Image-driven, patient-specific simulations in the right heart: Extracting valve geometry and kinematics from blood...

Although recent advancements in imaging technologies and computational power have significantly enhanced our ability to investigate intracardiac flow dynamics of the right heart (RH), inadequate imaging spatial and temporal resolution and reliance on experimentally derived constitutive relations pose challenges in creating patient-specific models capable of quantifying important flow parameters, particularly in pediatric cases. Image-driven computational models with prescribed kinematics offer a feasible and cost-efficient solution; however, accurate modeling of intricate structures including the tricuspid valve (TV) remains hindered by the resolution constraints of current imaging techniques.

We present a novel, image-driven, patient-specific direct numerical simulation (DNS) framework that integrates TV kinematics with RH motion. The prescribed TV kinematics are informed by the 4D-flow MRI velocity field. This approach avoids reliance on fully resolved valve imaging while preserving essential valve dynamics. The results are validated through direct comparison with the 4D-MRI measurements for both healthy control and repaired tetralogy of fallot (rToF) cases.

The proposed framework successfully reproduced physiologically relevant RH flow features. Simulated velocity fields showed good qualitative agreement with 4D-MRI measurements, demonstrating the ability of the workflow to accurately capture RH hemodynamics. Volume-averaged total kinetic energy and vorticity magnitude also showed good quantitative agreement with the 4D-MRI data.

This study introduces an image-driven, patient-specific modeling of TV driven by 4D-flow data. By integrating TV kinematics with RH, the proposed approach strengthens the link between medical imaging and high-fidelity simulation, enabling accurate characterization of RH flow dynamics without relying on structural property data.