一种脉搏波和心电信号时域基线漂移消除方法  

A method of eliminating baseline drift of the pulse waves and ECG signals in time domain

作　　者：郭垚垚 陈兆学[1] 

机构地区：[1]上海理工大学医疗器械与食品学院,上海200093

出　　处：《中国医学物理学杂志》2016年第2期167-172,共6页Chinese Journal of Medical Physics

基　　金：上海市教委科研创新项目(13YZ069)

摘　　要：目的:脉搏波和心电信号都是临床诊断的重要依据,但是基线漂移的存在影响了诊断的准确性,因此信号处理中首先要对信号去除基线漂移。方法:检测脉搏波和心电信号的时域特征点,将检测得到的特征点作为插值节点做插值得到时域的基线信号,原信号减去基线信号即得去除基线漂移的信号。为比较不同算法的差异,实验分别采用高斯滤波、中值滤波、形态学滤波和本文算法(插值拟合法)对同一信号进行基线漂移消除实验。结果:实验表明高斯滤波的结果最差,中值滤波和形态学滤波都能有效地去除基线漂移,但是失真却比较严重,而插值拟合法不仅能有效的去除基线漂移,同时也最大程度地保留了原信号的成分。结论:插值拟合法与其他3种方法相比,具有更好的鲁棒性和适应性,对于消除基线漂移有着良好的效果。Objective Both of the pulse wave and ECG signals are important basis for clinical diagnosis, but the existence of baseline drift has affected the diagnostic accuracy. So the baseline drift of signal should be firstly removed in signal processing. Methods The method firstly detects the time-domain feature point of the pulse wave and ECG signal, which will be the interpolation nodes of the interpolation algorithm to get the baseline components of the signal. Subtracting the baseline signal from the original signal, a signal for removing the baseline shift can be obtained. In this paper, the ECG signal baseline drift is removed by Gauss filtering, median filtering, morphological filtering and the algorithm introduced in this paper. Results Experiments have shown that the method based on Gaussian filter has the worst results, and by median filter or morphological filter based method, although the baseline drift can be effectively removed, it has all brought serious distortion, the produced algorithm can not only effectively remove baseline drift but also maximum retain the components of the original signal. Conclusion Our method has good robustness and adaptability with good effect on eliminating baseline drift.

关 键 词：脉搏波 心电 基线漂移 特征点检测 插值 

分 类 号：R318[医药卫生—生物医学工程] TN911.72[医药卫生—基础医学]