Real time system design of motor imagery brain-computer interface based on multi band CSP and SVM

作者：Li Zhao, Xiaoqin Li, Yan Bian · 发表于：AIP conference proceedings · 年份：2018 · DOI：10.1063/1.5033717 · 被引用次数：4 · 研究领域：EEG and Brain-Computer Interfaces、Gaze Tracking and Assistive Technology、Neuroscience and Neural Engineering

Motion imagery (MT) is an effective method to promote the recovery of limbs in patients after stroke. Though an online MT brain computer interface (BCT) system, which apply MT, can enhance the patient’s participation and accelerate their recovery process. The traditional method deals with the electroencephalogram (EEG) induced by MT by common spatial pattern (CSP), which is used to extract information from a frequency band. Tn order to further improve the classification accuracy of the system, information of two characteristic frequency bands is extracted. The effectiveness of the proposed feature extraction method is verified by off-line analysis of competition data and the analysis of online system.