Rapid diagnosis of chronic bronchitis based on spectral analysis of cough sound

Cough is a symptom that often causes patients to seek medical attention. Among the diseases, the main symptom of which is cough, chronic bronchitis (CB) leads. Currently, the generally accepted methods for assessing the etiology of cough are the collection of complaints, medical history and objective examination, and laboratory and instrumental studies. However, a large proportion of the information received about the patient is descriptive and subjective. The study aims to assess the possibility of using spectral analysis of cough sounds for rapid diagnosis of CB. Materials and methods. The main group consisted of patients suffering from mild non-obstructive catarrhal or mucopurulent CB in the acute stage (n = 64: 46.3% - men, 53.7% - women; age - 46.7 (45; 48) years). Comparison groups consisted of patients with moderate persistent asthma, uncontrolled, in the acute stage (n = 21), patients with COPD (chronic obstructive pulmonary disease), moderate exacerbation (n = 22), healthy individuals (n = 20), mild COVID-19 patients with less than 25% lung involvement by chest X-ray computed tomography, with no prior chronic respiratory tract disease (n = 20). The comparison groups were comparable in sex and age to the main group. Each subject recorded cough sounds. Using the fast Fourier transform algorithm, the frequencytime parameters of the sounds of the entire cough episode and for individual phases of the cough sound were estimated: duration (ms) (T, T1, T2, T3), the ratio of low and medium frequency energy (60-600 Hz) to high frequency energy (600-6000 Hz) (Q, Q1, Q2, Q3), the frequency of maximum energy sound (Hz) (Fmax, Fmax1, Fmax2, Fmax3). Study results. When comparing the cough sound of the main group and comparison groups, some features were revealed. The duration of coughing in CB is longer than in comparison groups. Duration of phase 2 cough is less than duration of phase 2 cough in patients with AD, COPD and induced cough of healthy individuals. Cough in patients with CB is dominated by lower frequencies compared to cough in AD and COVID-19, and compared to cough in COPD and induced cough in healthy individuals, highfrequency energy predominates. Analysis of the frequency of the maximum sound energy also revealed significant differences in this indicator of the main group and comparison groups. Based on the identified data, a model for the diagnosis of CB was developed in the form of a regression equation, when solving which the value of the integrated Y index was obtained. In patients with CB, the Y value ranged from 0.51 to one. Conclusion. The spectral characteristics of cough sounds are of differential diagnostic value, the method for assessing them is accurate, quantitative, and can significantly reduce the time for making a diagnosis of CB.

cough, spectral analysis of sound, chronic bronchitis