Topological Data Analysis of Murine Pulmonary Arterial Networks Under Hypoxia-Induced PH

Cardiovascular disease (CVD) is the world\u27s leading cause of mortality, annually claiming about 17.9 million lives. CVD conditions include structural and functional problems, diseased vessels, and blood clots, causing patients to experience heart failure and eventually leading to death. One incurable CVD is pulmonary hypertension (PH), defined as blood pressure above 25 mmHg in the main pulmonary artery and accompanied by vascular remodeling. This study uses topological data analysis (TDA) to analyze pulmonary arterial networks from control and PH mice. The networks are segmented and skeletonized from micro-CT images. Mice with PH had thicker vessels, causing their segmented networks to falsely appear to contain more arteries than control networks. To obtain networks of comparable size, we applied tree-pruning algorithms based on vessel radius and Strahler order. TDA is used to compute persistent homology of the networks. Results from a height filtration show total degree 0 persistence is lower in PH networks than in control. The alpha-complex filtration and bottleneck distances between persistence diagrams will also be discussed. This work was completed as part of an REU at NCSU