Font Size: a A A

Predictive Analysis Of Repeated Measures Of Resting-state EEG In Patients With Prolonged Disorders Of Consciousness

Posted on:2022-12-30Degree:MasterType:Thesis
Country:ChinaCandidate:K YuFull Text:PDF
GTID:2504306611478214Subject:Rehabilitation Medicine & Physical Therapy
Abstract/Summary:PDF Full Text Request
Objective The purpose of this study was to investigate the value of repeated measurements of resting state EEG in predicting the outcome of patients with prolonged disorders of consciousness.Methods 22 patients with prolonged disorders of consciousness were enrolled from June 1,2021 to October 31,2021 in Wangjiang Mountain campus of Zhejiang Province People’s Hospital,1 time each week to resting-state EEG acquisition,collect altogether 4 times andcoma recovery scale revised(CRS-R)was evaluated at the same time.The patients were followed up at 3 months and 6 months,respectively,and were assessed by the Glasgow Outcome Scale(GOS),in which recovery of consciousness(GOS ≥Ⅲ)was positive and nonrecovery of consciousness(GOS≤Ⅱ)was negative.The differences in power spectral density(PSD),approximate entropy(APEn),permutation entropy(PeEn),permutation Lempel-Ziv complexity(PLZC),coherence(COH),phase locking value(PLV)and weighted phase lag index(WPLI)in a band and θ band between consciousness recovery group and nonconsciousness recovery group were analyzed.The correlation between positive index and CRS-R and patient outcome was analyzed,and the predictive effect of CRS-R on the outcome of patients with prolonged disorders of consciousness was evaluated by using the receiver operating characteristic curve.Results The change value of CRS-R scores in conscious recovery group was significantly higher than that in unconscious recovery group(P=0.000).EEG data analysis showed that the main effect of PSD time in θ band was significant(P=0.024).The main effect in PeEn group was significant(P=0.042).The main effect of PLZC was significant in P4 channel time(P=0.038),and in T3 channel group(P=0.045).The main effect of COH connecting to T3-T4(P=0.031)and F3-C4(P=0.031)groups in θ band was significant,and the main effect of COH connecting to C3-T3(P=0.043)and T3-T4(P=0.013)groups in a band was significant.The main effect of COH(p3-T4)time in a band was significant(P=0.048).The main effect of PLV(connected T3-T4)group in θ band was significant(P=0.029).The main effect of PLV linking at α band was significant in groups C3-T3(P=0.047)and T3-T4(P=0.022).The main effect of WPLI(P4-C3)group in θ band was significant(P=0.012),and the interaction effect of WPLI(P3-T3)time*group in θ band was significant(P=0.026).The main effect of WPLI(F4-C4)time in a band was significant(P=0.036),the interaction effect of WPLI(F3-P3)time*group in a band was significant(P=0.041),and the interaction effect of WPLI(C3-P4)time*group in a band was significant(P=0.046).The interaction effect of WPLI(F3-T3)time*group in a band was significant(P=0.018).The change value of PSD(T4 channel)in θband within 1 month was positively correlated with GOS(R=0.532,P=0.034).There was a moderate positive correlation between the WPLI in a band and GOS in 1 month(R=0.56,P=0.024).There was a strong positive correlation between WPLI(F3-T3)in a band and GOS in 1 month(R=0.728,P=0.001),and a moderate positive correlation between CRS-R total score(R=0.512,P=0.043)and GOS in 1 month.The area under the curve for predicting pDoC outcome in θ band PSD(T4 channel),α band WPLI(F3-P3)and a band WPLI(F3-T3)in 1 month were 0.817(P=0.039),0.833(P=0.03)and 0.933(P=0.005),respectively.Conclusion The short-term changes of CRS-R scores,EEG power spectrum and connectivity were closely correlated with the outcome of patients with chronic disorder of consciousness.Power spectrum density and weighted phase lag index could be used as effective long-term prognostic tools for patients with chronic disorder of consciousness,with high predictive value.
Keywords/Search Tags:resting-state EEG, prolonged disorders of consciousness, prognosis
PDF Full Text Request
Related items