Pulmonary Artery

Research on pulmonary arteries focuses on accurately segmenting these vessels from medical images (CT scans, MRI) to improve diagnosis and treatment planning for pulmonary diseases. Current efforts leverage deep learning, employing architectures like U-Net, Swin UNETR, and hybrid image/graph convolutional networks to achieve fast and accurate segmentation, even from non-contrast CT scans. These advancements enable more efficient computational fluid dynamics simulations and reveal demographic associations with pulmonary vasculature anatomy, ultimately contributing to a better understanding of pulmonary physiology and improved patient care.