The widespread deployment of high-speed multi-detector CT (MDCT) scanners in the US market is creating opportunities for 3D analysis as single breath hold scans with isotropic voxels is routinely performed using these machines. This increase in the quality and quantity of lung scan data will demand routine use of quantitative tools to assess airway structures and parenchymal tissue distal to and subtended by the segmented airways. Integrated quantitative analysis of bronchial structures and lobar-level parenchymal tissue will allow for more accurate diagnosis of focal lung disease to aid existing interventions such as lung volume reduction surgery as well as newer intra-bronchial valve technology that is currently entering human trials.
Specific Aims : 1) Develop and demonstrate proof of concept for MDCT image analysis incorporating lung and lobe segmentation, airway tree segmentation, and intensity-based lung pathology assessment for emphysema. This will include: a) 3D methodologies for partitioning the lobes into surgically-accessible and surgically-inaccessible regions b) 3D methodologies for identification of low attenuation clusters (LACs) to assess distribution of """"""""hole"""""""" sizes on a lobar basis. c) application of these methods to both standard-dose and low-dose CT data sets. 2) Develop and demonstrate proof of concept for MDCT image analysis incorporating lung and lobe segmentation, airway tree segmentation, and intensity-based lung pathology assessment for the guidance of endobronchial interventions to treat emphysema. Long-term development of this technology in the future will include integration of our patented adaptive multiple feature method of parenchymal tissue analysis to facilitate integrated quantitative analysis of other lung pathologies and the development of longitudinal analysis methods to allow the use of these methods in ongoing characterization of COPD patients and those with emphysema in particular.