Journal of Cardiovascular Magnetic Resonance
DOI 10.1016/j.jocmr.2026.102773CE-MRA-FLOWnet: Fast and Accurate Generative AI-Based Aortic Hemodynamic Mapping from Contrast-Enhanced MRA
Evidence
Peer-reviewed research on anatomy-derived aortic hemodynamics.
The featured studies evaluate Third Coast Dynamics' physics-informed deep-learning technology for deriving aortic hemodynamic information from anatomic imaging.
Peer-reviewed journal publications
Radiology
DOI 10.1148/radiol.240714Anatomy-derived 3D Aortic Hemodynamics Using Fluid Physics–informed Deep Learning
Featured studies
Two studies at the center of the research program.
CE-MRA-FLOWnet: Fast and Accurate Generative AI-Based Aortic Hemodynamic Mapping from Contrast-Enhanced MRA
A retrospective, single-center study evaluated a fluid physics–informed model that derived aortic hemodynamic maps from contrast-enhanced MRA and compared them with paired reference MRI measurements. The study also explored associations with adverse outcomes in a bicuspid-aortic-valve subgroup.
Open the publication (opens in a new tab)
Anatomy-derived 3D Aortic Hemodynamics Using Fluid Physics–informed Deep Learning
A retrospective study evaluated a fluid physics–informed deep-learning approach for deriving systolic three-dimensional aortic hemodynamics from aortic geometry and compared the results with reference MRI-derived measurements.
Open the publication (opens in a new tab)1,765Full retrospective cohort
352Internal test sets
60External CE-MRA test set