Graph-based segmentation of the pediatric trachea in MR images to model growth

The upper airways are a major site of pediatric airway obstruction with its accompanying morbidity and mortality. The simplest approach to provide a stable airway is to perform a tracheotomy but it is a long recovery with its own complications. Other surgical procedures to reconstruct the airway require significant experience. The long-term objectives of this project are to develop a greater understanding of congenital abnormalities of the larynx and trachea. The objective of this thesis is to create a process to automatically segment and measure the pediatric trachea from MR images. Using 3DSlicer and ITK and program was created to perform the measurements. The software tool was optimized to produce similar results to that of CT image measurements from Pulmonary Workstation. The program was tested on a pediatric population and showed a significant correlation between cross-sectional area and age or height of the individual.