一种新的互补集合经验模态分解呼吸提取方法

A New Complementary Empirical Ensemble Mode Decomposition Method for Respiration Extraction

摘要

呼吸与心脏疾病、睡眠呼吸暂停和情绪有关,监测呼吸对于诊断和管理多种疾病具有重要意义.提出一种新的互补经验模态分解的呼吸提取方法,对心电信号和高斯白噪声分别进行EEMD分解,将得到的IMF分量进行复合降噪处理,选择呼吸频段内新的IMF与实测呼吸信号做相关分析,由相关性增量最大原则自适应地确定最佳幅值噪声系数,从心电中导出呼吸信号.在MIT-BIH多导睡眠数据库中进行实验验证,提出的呼吸提取方法与CEEMD算法对比,平均MSE减小3.95%、平均RMSE减小2.74%、平均MAE减小2.52%,计算耗时减少37.5%且IMF分量物理意义更明确.本算法具有较好的准确性、鲁棒性和自适应性,为呼吸信号的提取提供了一种新的解决思路.

Abstract

Respiration has been linked to heart disease, sleep apnea, and mood, and monitoring respiration is important for diagnosing and managing a variety of conditions. In this paper, a novel breath extraction method based on complementary empirical mode decomposition is proposed, the ECG signal and Gauss white noise were respectively decomposed by integrated empirical mode, and the obtained eigenmode function was combined with noise reduction. The new IMF in the respiratory band was selected for correlation analysis, and the optimal amplitude noise factor was determined adaptively by the principle of maximum correlation increment, and the respiratory signal was derived from the ECG. Experimental verification was carried out in MIT-BIH-Polysomnographic database. Compared with CEEMD method, the EDR extraction method proposed in this paper reduced the average MSE by 3.95%, the average RMSE by 2.74%, and the average MAE by 2.52%. The decomposition calculation time was decreased by 37.5% and the physical significance of IMF component was clearer. The method has good accuracy, robustness and adaptability, and provides a new solution for respiration signal extraction.

关键词

心电图 / 互补经验模态分解 / 心电导出呼吸

Key words

Electrocardiography (ECG) / new complementary ensemble empirical mode decomposition / ECG-derived respiration (EDR)

参考文献

[1] Qi M S, Xing L, Chao W, et al. Cardiopulmonary coupling analysis predicts early treatment response in depressed patients: A pilot study[J]. Psychiatry Research, 2019, 276: 6-11.

[2] Hietakoste S, Armanac-Julian P, Karhu T, et al. Acute cardiorespiratory coupling impairment in worsening sleep apnea-related intermittent hypoxemia[J]. IEEE Trans Biomed Eng, 2023, 99: 1-9.

[3] Maier C, Wenz H, Dickhaus H. Robust detection of sleep apnea from Holter ECGs. Joint assessment of modulations in QRS amplitude and respiratory myogram interference[J]. Methods Inf Med, 2014, 53(04): 303-307.

[4] 韩宇, 张兴敢. 基于改进变分模态分解的生命体征检测[J]. 南京大学学报(自然科学), 2022, 8: 680-688.

[5] 杨俊, 黄俊, 陶威. 基于FMCW雷达的自适应生命信号提取方法[J]. 雷达科学与技术, 2022, 20: 187-194.

[6] Gao Y, Yan H, Xu Z, et al. A principal component analysis based data fusion method for ECG-derived respiration from single-lead ECG[J]. Australas Phys Eng Sci Med, 2018, 41(01): 59-67.

[7] Zhao F, Xu F Y, Cai L. Application of KPCA and AdaBoost algorithm in classification of functional magnetic resonance imaging of Alzheimer’s disease[J]. Neural Computing and Applications, 2020(10): 5329-5338.

[8] Labate D, Foresta F L, Occhiuto G, et al. Emirical mode decomposition vs. wavelet decomposition for the extraction of respiratory signal from single-channel ECG: A comparison[J]. IEEE Sensors Journal, 2013, 13(07): 2666-2674.

[9] Wu Z, Huang N E. Ensemble empirical mode decomposition: a noise-assisted data analysis method[J]. Advances in Adaptive Data Analysis, 2009, 1: 1-41.

[10] Ying S, Jian Q, Li P, et al. A novel non-contact heart rate measurement method based on EEMD combined with FastICA[J]. Physiol Meas, 2023, 44(05): 055002.

[11] Yeh J R, Shieh J S, Huang N E. Complementary ensemble empirical mode decomposition: a novel noise enhanced data analysis method[J]. Advances in Adaptive Data Analysis, 2010, 2(02): 135-156.

[12] 杨克元, 邓忠文, 陈文军. 基于互补集合经验模态分解结合希尔伯特变换的光频扫描干涉信号相位提取方法[J]. 中国光学(中英文), 2023, 16(03): 682-700.

[13] Ichimaru Y, Moody G B. Development of the polysomnographic database on CD-ROM[EB/OL]. Psychiatry and Clinical Neurosciences 1999, 53: 175-177. https://physionet.org/content/slpdb/1.0.0/.

基金资助

国家自然基金项目(61901165)