计算机技术与发展2016，Vol.26Issue(6)：133-137,5.DOI:10.3969/j.issn.1673-629X.2016.06.029
融合互信息和支持向量机的癫痫自动检测算法
An Automatic Detection Algorithm for Epilepsy EEG Based on MI and SVM
摘要
Abstract
Electroencephalogram ( EEG) is the most important tools for seizure detection by recording spontaneous and rhythmic electrical activity of brain cells through electrodes. A new method for feature extraction and classification is proposed based upon Mutual Informa-tion ( MI) and Support Vector Machine ( SVM) ,which can distinguish epilepsy EEG from normal EEG quickly and efficiently. Then the comparison on the classification results is made using various combinations of feature vector in the same dimension and in the different di-mension. In addition,the classification results and efficiency are compared between proposed algorithm and other common algorithm. The experiment shows that the two-dimensional feature vectors combining mean and variance extracted from MI sequence of epilepsy EEG and normal EEG,has advantages of simple operation and high classification result,and this algorithm is also faster than others,which is useful for clinical seizure detection in real time.关键词
互信息/支持向量机/脑电信号/特征提取/癫痫自动检测Key words
MI/SVM/EEG/feature extraction/automatic epilepsy detection分类
信息技术与安全科学引用本文复制引用
沈洋洋,黄丽亚,郭迪,笪铖璐,陈志阳,戴加飞..融合互信息和支持向量机的癫痫自动检测算法[J].计算机技术与发展,2016,26(6):133-137,5.基金项目
国家自然科学基金资助项目(61003237) （61003237）
江苏省高校自然科学研究项目(10KJB510018) （10KJB510018）
