Patient-specific computational fluid dynamic modeling accurately predicts systolic pressure gradients across stenotic right ventricular outflow tracts

Background

Transcatheter options for interventions on stenotic right ventricular outflow tracts have become the standard of care for many individuals. Determining which patients should have interventions uses a combination of echocardiography derived gradients and axial imaging scans. Being able to predict catheter derived gradients from scans using computational fluid dynamic modeling can help with the pre-procedural planning of patients.

Objectives

The aim of this study is to validate a computational fluid dynamic modeling technique that accurately predicts pressure gradients when compared to in vivo cardiac catheterization measurements across a stenotic right ventricle outflow tract.

Methods

A retrospective chart review of patients who underwent TPVR (transcatheter pulmonary valve replacement) at our institution was performed. Five patients who underwent TPVR and had a CT scan prior to intervention were identified for this study. Each CT was processed using a 3D Slicer to build a 3-dimensional (3D) model. The model was then processed via ANSYS ICEM software to construct a mesh model for CFD testing. ANSYS Solver was then used to model the fluid dynamics through the heart model.

Results

CFD results were obtained for the models and compared to the catheterization data for the specific patient. Linear regression demonstrated a very strong correlation between our model gradients and the gradients obtained during cardiac catheterization with an adjusted R of 0.9959. Model coefficient values were beta-1 = 0.9329 and beta-0 = 3.2916 (p = 0.001).

Conclusions

This proof-of-concept study has shown that taking 3-dimensional imaging and building a CFD model can accurately and reliably predict the change in pressure across the right ventricular outflow tracts. This proof-of-concept model can hopefully be applied to the pre-catheterization planning prior to patients requiring a PVR once further refinement and validation have been performed.