures the brain's functional network hub structure and reverses the phase relationship between frontal and parietal regions. Using graph theoretical analysis of 21-channel electroencephalogram data from 10 volunteers, the study found that network topology—not connection strength—correlates with states of consciousness. After propofol administration, average path length, clustering coefficient, and modularity increased, long-range connections were disrupted, and hub node strength decreased. The primary hub shifted from parietal to frontal regions. The phase lead of frontal to parietal areas in the alpha frequency band (8-13 Hz) during wakefulness reversed direction after propofol and returned during recovery. Changes in network topology may be the primary mechanism for loss of frontal to parietal feedback during anesthesia.
Large-scale brain networks function near a critical state, and partial phase locking—a network science concept—shapes the characteristic functional connectivity and asymmetric anterior-posterior phase lag entropy (PLE) topography at this critical state, with low PLE for high-degree nodes and high PLE for low-degree nodes. Comparing PLE topography at criticality with electroencephalogram data from baseline consciousness, isoflurane anesthesia, ketamine anesthesia, vegetative state/unresponsive wakefulness syndrome, and minimally conscious state shows that topographical similarity and strength of PLE differentiate these pharmacologic and pathologic states of consciousness. This provides a theory-based metric to quantify how far a perturbed brain network deviates from criticality, rather than merely determining critical or non-critical state.
Permutation entropy (PE) and symbolic mutual information (SMI) analysis of fMRI data from 15 healthy participants under propofol sedation reveals that anesthesia differentially affects local and shared brain information. Global PE decreased from wakefulness to deep (unconscious) sedation and increased upon recovery, with greater reduction in subcortical than cortical networks. SMI's spatial pattern (topographic structure) reconfigures during unconsciousness, and the positive correlation between PE and SMI seen in conscious states is disrupted in deep sedation. PE changes preferentially occur in highly connected hub regions. These results suggest that altered local and shared information exchange is a mechanistic indicator of anesthetic-induced unconsciousness.