论文部分内容阅读
心电信号(ECG)智能分析非常有利于严重心脏病人的自动诊断。本文介绍了多变量回归模型(MAR)建模法,利用MAR模型从双导联ECG中提取特征对ECG信号进行分类。在分类时,利用MAR模型系数及其K-L变换(K-LMAR系数)作为信号特征,并采用了树状决策过程和二次判别函数(QDF)分类器。利用文中方法对MIT-BIH标准数据库中的正常窦性心律(NSR)、期收缩(APC)、心室早期收缩(PVC)、心室性心动过速(VT)和心室纤维性颤动(VF)各300个样本信号进行了建模和测试。结果表明,为了达到分类目的,MAR模型阶数取4是足够的,基于MAR系数的分类取得了比基于K-LMAR系数的分类稍好的结果。基于MAR系数的分类获得了97.3%~98.6%的分类精度。
Intelligent ECG analysis is very useful for the automatic diagnosis of severe heart disease. In this paper, a multivariable regression model (MAR) modeling method is introduced, and the MAR model is used to classify ECG signals by extracting features from a two-lead ECG. In the classification, the MAR model coefficients and K-L transform (K-LMAR coefficient) were used as signal features, and the tree decision process and QDF classifier were used. The normal sinus rhythm (NSR), phase contraction (APC), early ventricular contraction (PVC), ventricular tachycardia (VT) and ventricular fibrillation (VF) in the MIT- A sample signal was modeled and tested. The results show that for the purpose of classification, it is enough to choose 4 for the MAR model order, and the classification based on the MAR coefficient obtains slightly better results than the classification based on the K-LMAR coefficient. The classification accuracy based on the MAR coefficient obtained the classification accuracy of 97.3% -98.6%.