Research on Denoising Algorithm of Thoracic Impedance Data for COPD Diagnosis

COPD (Chronic Obstructive Pulmonary Disease) is a common chronic respiratory disease, which seriously affects health and life quality of patients. The limited availability of diagnostic methods restricts the symptomatic treatment of COPD. It has been found that the lung electrical impedance varies obviously with the amount of gases in the lung, impedance property has been considered a promising method for vital capacity measurement and further for lung function evaluation. As lung impedance is sensitive to the gas distribution in the lung, it could be possible to diagnose COPD symptom via the analysis of thoracic impedances. In the experimental studies, very often the thoracic impedance data have baseline drift and high frequency noise which made the extract of the peak-to-peak value of thoracic impedance difficult. At present, there are little research reported on quality improvement of thoracic impedance. This paper studies digital signal processing methods to denoise thoracic impedance data for COPD diagnosis. It is found that polynomial fitting method can effectively remove baseline drift and the five-spot triple smoothing method can effectively eliminate the high frequency noise from thoracic impedance data. The experimental studies on healthy volunteers and COPD patients have proved the effectiveness of the combination of the two methods, which laid the foundation for further investigation of COPD diagnosis based on thoracic impedance characteristics.