# Numerical validation of the applicability of the simplified ventricular model in the analysis of hemolysis in the mitral paravalvular leak

## TL;DR

This study shows that simplified heart models can accurately predict blood damage in heart valve leaks, saving time in medical diagnostics.

## Contribution

The paper demonstrates that simplified ventricular models can replace complex, patient-specific models for hemolysis analysis in mitral paravalvular leaks.

## Key findings

- Simplified ventricular models produced hemodynamic results comparable to complex, patient-specific models.
- Hemolysis-related shear stresses showed no significant differences across model types.
- Using static and universal meshes reduces computational and preprocessing time without sacrificing accuracy.

## Abstract

In this paper, we explore various approaches to model the hemodynamic changes during cardiac contraction in the presence of a mitral paravalvular leak. Using computational fluid dynamics and large deformation diffeomorphic metric mapping, we conducted simulations that represented ventricular motion in four distinct ways. Taking tomography data into account, we developed a heart model that accurately reproduced the actual heart structure. Two simplifications for ventricular geometry to streamline the modeling process were proposed: a static mesh and a universal geometry. The simulation results from the most intricate variant, the CT-based, real model with dynamic mesh, were compared with the outcomes from the simplified approaches, universal geometry and static mesh. The simulations described unsteady flow dynamics during contraction, using a non-Newtonian Carreau-Yasuda blood rheological model. As expected, the hemodynamic conditions and parameter values derived from the hemolysis criterion (shear stresses exceeding 300 Pa) demonstrated no significant discrepancies between the various models under scrutiny. This suggests that the analysis of this phenomenon can be simplified to employ a static and universal ventricular mesh, eliminating the necessity for patient-specific medical imaging of the ventricle. Such a simplification can significantly reduce preprocessing and computational time, making this model more practical for routine medical diagnostics.