WIP 48 - A Novel Pipeline for Three-Dimensional (3D) Modeling of Prenatal Echocardiograms to Characterize Diagnostic Factors of Coarctation of the Aorta

Poster Number: WIP 48
Publication Number: WIP 48.783

Background: Prenatal detection of congenital heart disease (CHD) significantly improves neonatal outcomes; however, prenatal CHD detection remains poor. Among CHD, coarctation of the aorta (CoA) is a specific high-risk lesion. CoA is a narrowing of the aortic arch that restricts blood flow, making up 7% of congenital heart disease cases. Limitations in standard-of-care fetal echocardiography lead to a high false positive prenatal detection rate, driving unnecessary NICU admissions and familial stress. Current cardiovascular computational modeling methods simulate morphology, but rely on 3D images created from 2D image series over an axis, such as MRI. There is no workflow to convert 2D echocardiograms into fetal aortic models.

Objective: We aimed to develop a methodology to create patient-specific pulsatile 3D fetal aortic models from echocardiograms to improve prenatal diagnosis of CoA. We expect this modeling pipeline to translate to clinical practice by providing anatomical and hemodynamic insight to inform postnatal management.

Design/Methods: The project is IRB approved. We applied image registration to align arch views. Cross-sectional diameters were measured along vessel centerlines to develop structural segmentations, from which a 3D geometric model was reconstructed using SimVascular. Inlet boundary conditions were prescribed using patients’ Doppler velocity waveforms, and the 3-Element Windkessel Method modeled effects of downstream vasculature on outflow. Displacement of blood and vessel domains was calculated using the Arbitrary Lagrangian-Eulerian finite element analysis method to simulate pulsatile flow over cardiac cycles. This methodology will be validated by comparing predicted pressure and flow waveforms, and vessel diameters from postnatal models with cardiac MRI and catheterization data from the same patient. After validation, we will model CoA prenatally suspected fetal and postnatal echocardiograms. Our cohort is 102 participants. The control group is 51 CoA prenatally suspected, postnatally healthy neonates; the experimental group is 51 confirmed CoA neonates.