EEG-Based Exoskeleton for Rehabilitation Therapy

Abstract: Functional restoration of arm and hand movements is a challenging goal of post-stroke/trauma rehabilitation therapy. The post-stroke/trauma conditions often may lead to partial or total loss of motor function. Therefore, the use of assistive brain-controlled robotic exoskeleton has recently gained a lot of interest from the bioengineering research community. In this study, electroencephalography (EEG) has been used for controlling the exoskeleton and for providing the brain–machine interface (BMI) roadmap. The brain signalswere recorded corresponding to different imagined movement by normal subjects using 10–20 standard electrode placement system. A sample of the EEG recorded at a frequency of at least 160 Hz was preprocessed to remove the line frequency interference and artifacts. Features were then extracted from the EEG signal and processed to actuate the exoskeleton, thus ultimately assisting the subject in his/her rehabilitation program. A linear discriminant analysis (LDA)-based classifier is used to map the extracted features to a specific task. This study has achieved the best accuracy of 97.101% using linear classifiers and 72.133% using quadratic classifiers. This paper presents system design and development along with an experimental evaluation of EEG-driven exoskeleton. This exoskeleton could then be used to assist in the rehabilitation program of stroke/trauma patients.

Download paper here