A Hybrid Brain-Computer Interface based Wearable Exoskeleton System for Fine-Grained Hand Rehabilitation

In the feld of medical rehabilitation, innovative technologies have emerged to assist in restoring motor functions especially for individuals with hand impairments caused by strokes or neuromuscular conditions. Conventional rehabilitation often depends on therapist-led manual therapies, which might not adequately apply brain plasticity or involve patients actively. This study proposed a novel hand rehabilitation system combining a hand exoskeleton, hybrid Brain-Computer Interface (BCI) technology, corticomuscular coherence (CMC) rehabilitation guidance, and brain plasticity training to enhance patient engagement and stimulate neuron activation for effective hand movement control. The proposed system addresses these by integrating EEG and EMG to control the exoskeleton. Experimental results show high offline accuracy (98% for 10 fine-grained gestures, $\mathbf{9 4. 8 \%}$ for the $\mathbf{1 0}$ fine-grained and 4 task-oriented ones) and real-time accuracy (90.5% for 10 fine-grained gestures), indicating the system’s technical feasibility. This breakthrough not only opens up new ways to significantly improve treatment outcomes, but also provide new research directions for the development of human-computer interaction technology.