The Hindu – Brain-computer interface enhanced for active rehabilitation of stroke patients

Researchers at IIT Palakkad developed an enhanced motor imagery-based Brain–Computer Interface aimed at supporting active rehabilitation for stroke patients. The approach uses a trained artificial neural network to identify the most informative EEG channels for classifying imagined hand movements, reducing the number of channels required while maintaining effective decoding performance.

Led by Prof. Vinod A. Prasad, the study reported an average motor imagery classification accuracy of 74.3% across 10 participants using 13 selected EEG channels, reduced from the original 32 channels. The work highlights the potential of motor imagery-based BCI to move stroke rehabilitation from passive therapy towards active rehabilitation, where a patient’s movement intention can be decoded and used to support recovery through brain plasticity.