标题:
基于函数系数非参数统计模型的心电R波检测R Wave Detection in ECG with Functional Coefficient Nonparametric Statistical Model
作者:
孙敏, 苏理云, 李晨龙
关键字:
心电信号, R波, 函数系数非参数统计模型, 自适应Electrocardiogram (ECG) Signal; R Wave; Functional Coefficient Nonparametric Statistical Model; Adaptive Method
期刊名称:
《Statistics and Application》, Vol.2 No.4, 2013-12-27
摘要:
本文提出了一种基于函数系数非参数统计模型的R波检测方法。在用数字带通滤波器对心电信号进行滤波和去噪以后，用函数系数非参数回归模型来拟合心电时域数据，通过差分选取适当的阈值，并且所选取的阈值具有自适应学习的特性，可根据不同需求调整阈值的大小，然后利用模型估计出的一阶导数位置检测出R波。和差分阈值的方法相比较，该方法在检测R波的过程中对数据本身就有一个很好的平整作用。最后还分别讨论了R波的多检和漏检的情况，并给出了相应的纠正方法以使得检测更加精确。我们利用临床试验数据对该方法进行验证，实验结果表明，该方法能够准确高效地检测出心电信号中的R波。A novel R wave detection method is presented based on functional coefficient nonparametric statistical model. Firstly, digital bandpass filter is used for denoising and filtering of Electrocardiogram (ECG) signal. Then ECG signal is fitted by using functional coefficient nonparametric statistical model and difference threshold. Finally, the derivative of fitting model is applied to detect R wave in ECG with signal processing and threshold. Compared to the old difference method for R wave detector, the nonparametric statistical regression approach is more accurate based on the derivative of each point. Clinical data are used to test the performances. The experimental results show that the proposed method for R wave detection is effective.