北京生物医学工程

基于小波阈值法的脉搏波去噪算法研究

Pulse Wave Denoising Algorithm Based On the Wavelet Threshold method

作者：吴星林林陈海军徐之标

单位：广东医科大学生物医学工程学院 (广东东莞 523808)

关键词：脉搏波；运动伪差；小波阈值法；基线漂移；信噪比；均方差；平滑度

分类号：R318.04

出版年·卷·期（页码）：2021·40·1（38-45）

摘要：

目的消除可穿戴式脉搏波监测设备在连续测量中由于运动造成的运动伪差，保证设备准确性和稳定性。方法通过选取合适的小波基、小波最大分解层数、阈值函数和阈值方法，对脉搏波信号进行小波阈值处理，提出了一种基于小波阈值法去除脉搏波噪声的算法。并针对在脉搏波信号采集过程中出现的基线漂移、工频干扰和运动伪差，与加窗傅里叶变换去噪后的结果进行对比。结果在信噪比、均方差和平滑度等关键指标上，小波阈值法的效果更优。利用db9小波基对脉搏波信号进行6层小波分解，设置启发式阈值所得到的处理效果最好。结论该算法能够有效抑制工频干扰和运动干扰，使信噪比提高22dB，均方差接近于0，且平滑度降为原来的11%，实现脉搏波信号采集中干扰的有效去除。

Objective To eliminate the error caused by motion in the continuous measurement of the wearable pulse wave monitoring device so that the accuracy and stability of the device can be ensured. Methods By choosing the suitable wavelet base, the biggest wavelet decomposition layer, the threshold function and threshold method, we present an algorithm based on the wavelet threshold method to remove pulse wave noise and compare the results of windowed Fourier transform denoising with the baseline drift interference and motion error in pulse wave signal acquisition. Results The wavelet threshold method is more effective in key indicators such as the signal-to-noise ratio, mean square deviation and smoothness. In addition, after using DB9 wavelet base to decompose pulse wave signal with a 6-level wavelet and setting the heuristic threshold, we find that the processing effect is the best. Conclusions This algorithm can effectively suppress power frequency interference and motion interference so that the signal-to-noise ratio will be increased by 22dB, the mean square deviation will get close to 0, and the smoothness will increase to the original 11%, which will realize the effective removal of interference in the pulse wave signal acquisition.

参考文献：

[1] Fedotov AA. Selection of parameters for filtering distal arterial pulse signal using multi-resolution wavelet transforms[J]. Biomedical Engineering, 2013, 47(3):146-149.

[2] Hoseini MR, Zuo MJ, Wang X. Denoising ultrasonic pulse-echo signal using two-dimensional analytic wavelet thresholding[J]. Measurement, 2012,45(3):255-267.

[3] Wang YF, Guo GL, Li ZQ. Study on pulse-signal detection methods using wavelet transform and Hilbert Huang transform[J]. Advanced Materials Research, 2013,860-863:2918-2923.

[4] Andreev DA, Bozhokin SV, Venevtsev ID,et al. Gabor transform and continuous wavelet transform for model pulsed signals[J]. Technical Physics, 2014, 59(10):1428-1433.

[5] Amin HU, Malik AS, Ahmad RF, et al. Feature extraction and classification for EEG signals using wavelet transform and machine learning techniques[J]. Australasian Physical & Engineering Sciences in Medicine, 2015, 38(1):1-11.

[6] Bhattacharyya A, Pachori R, Acharya U. Tunable-Q Wavelet transform based multivariate sub-band fuzzy entropy with application to focal EEG signal analysis[J]. Entropy, 2017,19(3):1-14.

[7] Hayes MJ, Smith PR. A new method for pulse oximetry possessing inherent insensitivity to artifact[J]. IEEE Transactions on Biomedical Engineering[J]. 2001, 48(4):452-461.

[8] Lee S, Ibey BL, Xu W, et al. Processing of pulse oximeter data using

discrete wavelet analysis[J]. IEEE Transactions on Biomedical Engineering[J]. 2005, 52(7):1350-1352.

[9] Gibbs P, Ashada HH. Reducing motion artifact in wearable bio-sensors using mems accelerometers for active noise cancellation[C] // Proceedings of the 2005, American Control Conference, Portland ：IEEE，2005：1581-1586.

[10] Yan YS，Zhang YT. An efficient motion-resistant method for wearable pulse

oximeter[J]. IEEE Transactions on Information Technology in Biomedicine， 2008, 12(3):

399-405.

[11] Lee SH, Lim JS, Kim JK, et al. Classification of normal and epileptic seizure EEG signals using wavelet transform, phase-space reconstruction, and Euclidean distance.[J]. Computer Methods & Programs in Biomedicine, 2014, 116(1):10-25.

[12] Boggess A ，Narcowich FJ. 小波与傅里叶分析基础[M]. 芮国胜,康健译. 北京:电子工业出版社,2004.

[13] 蒋曲博, 甘永进, 张翠娜. 反射式血氧饱和度检测系统研制[J]. 中国医学物理学杂志, 2017, 34(1):58-64.

Jiang QB, Gan YJ, Zhang CN. Development of reflective blood oxygen saturation detection system [J]. Chinese Journal of Medical Physics,2017, 34(1):58-64.

[14] 罗志昌, 张松, 杨益民. 脉搏波的工程分析与临床应用[M]. 北京: 科学出版社, 2006.

[15] 张爱华, 王平, 丑永新. 基于动态差分阈值的脉搏信号峰值检测算法[J]. 吉林大学学报: 工学版, 2014, 44(3): 847-853.

Zhang AH, Wang P, Chou YX. Peak detection of pulse signal based on dynamic different threshold[J]. Journal of Jilin University (Engineering and Technology Edition)， 2014, 44(3): 847-853.

[16] Chen HK, He XY, Tang HM, et al. Study on wavelet denoising of pulse impactforce of non-viscosity debris flow[J]. Applied Mechanics and Materials, 2012, 249-250:1040-1046.

[17] 金燕, 王丹浓, 刘国越. 基于迭代算法的暂态电能质量扰动信号消噪[J]. 浙江工业大学学报, 2011, 39(1):92-96.

Jin Y, Wang DN, Liu GY. Transient power quality disturbance signal denoising based on a recursive algorithm[J]. Journal of Zhejiang University of Technology, 2011,39(1):92-96.

[18] Dey D . Comparison of FFT, DCT, DWT, WHT compression techniques on electrocardiogram and photoplethysmography signals[C]// International Conference on Computing. Communication and Sensor Network. Rourkela, India ：CCSN, 2012:6-11.

[19] Sidhik S. Comparative study of Birge-Ma ssart strategy and unimodal thresholding for image compression using wavelet transform[J]. Optik, 2015, 126(24):5952-5955.

服务与反馈：

【文章下载】【加入收藏】

提示：您还未登录，请登录！点此登录