Computer simulations of lung airway structures using data-driven surface modeling techniques.

Abstract

Knowledge of human lung morphology is a subject critical to many areas of medicine. The visualization of lung structures naturally lends itself to computer graphics modeling due to the large number of airways involved and the complexities of the branching systems. In this study, a method of generating three-dimensional computer simulations of human lung airway networks using data-driven, surface modeling techniques is presented. By simulating the tubular airway structures and realistic bifurcation shapes, anatomically accurate representations of human lungs are obtained. These computer models are designed for use in computational fluid dynamic applications and particle trajectory analyses, and to be complimentary to medical imaging (gamma scintigraphy) protocols.