Comparative CT Densitometry in Murine Pulmonary Disease Models

Abstract

Using micro-computed tomography it is possible to detect the presence of pathologies which alter the lung's normal density. The density of the lungs can be altered depending on the amount of air, tissue, cells or fluid they contain. Using established mouse models of house dust mite (HDM) induced asthma, TGF-J31 induced pulmonary fibrosis (PF) and lipopolysaccharide (LPS) induced neutrophilic inflammation, this thesis examines if CT densitometry can distinguish between different pathophysiological processes. An airway segmentation method was applied to the CT images and data from these regions were assessed to determine: first, if pathologies can be detected compared to control animals; secondly, if pathological progression within each model can be measured; and finally, if it is possible to distinguish between the pathologies themselves. Lung histology and bronchoalveolar lavage fluid cytology, and total lung resistance (for the asthma model only) were assessed to confirm the disease models. The results showed that a healthy lung can be distinguished from a diseased lung in all three models. Pathological progression and resolution were also visible in the asthma and LPS groups. No changes were noted between the examined time points in the PF model. This corresponded to histological findings. It is also possible to distinguish between many of the pathologies based on the density profiles alone. Thus, CT densitometry affords a non-invasive method to longitudinally assess disease progression and resolution which is useful for the testing of novel therapeutics within the same subject. Regional CT density assessment, allows for the detection of localized pathologies around the airways which whole lung assessments may not be sensitive enough to detect.