運用腦電波研究中風病人的復健成效 與持續情形

Using EEG to evaluate the stroke rehabilitation efficacy : a longitudinal study

呂億綸(Yi-Lun Lu)

指導教授 ： 陳純娟

國立中央大學/工學院/生物醫學工程研究所/碩士(2015年)

摘要

虛擬實境復健近來受到許多的注目，但是其復健功效卻是眾說紛紜，本研究的目的在於運用EEG訊號搭配定性定量的動態因果模型來研究中風病人在復健之後，復健的成效以及復健後一個月復健成效是否持續，並比較兩種復健的方式-虛擬實境以及傳統復健的差異。
本研究為縱貫性研究（longitudinal study），採用EEG訊號搭配定性定量的動態因果模型來研究大腦內運動神經網路因應不同復健方式改變的情形與其效果持續的狀況。本研究招募中風病人30人（女性8人，男性22），隨機的情狀下分成study 以及control 組，其中16人做虛擬實境（Study），另14人做傳統復健（Control），皆為24小時復健療程。我們收集復健前後和追蹤一個月後的EEG與臨床量表復原成效。研究結果顯示，以傳統臨床量表來看復健的效果，虛擬實境與傳統復健皆有顯著的復健療效，且都可以持續至少復健完一個月，驗證虛擬實境復健確有療效。 EEG與動態因果模型的分析結果發現兩組病患在後測時皆展現出跟健康正常人類似的大腦連結狀況，但是到了追蹤時，卻趨向於同側優勢，其結果指向病人受傷的大腦為了要能與年輕正常大腦相同程度的運動，必須活化更多的大腦皮質來參與工作達到目的，此結果研究在兩組病患間並無差異，顯示定性上來說，虛擬實境復健所誘發之神經網路重組與傳統無異。但是如果以大腦連結的情形來探討復健方法所造成的差異時，我們發現傳統復健會使大腦神經功能趨向於健側，以健側來輔助患側達到中風前的功能，而虛擬實境復健則會刺激兩側的腦區；此外，在追蹤一個月後，大腦會有整體連結強度減弱的情況，顯示虛擬實境復健對於健側輔助患側的機制減少。我們研究的結果顯示，運用客觀定量的腦波與神經網路改變分析，有較高的敏感度，可以更進一步探討大腦網路改變的機制，增進我們對大腦復原歷程的了解。

關鍵字

動態因果模型 ； 中風 ； 復健方式 ； 虛擬實境復健系統

並列摘要

Virtual reality based rehabilitation has drawn a lot of attention recently but the efficacy of it is still under debate. In this study, we set up to test if there is any difference of rehabilitation efficacy between convention and VR based intervention. Specifically, we examine whether the efficacy of the two intervention changes with time by using EEG and dynamic causal modelling for induced response (DCM_IR).
This is a prospective study. 30 subjects (8 females, 22males) were recruited and divide randomly into either VR (study) or conventional (control) training group, resulting in 16 and 14 in the study and control group, respectively. All subjects underwent a total 24 hours training program with the frequency of 1 hour a day, five days a week. 30 channel EEG were measured three times at pre-(before rehabilitation) , post- ( after rehabilitation ) intervention and one month after as the follow-up when patients performed either the shoulder or elbow flexion-extension using their affected hand. Clinical measures included FMA, TEMPA, WOLF were also conducted followed EEG measurement. The EEG data were pre-processed and then entered DCM_IR for network identification. Six plausible models, comprising bilateral primary motor cortex (M1), premotor cortex(PM) and (supplementary motor area) SMA, were tested and Bayesian model selection (BMS) was used to selected the model.
There has a significant effect on the efficacy of both groups, and this treatement effect can last at least one month, indicating the effectiveness of VR based intervention. The Bayesian Model Selection (BMS) identified the model with ipsilesional M1 dominating the network structure for both groups after rehab, suggesting that the strategy used by the brain for functional restoration is identical: to be as normal as before, irrelevant to the means of intervention. However, the best model switched into a model with more contralesional M1 engaged in the network structure for both groups at follow-up, indicating the compensation mechanism occurred after stopping rehab training. Qualitatively, there has no difference on the best model between two groups. In terms of the alternations of motor network after rehabilitation, the contralesional hamisphere engaged more for moving the paretic hand in the control group while, in the study, the activation patterns were more bilateral. Nevertheless, the overall coupling strength decrease in both groups at follow-up. Our result puts forward our understanding of the recovery process after stroke.

並列關鍵字

dynamic casual modeling ； stroke ； rehabilitation ； virtual reality system

參考文獻

1. Dennis, M.S., et al., Long-term survival after first-ever stroke: the Oxfordshire Community Stroke Project. Stroke, 1993. 24: p. 796-796.

2. Ferrucci, L., et al., Recovery of functional status after stroke. A postrehabilitation follow-up study. Stroke, 1993. 24（2): p. 200-205.

3. Wilkinson, P.R., et al., A long-term follow-up of stroke patients. Stroke, 1997. 28（3): p. 507-512.

4. Wade, D.T., et al., Physiotherapy intervention late after stroke and mobility. Bmj, 1992. 304（6827): p. 609-613.

5. Turolla, A., et al., Virtual reality for the rehabilitation of the upper limb motor function after stroke: a prospective controlled trial. J Neuroeng Rehabil, 2013. 10: p. 85.