1348525 - Computational Simulation of Vorticity and Viscous Energy Loss in Repaired Tetralogy of Fallot using Conventional Cine

Repaired Tetralogy of Fallot (rTOF) patients may develop right ventricle (RV) dilation from chronic pulmonary insufficiency and eventually require pulmonary valve replacement (PVR)(1). Cardiac magnetic resonance (CMR) studies with 4D flow have implicated abnormal vorticity and viscous energy loss (VEL) in RV dysfunction (2–4); thus these parameters have the potential to guide PVR therapy. However, 4D flow sequences have limitations in spatial and temporal resolution that preclude detailed analyses. Simulation of intracardiac flows with computational fluid dynamics (CFD) holds promise for highly resolved and meaningful hemodynamic RV assessments.We aim to use conventional CMR imaging and a unique immersed-boundary method CFD solver (5) to simulate intracardiac flow in rTOF patients compared to the normal RV.

Methods: Conventional CMR datasets from rTOF patients and normal controls were used for this IRB-approved study.. A three-dimensional (3D) RV model was segmented from noncontrast, 3D steady-state free precession images. Feature-tracking (QStrain, Medis, Netherlands) was used to capture RV wall motion from CMR long-axis and short-axis cine stacks. A kinematic model of the RV over the cardiac cycle was then created via diffeomorphic mapping (Deformetrica, open-source, deformetrica.org). 2D phase contrast imaging provided the outlet boundary condition at the main pulmonary artery. Fully-resolved direct numerical simulations were performed over multiple cardiac cycles, with the RV kinematic model immersed in a Cartesian domain. Systolic/diastolic vorticity and VEL (averaged over RV volume and 5 cardiac cycles) were measured and compared. Q-Criterion was used to qualitatively evaluate vortex formation.

Results:

Twelve rTOF patients (22 ± 6 years, BSA 1.8 ± 0.2 m²) and twelve normal controls (17 ± 2 years, BSA 1.8 ± 0.2 m²) were included for analysis. rTOF had significantly elevated systolic/diastolic vorticity and VEL compared to normal controls (Figure 1).  Qualitative visualization demonstrated pulmonary insufficiency jet propagation into the RV apex, limiting the typical tricuspid ring-vortex seen in normal controls (5 examples in Figure 2). Five of the rTOF patients underwent PVR, and subsequent CMR analyses  demonstrated interval reorganization of the tricuspid ring-vortex (3 examples in Figure 3).

Conclusion:

Our CFD methodology can simulate the RV intracardiac flow environment in rTOF patients. rTOF patients have abnormal intracardiac vorticity and VEL in the RV. This computational framework has the potential to predict intracardiac changes after PVR, improving its clinical indications as guided by CMR.