The numerical acoustic and fluid ow analysis on a CT-scan derived upper airway model of a stridor patient

Abstract

In this report the acoustic numerical analysis of a stridor patient is discussed. Stridor is an abnormal, noisy or toned/voiced breathing sound which is caused by obstruction or malformation of pharynx, larynx, trachea or the bronchia; whereby the turbulence of air and vibration of surrounding tissue are the source of the observed sound. There is a need for uninvasive diagnostics because of risks involved using anaesthetics, especially on young children. To find the obstruction from a sound signal, first one needs to know what kind of sound is produced by what kind of geometry. Doing this numerically has the advantage of allowing one to change only one parameter at the time. Using a hybrid approach, combining LES and the Lighthill equation, the airflow acoustics for internal flows can be predicted. The acoustics are validated on a reed instrument and diaphragm. The sound pressure level is of the right magnitude and also good agreement is found for the distribution of power. The lower frequencies are however over predicted. The fluid part is validated using PIV measurements on a simplified upper airway model. Good agreement was found. The fluid and acoustic analysis on the different upper airway models showed that probably for increasing constriction the sound production mechanism shifts from turbulent sound production to tissue vibration. Therefore a rigid wall approach is not good enough to predict the sounds generated by obstructions in stridor patients.