EEG Network Predicts The Outcome Of Diagnosis And Prognosis In Disorders Of Consciousness

ObjectivesPrevious studies have shown disorders of consciousness(DOC)appear to be a complex dysfunctional connectivity syndrome.Accurate diagnosis and prognosis prediction based on brain network analysis have tremendous significance in guiding the management of behaviorally unresponsive patients.However,a more precise,hierarchical and dynamic detection profile of consciousness and network connectivity is still lacking.Moreover,the utility of network tests remains to be validated in the clinical setting since the results vary significantly in different researches.Hence,our experiments sought to dynamically unmask subtle neural processing and residual network connectivity in patients with different level of dissolution of consciousness.Moreover,we investigated the spatial topology pattern of the brain network and then examined the potential of spatial pattern of the network(SPN)for diagnosis and prediction of prognosis in patients with DOC.Part Ⅰ:Dynamic EEG network analysis in patients withdisorders of consciousnessMethods42 patients with DOC were included.Herein,taking etiology into account,we assessed the effective network connectivity mechanisms underlying emotional sound stimulation using time-varying electroencephalography(EEG)network analysis.ResultsCompared with minimally conscious state(MCS),vegetative state/unresponsive wakefulness syndrome(VS/UWS)exhibited impaired frontal-parietal networks both for traumatic and nontraumatic groups during the middle-late emotion processing period and surprisingly,traumatic unconsciousness patients suffered from greater top-down connectivity reduction than nontraumatic subjects.We also demonstrated that the brain network properties were significantly correlated with behavioral assessment scores in patients with traumatic injuries(P<0.05),while not in the nontraumatic patients(P>0.05).ConclusionsThe reduction of frontoparietal connectivity may be a potential biomarker for identification of MCS and VS/UWS with same damage cause.In particular,the damage is crucial in traumatic pathogenesis.Part Ⅱ:Spatial Pattern of EEG Functional Networks Predictsthe Outcome of Diagnosis and Prognosis in Disorders ofConsciousnessMethodsThis prospective study included 23 patients with DOC.Clinical behavior was evaluated with the Coma Recovery Scale-Revised(CRS-R)and an ongoing EEG was recorded during a passive emotional paradigm.We compared the SPN assessments to the results of the EEG network properties and the event-related potentials(ERP).Thereafter,we evaluated the outcome at a 3-month follow-up with the CRS-R and Glasgow Outcome Scale(GOS),and assessed the prognostic value of the EEG-SPN in predicting recovery from DOC.ResultsBoth the network properties and ERPs demonstrated unsatisfactory performances for consciousness indexing and outcome prediction.In contrast,the SPN achieved an accuracy of 95%when classifying the patients into MCS and VS/UWS,and when predicting the corresponding 3-month prognosis(i.e.,outcome-positive or outcome-negative).ConclusionsThe SPN can effectively capture the electrical activity differences in the scalp EEG that underlie different consciousness conditions.It,thus,offers a reliable,independent tool to synergize with bedside examinations to discriminate MCS from VS/UWS and to predict the long-term recovery of patients with DOC.