Relationship between airflow and frequency-based features of tracheal respiratory sound

The relationship between tracheal respiratory sounds and airflow has always been of great interest for physicians and researchers due to its diagnostic applications. Despite the magnitude of research in this field the exact relationship is still in debate. The goal of this study was to review the most common frequency-based models describing such a relationship. For this purpose, tracheal sounds and airflow signals were recorded simultaneously from 13 healthy subjects at four different target flow rates. The power spectrum density (PSD) of tracheal sound was calculated for every 100 ms segment with 50 % overlap between successive segments, and frequency-based features were derived from the PSD of each segment. Exponential and second-degree polynomial relationships between target airflow and these frequency-based features were investigated. For each model, the mean square error (MSE) was calculated and averaged between the subjects. The results suggest the average power of tracheal sound (P_ave), and the maximum frequency have strong relationship with airflow in exponential and second-degree polynomial models, respectively. The effect of age on the maximum frequency was also investigated and it was found that the model parameters change with age.