Correlation between trachea internal airway geometry and the auscultation signal response

Human auscultation represents an easy and prompt tool for physicians to diagnose the existence of a pulmonary pathology. The contemporary practice of lung auscultation as an essential part of physical examination can be traced back to the time of Hippocrates (Bohadana et al, 2014). But revolutionary tool called “stethoscope” was only introduced by Rene Laennec in 1816 and gradually refined ever since (Wilks 1883 and Bishop 1980). The process of new physician training in acquiring pulmonary auscultation skills is still time consuming. Moreover, the success rate as revealed by the training examination is also unsatisfactorily elusive. To avoid the subjective nature of pulmonary auscultation, efforts have been carried out to digitally record and analyze lung sounds and further explore the correlation between the acoustic characteristics with the clinical symptoms. In our project, we started to tackle the problem using a microphone, and instead of carrying out experiments on human beings, we chose to use a 3D printed model from data collected from a real human patient by Pacific Northwest National Laboratory. We want to understand how the geometry of the trachea wall interacts with the air flow and thus, impacts the acoustic signal generated and acquired by the microphone. Once we understand the pattern for a healthy trachea, we alter its internal geometry to simulate a tracheal disease. For this, we investigated a simplified model of stenosis by simply create a small obstruction inside the trachea. Using an A/D board NI USB 6003 with maximum sampling rate of 100 kHz and 16-bit resolution and by processing the signal using “Labview” and Matlab, we established a correlation between the different tracheal shapes and the generated acoustic signal. Using an Ansys Fluent Turbulent CFD simulation, we matched the air flow events inside the trachea with the Power Spectral Density peaks of the tracheal sound.