DocumentCode :
3017143
Title :
E-Health Design of EEG Signal Classification for Epilepsy Diagnosis
Author :
Chi-Chou Kao
Author_Institution :
Dept. of Comput. Sci. & Inf. Eng., Nat. Univ. of Tainan, Tainan, Taiwan
fYear :
2013
fDate :
2-5 July 2013
Firstpage :
67
Lastpage :
71
Abstract :
Epilepsy, which is caused by abnormal discharges in the brain, is one of the most common neurological disorders. To diagnose efficiently epilepsy, it is valuable to classify electroencephalogram signal. In this paper, we proposed a new e-health design of Electroencephalogram (EEG) signal classification for epilepsy diagnosis. The design is based on support vector machine to classify electroencephalogram signal. We first decompose electroencephalogram signal into bands by using discrete wavelet transform and compute the approximate entropy values in the bands. Next, by proposed feature selection method, the feature vectors are selected adaptively from statistical wavelet coefficients and approximate entropy values. Finally, the support vector machine is used to classify the selected features. The experimental results showed the proposed system has great performance and reliability and the total accuracy of classification can achieve 98%.
Keywords :
approximation theory; discrete wavelet transforms; electroencephalography; entropy; medical signal processing; neurophysiology; signal classification; statistical analysis; support vector machines; EEG signal classification; approximate entropy value; brain; discrete wavelet transform; e-health design; electroencephalogram signal classification; electroencephalogram signal decomposition; epilepsy diagnosis; feature selection method; feature vector; neurological disorder; statistical wavelet coefficients; support vector machine; Discrete wavelet transforms; Electroencephalography; Entropy; Epilepsy; Feature extraction; Support vector machines; Tumors; e-health; electroencephalogram signal; epilepsy; support vector machine;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Biometrics and Security Technologies (ISBAST), 2013 International Symposium on
Conference_Location :
Chengdu
Print_ISBN :
978-0-7695-5010-7
Type :
conf
DOI :
10.1109/ISBAST.2013.13
Filename :
6597668
Link To Document :
بازگشت