The Relationships Between Characteristics of Wheezes and Spectral-Temporal Parameters of Forced Expiratory Tracheal Noises

Computer-based breathing sound analysis has a high potential for respiratory system assessing. However, insufficient knowledge of the origin of breathing noises, the difficulty of standardization of their registration and digital signal processing methods slow down the widespread use of the acoustic methods. Forced expiratory tracheal noises (FE) analysis showed its effectiveness for detecting hidden bronchial obstruction and monitoring of the pulmonary ventilation after underwater dives. In addition, of interest is the study of forced expiratory wheezes (FEWs) as indicators of airflow limitation in normal and with bronchial obstruction. However, the estimation of their parameters seems to be insufficiently reliable due to non-stationarity. An original processing algorithm was developed with splitting the signal into 200-Hz bands in the range of 200–2000 Hz and automatically determining the duration and energy of noise in these bands. The aim of this work was to study the relationship between the semi-automatically determined spectral characteristics of FEWs and the frequency-time parameters of tracheal FE noises in 200 Hz bands in a sample of healthy individuals of both sexes aged 20–59 years. Significant correlations were found between spectral characteristics of wheezes and the energy and power of tracheal noises in the bands. Moreover, the strongest relationships were observed in the mid-frequency range. A linear regression model of acceptable quality was constructed, which showed that up to 70% of the variability in the average power of tracheal noises in the 200–600 Hz band can be explained by the influence of the amplitudes of the mid-frequency FEWs maxima.