Abstract: Objective  To evaluate the rehabilitation effects of brain-computer interface (BCI)-assisted rehabilitation training based on motor imagery (MI) and steady-state visual evoked potential (SSVEP) on upper limb motor function and activities of daily living in stroke patients.
Methods  This study was a prospective, single-center, randomized controlled trial, which enrolled stroke patients with hemiplegia within 2 weeks to 12 months after onset. These patients were randomly divided into an experimental group and a control group. Both groups received conventional rehabilitation training, including physical therapy, occupational therapy with each training lasting 30 minutes, 5 days per week, for 2 consecutive weeks, totaling 10 sessions. On the basis of conventional rehabilitation training, the experimental group additionally received 10 sessions of BCI-assisted rehabilitation training (each session lasting 45 minutes, 5 days per week, for 2 consecutive weeks, totaling 10 sessions). A dual-modal closed-loop system based on MI and SSVEP was used to control an exoskeleton manipulator for performing grasping tasks. The primary outcome measures were the Fugl-Meyer motor assessment-upper extremity (FMA-UE) to evaluate upper limb motor function and the modified Barthel index (mBI) to evaluate activities of daily living. The secondary outcome measures included the modified Ashworth scale (MAS) for evaluating muscle tone, the Hamilton anxiety scale (HAMA) for evaluating anxiety status, and the Hamilton depression scale (HAMD) for evaluating depressive status. All patients completed the assessments before intervention, after intervention, and 2 weeks after the end of intervention (follow-up period).
Results  A total of 36 stroke patients with hemiplegia were enrolled, with 18 patients in the experimental group and 18 in the control group. There were no statistically significant differences in baseline data and scale assessment results between the two groups. Analysis of primary outcome measures showed that both FMA-UE scores and mBI scores of the two groups had significant time effects (FMA-UE scores: F=58.519, P<0.001; mBI scores: F=129.935, P<0.001) and time×group interaction effects (FMA-UE scores: F=19.551, P<0.001; mBI scores: F=15.661, P<0.001). The scores of both groups showed an increasing trend with the extension of time. Specifically, the score improvements of the experimental group were better than those of the control group after intervention (FMA-UE scores: P=0.049; mBI scores: P=0.035) and during the follow-up period (FMA-UE scores: P=0.005; mBI scores: P=0.002). Analysis of secondary outcome measures showed that after intervention, the HAMA scores and HAMD scores of both groups improved, and the improvement trend in the experimental group was more obvious. However, there were no statistically significant differences in HAMA scores, HAMD scores, and MAS grades between the two groups. The BCI feedback accuracy of the experimental group increased from (65.97±14.70)% before intervention to (76.34±12.16)% after intervention, and the difference was statistically significant (P=0.008).
Conclusions  For stroke patients with hemiplegia, MI-SSVEP-based BCI-assisted rehabilitation training combined with conventional rehabilitation training can effectively improve their upper limb motor function and activities of daily living. 

Key words: Stroke; Brain-computer interface; Upper limb motor function; Rehabilitation

摘要： 目的 评估基于运动想象（motor imagery，MI）和稳态视觉诱发电位（steady-state visual evoked potential，SSVEP）的脑机接口（brain-computer interface，BCI）辅助康复训练对卒中患者上肢运动功能及日常生活活动能力的康复效果。
方法 本研究为前瞻性单中心随机对照试验，纳入了发病2周～12个月的卒中偏瘫患者，并将其随机分为试验组和对照组。两组均接受常规康复训练，包括物理治疗和作业治疗，每个项目每次训练30 min，每周训练5 d，持续2周，总训练次数为10次。试验组在常规康复训练的基础上，接受10次BCI辅助康复训练（每次训练45 min，每周训练5 d，持续2周，总训练次数为10次），通过MI和SSVEP双模态闭环系统控制外骨骼机械手执行抓握任务。主要结局指标：采用Fugl-Meyer运动功能评定量表上肢部分（Fugl-Meyer motor assessment-upper extremity，FMA-UE）评估上肢运动功能；采用改良巴塞尔指数（modified Barthel index，mBI）评估日常生活活动能力。次要结局指标：采用改良阿什沃思量表（modified Ashworth scale，MAS）评估肌张力；采用汉密尔顿焦虑量表（Hamilton anxiety scale，HAMA）评估焦虑状态；采用汉密尔顿抑郁量表（Hamilton depression scale，HAMD）评估抑郁状态。所有患者分别于干预前、干预后及干预结束2周后（随访期）完成上述指标评估。
结果 共纳入36例卒中偏瘫患者，试验组和对照组各18例。两组的基线资料及量表评估结果的差异均无统计学意义。主要结局指标分析显示，两组患者的FMA-UE评分和mBI评分均存在显著的时间效应（FMA-UE评分：F=58.519，P<0.001；mBI评分：F=129.935，P<0.001）和时间×组别交互效应（FMA-UE评分：F=19.551，P<0.001；mBI评分：F=15.661，P<0.001）；两组评分均随时间延长呈升高趋势，且试验组干预后（FMA-UE评分：P=0.049；mBI评分：P=0.035）和随访期（FMA-UE评分：P=0.005；mBI评分：P=0.002）的评分改善均优于对照组。次要结局指标分析显示，两组干预后HAMA评分和HAMD评分均有所改善，且试验组的改善趋势更明显，但两组在HAMA、HAMD评分及MAS分级上的差异均无统计学意义。另外，试验组的BCI反馈准确率由干预前的（65.97±14.70）%提升至干预后的（76.34±12.16）%，差异具有统计学意义（P=0.008）。
结论 对于卒中偏瘫患者，基于MI-SSVEP的BCI辅助康复训练联合常规康复训练可有效改善其上肢运动功能和日常生活活动能力。

关键词: 卒中; 脑机接口; 上肢运动功能; 康复