Lung tissue has unique mechanical properties that make it difficult to impossible, to image using conventional modalities (ultrasound, magnetic resonance, x-ray computed tomography). Understanding acoustical wave propagation in the pulmonary system and chest region is an important step towards building diagnostic techniques based on audible frequency acoustics that go beyond that of simple stethoscopic auscultation of lung sounds or chest percussion. Envisioned noninvasive and inexpensive techniques that would rely on a much-improved understanding of pulmonary and chest acoustics could incorporate such things as multi-sensor measurements for forming 3-dimensional acoustic “images”. More sophisticated acoustic measurements and interpretations could also be part of a multi-mode imaging strategy complementing or adding value to other conventional imaging methods.

Computational meshing tools for discretizing the lung acoustical problem domain are developed for simple geometries and for actual human lung geometry taken from the National Library of Medicine Visible Human Project. An acoustic boundary element (BE) model for the lungs that utilizes these meshes is developed and evaluated numerically and experimentally. A source localization algorithm using the BE model is implemented to be used with acoustic sensor array measurements on the chest surface to help identify lung pathologies. In order to experimentally validate the computational developments, materials having similar acoustic properties to human tissue are identified and experimentally quantified. Mechanical phantom models are designed, built and tested to simulate pneumothorax and hydrothorax conditions and their diagnosis via acoustic transmission and acoustic percussion approaches.

Use of the boundary element method for simulating complicated acoustic wave propagation in the lung tissue is a unique contribution relative to the overly-simplistic free field approach that has been used previously in literature. With some improvements, the developed computational analysis and experimental measurement techniques may aid in the development of improved noninvasive acoustic methods for identifying life-threatening pulmonary conditions.