Computational fluid dynamics (CFD) are useful in the assessment of blood flow conditions in patients with congenital heart disease. A necessary, time-consuming step in the creation of CFD models is the segmentation of the anatomy of interest. In this work, a neural network was trained to segment the aorta and the pulmonary arteries in 3D MRI, and its performance was evaluated in the context of a CFD application. The network performs well in terms of Dice score and is shown to lead to accurate pressure and flow velocity fields, with errors at the level of inter-observer variability.
This abstract and the presentation materials are available to members only; a login is required.