基于小波变换和相关分析的心电信号检测

QRS complexes detection based on wavelet transform and correlation analysis

目的 根据信号奇异点 (R波 )与其小波变换模极大值的对应关系 ,以及 QRS波与其模板具有较高相关性 ,将小波变换和相关分析方法应用于心电信号的 QRS波检测 .方法 小波变换是利用二次样条小波对心电信号按 Mallat算法进行变换 .相关分析方法是定义 QRS波的波形模板 ,计算 QRS波与模板的相关系数 .同时还比较了 6个检测指标(小波变换正极大值 ,负极小值 ,QRS波与模板的相关系数 ,QR幅度 ,正极大值 /负极小值点之间的斜率 ,QRS波均值 )的相对变化范围 .结果 在上述 6个指标中 ,前四项检测指标的变化范围较小 ,这为在检测策略中合理选取检测指标及其阈值提供了参考 .经实际心电数据验证 ,QRS波正确检测率达到 99.99% .结论 采用小波变换和相关分析方法可以提高心电信号

AIM To develop wavelet transform and correlation analysis methods for QRS complexes detection. METHODS Quadratic spline wavelet and Mallat algorithm were used for wavelet transform. The correlation coefficient between QRS complexes and QRS template was calculated for correlation analysis. Six indices for detection were compared, namely, wavelet transform of the maximum, wavelet trans form of the minimum, the correlation coefficient, the amplitude from Q wave to R wave, the slope between the points where the maximum and minimum of wavelet transform locate, and the average of QRS complexes. RESULTS The relative variety scale of the first four indices was small. The probability of correct detection reached 99.99%. CONCLUSION The probability of correct detection can be raised by using wavelet transform and correlation analysis methods.