论文部分内容阅读
在用模拟方法比较了极大似然法和基于Ｆｏｕｒｉｅｒ变换的传统方法的分辨率及与传感器位置误差的敏感性后，采用了Ｆｏｕｒｉｅｒ方法，将零延时波数谱分析应用于人体体表阵列心电信号的分析。本文从零延时波数谱出，发提出新的特征量，能够很好地区分正常人和两类心梗病人，且该特征量与单偶极子的定性解释理论一致。利用该特征和最小距离分类器，对现有４４例数据进行统计识别，得到了８８．６％的正确率。
After comparing the resolution of traditional method based on maximum likelihood method and Fourier transform and the sensitivity to sensor position error by using simulation method, the Fourier method was applied to apply zero-delay wave number spectrum analysis to human body surface array heart Analysis of electrical signals. In this paper, we present a new feature quantity from the zero-delay wave number spectrum, which can distinguish normal people from two types of myocardial infarction patients well, and the characteristic quantity is consistent with the qualitative explanation theory of single dipole. Using this feature and the minimum distance classifier, the existing 44 cases of data were statistically identified, the correct rate of 88.6% was obtained.