标题:Classification of Motor Imagery Electrocorticogram Signals for Brain-Computer Interface
作者:Zheng, Wenfeng; Xu, Fangzhou; Shu, Minglei; Zhang, Yingchun; Yuan, Qi; Lian, Jian; Zheng, Yuanjie
通讯作者:Xu, FZ;Xu, FZ;Xu, FZ
作者机构:[Zheng, Wenfeng] Qilu Univ Technol, Shandong Acad Sci, Sch Elect Engn & Automat, Jinan, Shandong, Peoples R China.; [Xu, Fangzhou] Qilu Univ Technol 更多
会议名称:9th IEEE/EMBS International Conference on Neural Engineering (NER)
会议日期:MAR 20-23, 2019
来源:2019 9TH INTERNATIONAL IEEE/EMBS CONFERENCE ON NEURAL ENGINEERING (NER)
出版年:2019
卷:2019-March
页码:530-533
DOI:10.1109/NER.2019.8716963
摘要:In recent several decades, brain-computer interface (BCI) technology continually yield fruitful results. The electrocorticogram (ECoG) has attracted considerable interest because of its advantages of higher signal-to-noise ratio and greater long-term stability than electroencephalography (EEG) signals. We present an optimal scheme of ECoG signals for motor imagery (MI) classification. The time-frequency features are first extracted by the modified S-transform (MST) algorithm, and then a classifier is trained by using the support vector machine (SVM). In addition, channel selection is performed to reduce the computational complexity of MI-based BCI scheme. This method was tested on BCI Competition III dataset I. The MST coupled with the SVM can obtain the satisfactory classification of 95%. Channel selection can greatly reduce the computational burden of classification and enable this scheme to classify MI tasks in real time.
收录类别:CPCI-S;EI;SCOPUS
资源类型:会议论文;期刊论文
原文链接:https://www.scopus.com/inward/record.uri?eid=2-s2.0-85066763242&doi=10.1109%2fNER.2019.8716963&partnerID=40&md5=d1afde88490a7b779801414283ba8a53
TOP