Motor imagery training protocol for brain-computer interface-based wearable robotic glove control

Abstract

Research in the field of Brain-Computer Interface (BCI) for rehabilitation and control using wearable robots is actively conducted to support individuals with motor disabilities. This study explores an optimal training protocol for collecting high-quality motor imagery (MI) data to enhance the control performance of MI-BCI-based wearable robot gloves. We collected EEG data from 16 healthy participants engaged in MI tasks across four experimental MI training protocols, including single and repetitive imagery involving both video and actual glove stimulation of wearable robotic gloves movement. Video stimulation with repetitive imagination showed the most significant neural activation in Power Spectral Density(PSD) and Event-Related Desynchronization (ERD) analyses with achieved higher average classification accuracy. Moreover, participants expressed a preference for video stimulation in the survey. These findings emphasize the importance of user-friendly training protocols for enhancing MI-based wearable robot control performance, highlighting the significant role of external stimuli, especially visual signals, in improving MI performance. The study contributes valuable insights to the growing field of MI-BCI-based wearable robotic glove research for broader implementation in real-world applications.