Automatic Detection of Epileptic Sharp-slow by Wavelet and Approximate Entropy
The automatic epileptiform activities detection in EEG is significant in clinical application. Epileptic sharp-slow complex wave is one of typical presence of epileptiform activities, which has different time-frequency property compared with spike and spike-slow complex wave. A new scheme is presented for detecting epileptic sharp-slow wave in 8-channel EEG data from normal subjects and epileptic patients. The scheme is based on the characteristic of a multi-resolution and approximate entropy (ApEn) analysis of EEG signals. The EEG signals on each channel are decomposed into three levels using multi-resolution wavelet analysis, and then ApEn values of the detail coefficients are computed. Distinct differences are found between the ApEn values of the epileptic sharp-slow and the normal EEG. The EEG signals are detected by Neyman-Pearson (NP) criteria. The optimal detection rule of detecting sharp-slow is achieved, and it assures a higher detection rate with a lower false detection rate.
Chunmei Wang Junzhong Zou Jian Zhang Zhisuo Zhang
Department of Automation,East China University of Science and Technology,Shanghai 200237,China Depar Department of Automation,East China University of Science and Technology,Shanghai 200237,China Changhai Hospital Attached to the Second Military Medical University,Shanghai 200433,China
国际会议
2009 IEEE International Conference on Information and Automation(2009年 IEEE信息与自动化国际学术会议)
珠海、澳门
英文
1269-1273
2009-06-22（万方平台首次上网日期，不代表论文的发表时间）