EEMD在同时消除脉搏血氧检测中脉搏波信号高频噪声和基线漂移中的应用  

Using EEMD to Eliminate High Frequency Noise and Baseline Drift in Pluse Blood-oximetry Measurement Simultaneously

作　　者：韩庆阳[1] 王晓东[1] 李丙玉[1] 周鹏骥 

机构地区：[1]中国科学院长春光学精密机械与物理研究所光电技术研发中心,长春130033

出　　处：《电子与信息学报》2015年第6期1384-1388,共5页Journal of Electronics & Information Technology

摘　　要：人体血氧饱和度是基于脉搏波信号测量得到的,然而在脉搏波信号采集的过程中存在着由人体呼吸和仪器本身热噪声等带来的基线漂移和高频噪声,影响人体血氧饱和度的测量精度。因此,该文提出一种总体平均经验模态分解(Ensemble Empirical Mode Decomposition,EEMD)与基于排列熵(Permutation Entropy,PE)的信号随机性检测相结合的方法,同时消除基线漂移和高频噪声。对脉搏波信号进行EEMD分解,计算分解到得到的内在模式分量的排列熵,选取阈值,分别判断并剔除代表高频噪声和基线漂移的内在模式分量。最后信号重构就得到同时消除高频噪声和基线漂移的脉搏波信号。通过自行研制的测量装置所采集的脉搏波信号进行实验验证,利用信号的频谱和交直流比R评价效果。结果表明:该方法有效地同时消除了脉搏波信号中的高频噪声和基线漂移,这将有利于人体血氧饱和度测量精度的提高。The measurement of blood-oxygen saturation is based on the factors impact the accuracy of t, such as high frequency noise pulse wave signal, but there are many caused by instrument thermal noise and baseline drift caused by the breath. A method which combines Ensemble Empirical Mode Decomposition （EEMD） and Permutation Entropy （PE） is proposed, it can decrease high frequency noise and baseline drift. The pulse wave signal is decomposed by EEMD, the PE of each Intrinsic Mode Function （IMF） is calculated and the threshold value of PE is chosen. Then the IMFs which present high frequency noise and baseline drift are judged and decreased. Finally, the signal without high frequency noise and baseline drift is achieved. A self-developed measurement device is used to obtain the pulse wave for testing validation, and the signal spectrum and AC-DC modulation ratio value are adopted to evaluate the effect. The result shows that this method could effectively remove high frequency noise and baseline drift, which is conducive to improve the accuracy of blood-oxygen saturation.

关 键 词：脉搏波信号 人体血氧饱和度 高频噪声 基线漂移 总体平均经验模态分解 排列熵 

分 类 号：R318.04[医药卫生—生物医学工程]