Time-frequency analysis of breathing signals: in vitro airway model

Sound signals of respiratory airflow represent summations of acoustic waves of various frequencies, which basically depend on the characteristics of the flow and on those of the surrounding tissue. This study was designed to examine the capability of time-frequency distribution (TFD) of respiratory signals in order to differentiate between unobstructed and obstructed upper airways. In order to investigate the TFD characteristics of defined upper airway geometry we conducted a controlled basic study in a laboratory system with an in vitro isolated airway model, which was either unobstructed or had concentric obstructions of various degrees at different locations along the tube. Pressure fluctuations were acquired with a microphone proximal to the airway opening. A short-term Fourier transform was used to study the TFDs of these signals. The results of the in vitro study showed that the energy of the higher frequencies increased for relatively small incremental changes in: i) reduction of the lumen cross-section, ii) decrease of distance from measurement site to obstruction, and iii) increase of breathing effort. Further development of this method may lead to noninvasive clinical techniques for early diagnosis of upper airway obstructions.