Brain–computer interface (BCI)-guided robot-assisted upper-limb training has been increasingly applied to stroke rehabilitation. However, the induced long-term neuroplasticity modulation still needs to be further characterized. This study investigated the functional reorganization and its structural base after BCI-guided robot-assisted training using resting-state fMRI, task-based fMRI, and diffusion tensor imaging (DTI) data. The clinical improvement and the neurological changes before, immediately after, and six months after 20-session BCI-guided robot hand training were explored in 14 chronic stroke subjects. The structural base of the induced functional reorganization and motor improvement were also investigated using DTI. Repeated measure ANOVA indicated long-term motor improvement was found (F[2, 26] = 6.367, p = 0.006). Significantly modulated functional connectivity (FC) was observed between ipsilesional motor regions (M1 and SMA) and some contralesional areas (SMA, PMd, SPL) in the seed-based analysis. Modulated FC with ipsilesional M1 was significantly correlated with motor function improvement (r = 0.6455, p = 0.0276). Besides, increased interhemispheric FC among the sensorimotor area from resting-state data and increased laterality index from task-based data together indicated the re-balance of the two hemispheres during the recovery. Multiple linear regression models suggested that both motor function improvement and the functional change between ipsilesional M1 and contralesional premotor area were significantly associated with the ipsilesional corticospinal tract integrity. The results in the current study provided solid support for stroke recovery mechanism in terms of interhemispheric interaction and its structural substrates, which could further enhance the understanding of BCI training in stroke rehabilitation. This study was registered at https://clinicaltrials.gov (NCT02323061).

中文翻译：

---

BCI指导的慢性卒中上肢训练后半球功能重组及其结构基础

脑机接口（BCI）指导的机器人辅助上肢训练已越来越多地应用于中风康复。但是，诱导的长期神经可塑性调节仍需进一步表征。这项研究调查了BCI指导的机器人辅助训练后使用静止状态功能磁共振成像，基于任务的功能磁共振成像和扩散张量成像（DTI）数据后的功能重组及其结构基础。在14例慢性卒中患者中，探讨了在20场BCI指导的机器人手部训练之前，之后和之后六个月的临床改善情况和神经系统变化。还使用DTI研究了诱导的功能重组和运动改善的结构基础。重复测量方差分析表明发现了长期的运动改善（F [2，26] = 6.367，p = 0.006）。在基于种子的分析中，在同病运动区（M1和SMA）与一些对病区（SMA，PMd，SPL）之间观察到显着调节的功能连通性（FC）。同侧M1的调制FC与运动功能改善显着相关（r = 0.6455，p = 0.0276）。此外，静止状态数据增加了感觉运动区域的半球间FC，而基于任务的数据则增加了侧向指数，表明恢复期间两个半球的重新平衡。多元线性回归模型表明，运动功能的改善以及同侧M1和对侧运动前区之间的功能变化均与同侧皮质脊髓束完整性密切相关。本研究的结果为脑卒中恢复机制的半球间相互作用及其结构基质提供了坚实的支持，这可以进一步增强对脑卒中康复中BCI训练的理解。该研究在https://clinicaltrials.gov （NCT02323061）。