University of Twente / Politecnico di Milano
Model weights not public. Contact creators for more information.
Machine-learning surrogate model for estimating pulsatile hemodynamic fields (velocity, pressure) in coronary arteries from a steady-state computational fluid dynamics (CFD) prior, avoiding the high computational cost of full pulsatile CFD. The model, a neural field conditioned on hemodynamic boundary conditions, is discretisation-independent and can be parametrised with message-passing or self-attention layers by relaxing point-wise action to permutation-equivariance. Evaluated on 74 stenotic coronary arteries from coronary CT angiography (CCTA) with patient-specific pulsatile CFD as ground truth, the model produced accurate, discretisation-independent estimates of pulsatile velocity and pressure fields.
Architecture
Hybrid
Neural field / deep vectorised operator conditioned on steady-state hemodynamic boundary conditions, parametrised via message-passing or self-attention layers for discretisation-independent, permutation-equivariant estimation of pulsatile flow fields
Framework
PyTorch
Added to catalog
2026-08-14
CCTA Stenotic Coronary Artery Cohort (patient-specific CFD ground truth)
Stenotic coronary arteries extracted from CCTA with patient-specific pulsatile CFD ground truth; count is arteries, not necessarily unique patients
Pulsatile hemodynamic (velocity and pressure) fields in coronary arteries estimated from a steady-state CFD prior