Research on denoising algorithm of thoracic impedance signal for respiratory monitoring during running exercise

• Published in: Biomedical signal processing and control Vol. 70; p. 102941
• Main Authors: Ge, Hao, Qin, Hui, Xue, Shan, Liu, Enkang, Zhang, Mingzhu, Bai, Zixuan, Ma, Yixin
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.09.2021
• Subjects: Bioimpedance technology; Detrend; Pulmonary function exercise test (PFET); Signal denoising; Wavelet decomposition method; Wavelet decomposition method; Pulmonary function exercise test (PFET); Bioimpedance technology; Signal denoising; Detrend
• ISSN: 1746-8094; 1746-8108
• DOI: 10.1016/j.bspc.2021.102941

•Monitoring lung ventilation during exercise has clinical significance.•Thoracic impedance is a capable measure of lung ventilation.•Extracting features of noise in thoracic impedance caused by running exercise.•Wavelet and polynomial smoothing methods based Thoracic Impedance Denoising.•wpsTID is tested to be effective using synthetic and real thoracic impedance. Pulmonary Function Exercise Test (PFET) or Cardio Pulmonary Exercise Test (CPET) using spirometer, as a method for early detection of pulmonary dysfunction or an effective supplement method to conventional pulmonary function test, has considerable clinical significance. However, the spirometer that uses air flowmeter to measure exhale/inhale air flowrate and flow velocity, increases respiratory resistance and is not comfortable for using during physical exercise. Thoracic impedance is sensitive to pulmonary ventilation and has no disturbance to human airway, hence is more suitable for respiratory monitoring, nevertheless it is influenced by many other factors, such as body movement during running exercises. In this paper, we first analyze the characteristics of thoracic impedance collected during running exercise, then put forward a wpsTID (wavelet and polynomial smoothing methods based Thoracic Impedance Denoising) algorithm to remove high-frequency alternating noises and motion artifacts to obtain clean thoracic impedance information that varies only with the pulmonary ventilation. The wpsTID algorithm is compared with a few other denoising algorithms using synthetic thoracic impedance data that containing known noises, and demonstrated to have the best performance in retaining respiratory information from the original signal and suppressing noises, which lays a foundation for the undisturbed monitoring of pulmonary ventilation during running with impedance method. Finally, the wpsTID algorithm is further verified with thoracic impedance signals of healthy volunteers measured during different kinds of motions.