基于EEG模糊相似性的癫痫发作预测  

Epileptic Seizure Prediction using Fuzzy Similarity Measure on EEG Recordings

作　　者：李小俚[1] 欧阳高翔[1] 关新平[1] 李岩[1] 

机构地区：[1]燕山大学电气工程学院,秦皇岛066004

出　　处：《中国生物医学工程学报》2006年第3期346-350,381,共6页Chinese Journal of Biomedical Engineering

基　　金：国家自然科学基金资助项目(60575012)

摘　　要：本研究提出基于EEG序列模糊相似性指数方法预测癫痫发作。首先,结合复自相关法和Cao法对EEG序列进行了相空间重构;然后,计算相关积分时用Gaussian函数代替Heavyside函数,克服了Heavyside函数的刚性边界问题,使得计算相似性指数更加准确和可靠;最后,分析大鼠癫痫EEG信号,检测癫痫发作前期状态。分析结果表明模糊相似性指数方法能够比动态相似性指数方法获得更长的预测时间和更低的错误预测率。This paper proposes a fuzzy similarity method to predict epileptic seizures with electroencephalography （EEG）. First, muhiple-autocorrelation and Cao＇s method are employed to reconstruct a phase space of EEG recordings. Second, instead of Heavyside function is Gaussian function used in correlation integral for calculating a similarity index, so the crisp boundary of the Heavyside function is eliminated to make the similarity index is more accurate and reliable. Finally, the fuzzy similarity index is applied to indicate the preictal state of nine rats with EEG signals. The result shows that the fuzzy similarity index is better than dynamical similarity index in increasing anticipation time and decreasing false prediction rate for the prediction of epileptic seizure.

关 键 词：EEG信号 模糊相似性指数 癫痫发作 预测 相空间重构 

分 类 号：R318.03[医药卫生—生物医学工程]