典型失神癫痫的脑电定量分析及分类研究

发布时间：2018-11-23 16:25

【摘要】：目的采用定量脑电图的方法对失神发作期脑电和发作间期脑电进行分析,提出一种基于去趋势波动分析(detrended fluctuation analysis,DFA)、能量指数和支持向量机(support vector machine,SVM)的癫痫分类方法。方法对23例典型失神癫痫患儿发作期脑电及发作间期脑电进行定量分析,首先采用DFA方法计算脑电的标度指数α,然后对脑电进行小波分解,对δ(0.5~3.5Hz)、θ(4~7.5Hz)、α(8~13Hz)、β(14~30Hz)频段的脑电进行重构,计算各频段脑电能量指数,最后使用SVM分类器实现癫痫脑电的自动分类。结果对儿童失神发作期脑电与发作间期脑电的分类准确率达到96.9%,实现了儿童失神发作期脑电的自动识别,其准确率达到临床检验标准。结论该方法具有较强自动识别能力,将来可用于临床失神发作期脑电图的定量分析及自动识别,具有一定的临床应用前景。
[Abstract]:Objective to analyze EEG during aphasia and interictal EEG by using quantitative electroencephalogram (QEEG), and to propose a new method based on (detrended fluctuation analysis,DFA), energy index and support vector machine (support vector machine,). SVM). Methods the EEG and interictal EEG of 23 children with typical aphasia were quantitatively analyzed. The scalar index 伪 of EEG was calculated by DFA method, then the EEG was decomposed by wavelet, and 未 (0.5~3.5Hz) was analyzed. The EEG of 胃 (4~7.5Hz), 伪 (8~13Hz), 尾 (14~30Hz) were reconstructed, and the energy index of EEG was calculated. Finally, the SVM classifier was used to realize the automatic classification of epileptic EEG. Results the classification accuracy of EEG in children with aphasia and interictal EEG was up to 96.9. the automatic recognition of EEG in children with aphasia was realized, and the accuracy rate was up to the standard of clinical test. Conclusion this method has strong automatic recognition ability and can be used in quantitative analysis and automatic recognition of electroencephalogram in the period of clinical aphasia in the future. It has a certain clinical application prospect.
【作者单位】：武汉市妇女儿童医疗保健中心神经电生理室;武汉市妇女儿童医疗保健中心神经内科;
【基金】：武汉市卫生局科研基金资助项目(No.WX09C20) 武汉市科技创新平台-儿童神经疾病临床医学研究中心资助项目(武科计2014-160号) 武汉黄鹤英才医疗卫生计划专项经费资助项目(武人才办2014-10号)
【分类号】：R742.1;R741.044

【参考文献】