Title :
Motor Imagery BCI Research Based on Sample Entropy and SVM
Author :
Lei Wang ; Guizhi Xu ; Shuo Yang ; Miaomiao Guo ; Weili Yan ; Jiang Wang
Author_Institution :
Province-Minist. Joint Key Lab. of Electromagn. Field & Electr. Apparatus Reliability, Hebei Univ. of Technol., Tianjin, China
Abstract :
Brain Computer Interface (BCI) is a system that provides an artificial communication between the human brain and the external world. It may give disabled people direct control over a neuro-prosthesis by their intentions that are reflected in their brain signals. In this paper, brain electric field data (EEG) was recorded though 28 electrodes placed on the scalp. According to the fact that EEG is non-stationary and non-linear; a non-linear dynamic method called Sample Entropy (SampEn) was applied to extract the features of EEG. A Support Vector Machine (SVM) classifier was structured for pattern classification. The final results show that SampEn is an effective method to extract the feature of different brain states.
Keywords :
brain-computer interfaces; electroencephalography; feature extraction; medical signal processing; prosthetics; support vector machines; EEG; SVM; SVM classifier; SampEn; artificial communication; brain computer interface; brain electric field data; feature extraction; motor imagery BCI research; neuro-prosthesis; pattern classification; sample entropy; support vector machine classifier; Accuracy; Electroencephalography; Entropy; Feature extraction; Support vector machines; Time series analysis; Vectors;
Conference_Titel :
Electromagnetic Field Problems and Applications (ICEF), 2012 Sixth International Conference on
Conference_Location :
Dalian, Liaoning
Print_ISBN :
978-1-4673-1333-9
DOI :
10.1109/ICEF.2012.6310370