Computer reconstruction of a human lung boundary model from magnetic resonance images

A mathematical description of the morphology of the lung is necessary for modeling and analyzing the deposition of inhaled aerosols. A model of the lung boundary was generated from magnetic resonance images, with the goal of creating a framework for anatomically realistic morphological models of the human airway network. We used data visualization and analysis software to reconstruct the lung volume from a series of transverse magnetic resonance images collected at many vertical locations in the lung, ranging from apex to base. The lung model was then built using isosurface extraction techniques. These modeling methods may facilitate the creation of customized morphological models for indi-vidual subjects, resulting in improved interpretation of aerosol distribution data from single-pho-ton-emission computed tomography (SPECT). Such customized models could be developed for children and for patients with respiratory diseases, thus aiding in the study of inhaled medications and environmental aerosols in these sensitive populations. Key words: magnetic resonance imaging; single-photon-emission computed tomography; lung, modeling; computer simulation; models, theoreti-cal; models, anatomical. [Respir Care 2004;49(2):180–185. © 2004 Daedalus Enterprises