用于心电波形分析的相似性度量策略

Similarity Measurement Strategie for ECG Waveform Analysis

为表现2个形态相似的心电波形在临界点上的差别程度,根据心电波形的自身特点,提出一种基于Hausdorff距离的分段加权策略。设计心电波形的分类算法,利用包容度描述不同度量方法在相似波形临界点的状况。实验数据显示,该策略的平均分类敏感度大于90%。结果表明,与其他度量方法相比,该策略的分类准确率高、包容度强。

In order to perform differences between two similar Electrocardiogram（ECG） waves on partition points,a method based on Haussdorff distance is proposed according to wave characteristics.This method differentiates the critical point among different kinds of ECG waves.A description method named inclusive degree is proposed based on repeated experiments.This method is used to describe situations about similar waves in the vicinity of partition points among different measurement methods.Experimental results show that the average classification sensitivity of this measurement method is above 90%.Comparing with other measurement methods,the result shows that the method has the better discrimination.