Feasibility of a hybrid SSVEP-motor imagery BCI with robotic feedback for upper limb motor rehabilitation in stroke patients.

BACKGROUND

Stroke remains a leading cause of long-term disability, necessitating innovative neurorehabilitation strategies to address persistent motor deficits. Traditional therapies often exhibit limited efficacy due to therapeutic plateau, highlighting the critical need for alternative rehabilitation paradigms.

NEW METHOD: This study assesses the feasibility of a novel hybrid brain-computer interface (BCI) that integrates motor imagery (MI) and steady-state visual evoked potentials (SSVEP), with robotic glove-assisted feedback used to optimize overall system performance. Thirty-two stroke patients were divided into a control group (conventional therapy) and an experimental group (conventional therapy plus BCI intervention with 10- or 20-day cycles).

RESULTS

Outcomes assessed via Fugl-Meyer Assessment (FMA) scores, electroencephalography (EEG) classification accuracy, laterality coefficients (LC), and weighted brain connectivity analysis indicated promising trends. The experimental group showed considerable improvements in FMA scores compared with the control group. The proposed BCI system successfully achieved satisfactory EEG classification accuracy (maximum value of 98.08%) and robust system operation. Furthermore, increases in EEG accuracy, normalization of laterality coefficients (LC), and reinforcement of task-specific weighted brain connectivity were observed, particularly after prolonged training.

COMPARISON WITH EXISTING

METHODS

The proposed hybrid BCI system demonstrates a potential to overcome the limitations of conventional therapies and traditional single-modality BCIs, offering a more engaging and adaptive rehabilitation approach.

CONCLUSIONS

These findings demonstrate the feasibility of the proposed hybrid BCI system and the observed improvements in motor function, neurophysiological markers, and brain connectivity underscore its promise as a novel paradigm to enhance neuroplasticity and recovery outcomes.