Identification of Motor Imagery EEG Signal

Author

Xiao, Dan ; Hu, Jianfeng

Author_Institution

Inst. of Inf. & Technol., JiangXi Blue Sky Univ., Nanchang, China

fYear

2010

fDate

23-25 April 2010

Firstpage

1

Lastpage

4

Abstract

To identify subjects by classifying motor imagery EEG signal. Second-order blind identification (SOBI), a blind source separation (BSS) algorithm was applied to preprocess EEG data in for higher signal-to-noise ratio. Subsequently, Fisher distance was used to extract features. Finally, classification of extracted features was performed by back-propagation neural networks. Four types motor imagery EEG of three subjects was classified respectively. The results showed that the average classification accuracy achieved over 80%, and the highest was 88.1% on tongue movement imagery EEG.

Keywords

backpropagation; blind source separation; brain-computer interfaces; electroencephalography; feature extraction; medical signal processing; neural nets; signal classification; Fisher distance; backpropagation neural networks; blind source separation algorithm; brain-computer interface; feature extraction; motor imagery EEG signal identification; second-order blind identification; signal classification; signal-to-noise ratio; Biological neural networks; Biometrics; Blind source separation; Data mining; Electroencephalography; Feature extraction; Foot; Signal processing; Signal to noise ratio; Tongue;

fLanguage

English

Publisher

ieee

Conference_Titel

Biomedical Engineering and Computer Science (ICBECS), 2010 International Conference on

Conference_Location

Wuhan

Print_ISBN

978-1-4244-5315-3

Type

conf

DOI

10.1109/ICBECS.2010.5462405

Filename

5462405