Coherence in resting-state EEG as a predictor for the recovery from unresponsive wakefulness syndrome.

We investigated differences of EEG coherence within (short-range), and between (long-range) specified brain areas as diagnostic markers for different states in disorders of consciousness (DOC), and their predictive value for recovery from unresponsive wakefulness syndrome (UWS). EEGs of 73 patients and 24 controls were recorded and coma recovery scale- revised (CRS-R) scores were assessed. CRS-R of UWS patients was collected after 12 months and divided into two groups (improved/unimproved). Frontal, parietal, fronto-parietal, fronto-temporal, and fronto-occipital coherence was computed, as well as EEG power over frontal, parietal, occipital, and temporal areas. Minimally conscious patients (MCS) and UWS patients could not be differentiated based on their coherence patterns or on EEG power. Fronto-parietal and parietal coherence could positively predict improvement of UWS patients, i.e. recovery from UWS to MCS. Parietal coherence was significantly higher in delta and theta frequencies in the improved group, as well as the coherence between frontal and parietal regions in delta, theta, alpha, and beta frequencies. High parietal delta and theta, and high fronto-parietal theta and alpha coherence appear to provide strong early evidence for recovery from UWS with high predictive sensitivity and specificity. Short and long-range coherence can have a diagnostic value in the prognosis of recovery from UWS.