TY - GEN
T1 - A Motor Imagery-based Lower Limb Rehabilitation Robot System
AU - Wang, Su
AU - Su, Hao
AU - Huang, Mengjie
AU - Chen, Yuqing
AU - Zheng, Yuting
AU - Yang, Rui
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - Patients with motor impairments resulting from brain injuries or neurological disorders often require specialized rehabilitation to regain mobility and function. The primary challenge is ensuring an immersive yet effective rehabilitation experience. Rehabilitation robotics, designed for assisting rehabilitation training process, are beneficial for performing rehabilitation exercises while providing real-time assistance and guidance. These robotics could achieve even better results by capitalizing on patients' initiative. This work introduces a solution by leveraging Motor Imagery (MI) - a representation of human motor intention - to directly control rehabilitation robotics using brain activities. Specifically, the study showcases the development of a lower limb rehabilitation robot system based on motor imagery, which aims to provide a seamless, immersed rehabilitation experience for patients with motor impairment in lower limb. Three experiments have been conducted in this work: an offline paradigm for data collection and training, a pseudo-online paradigm for quantitative real-time performance analysis, and an online paradigm to evaluate the system's practical performance in real-world scenarios. The offline paradigm achieved an accuracy of 83% for MI detection and 82.15% for classification. The pseudo-online paradigm demonstrated an accuracy of 66.95%, showing its possible effectiveness in real-time situations. In the online paradigm experiment, participants managed to performance rehabilitation training with this system, but issues with delayed or incorrect response were reported. With enhanced datasets and further optimizations, the combination of MI and rehabilitation robotics can potentially revolutionize the outcomes for patients, offering an immersive and effective rehabilitation experience.
AB - Patients with motor impairments resulting from brain injuries or neurological disorders often require specialized rehabilitation to regain mobility and function. The primary challenge is ensuring an immersive yet effective rehabilitation experience. Rehabilitation robotics, designed for assisting rehabilitation training process, are beneficial for performing rehabilitation exercises while providing real-time assistance and guidance. These robotics could achieve even better results by capitalizing on patients' initiative. This work introduces a solution by leveraging Motor Imagery (MI) - a representation of human motor intention - to directly control rehabilitation robotics using brain activities. Specifically, the study showcases the development of a lower limb rehabilitation robot system based on motor imagery, which aims to provide a seamless, immersed rehabilitation experience for patients with motor impairment in lower limb. Three experiments have been conducted in this work: an offline paradigm for data collection and training, a pseudo-online paradigm for quantitative real-time performance analysis, and an online paradigm to evaluate the system's practical performance in real-world scenarios. The offline paradigm achieved an accuracy of 83% for MI detection and 82.15% for classification. The pseudo-online paradigm demonstrated an accuracy of 66.95%, showing its possible effectiveness in real-time situations. In the online paradigm experiment, participants managed to performance rehabilitation training with this system, but issues with delayed or incorrect response were reported. With enhanced datasets and further optimizations, the combination of MI and rehabilitation robotics can potentially revolutionize the outcomes for patients, offering an immersive and effective rehabilitation experience.
KW - brain-computer interface (BCI)
KW - lower limb rehabilitation robot
KW - motor imagery (MI)
KW - rehabilitation training
UR - http://www.scopus.com/inward/record.url?scp=85184661592&partnerID=8YFLogxK
U2 - 10.1109/IBITeC59006.2023.10390931
DO - 10.1109/IBITeC59006.2023.10390931
M3 - Conference Proceeding
AN - SCOPUS:85184661592
T3 - 2023 IEEE International Biomedical Instrumentation and Technology Conference, IBITeC 2023
SP - 80
EP - 85
BT - 2023 IEEE International Biomedical Instrumentation and Technology Conference, IBITeC 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 3rd IEEE International Biomedical Instrumentation and Technology Conference, IBITeC 2023
Y2 - 9 November 2023 through 10 November 2023
ER -