基于Hurst指数和SVM的癫痫脑电检测方法

Approach of Epileptic EEG Detection Based on Hurst Exponent and SVM

癫痫脑电波的自动检测对于患者诊断和减轻医生工作强度都具有重要的意义。提出一种基于Hurst指数和SVM的癫痫脑电检测算法。首先提取脑电信号的Hurst指数,然后对脑电进行3 Hz～8.5 Hz、8.5 Hz～16.5 Hz、16.5 Hz～29 Hz带通滤波并分别计算波幅的相对均值,最后使用SVM分类器实现癫痫脑电波的自动检测。对临床脑电信号的实验表明,该方法具有较强的检测性能和良好的实时性,准确率达到98.75%。所提出的Hurst指数和波幅相对均值作为特征,采用SVM的分类方法能有效实现癫痫脑电的检测,值得更深入的研究。

Automatic seizure detection using electroencephalogram（EEG） can effectively assist medical diagnosis and alleviate work intensity of hospital doctors.This paper proposes a new seizure detection approach based on Hurst exponent and support vector machine（SVM）.We extract Hurst exponent of EEG as the feature of epileptic waves first.Then EEG is filtered with 3 Hz ～ 8.5 Hz,8.5 Hz ～ 16.5 Hz and 16.5 Hz ～ 29 Hz band-pass filters and the relative average amplitudes within different frequency bands are calculated.In the end,we apply a SVM classifier to detect seizure automatically.The experimental results reveal that the proposed algorithm is relatively accurate and efficient and the recognition rate is up to 98.75%.