Video-based valve motion combined with computational fluid dynamics gives stable and accurate simulations of blood flow in the Realheart total artificial heart

Background Patients with end-stage, biventricular heart failure, and for whom heart transplantation is not an option, may be given a Total Artificial Heart (TAH). The Realheart (R) is a novel TAH which pumps blood by mimicking the native heart with translation of an atrioventricular plane. The aim of this work was to create a strategy for using Computational Fluid Dynamics (CFD) to simulate haemodynamics in the Realheart (R), including motion of the atrioventricular plane and valves. Methods The accuracies of four different computational methods for simulating fluid-structure interaction of the prosthetic valves were assessed by comparison of chamber pressures and flow rates with experimental measurements. The four strategies were: prescribed motion of valves opening and closing at the atrioventricular plane extrema; simulation of fluid-structure interaction of both valves; prescribed motion of the mitral valve with simulation of fluid-structure interaction of the aortic valve; motion of both valves prescribed from video analysis of experiments. Results The most accurate strategy (error in ventricular pressure of 6%, error in flow rate of 5%) used video-prescribed motion. With the Realheart operating at 80 bpm, the power consumption was 1.03 W, maximum shear stress was 15 Pa, and washout of the ventricle chamber after 4 cycles was 87%. Conclusions This study, the first CFD analysis of this novel TAH, demonstrates that good agreement between computational and experimental data can be achieved. This method will therefore enable future optimisation of the geometry and motion of the Realheart (R).

Automated summary

This study presents a novel strategy using Computational Fluid Dynamics (CFD) to simulate the haemodynamics in the Realheart (R) total artificial heart. By comparing different computational methods, it was found that video-prescribed motion was the most accurate in predicting chamber pressures and flow rates. This method shows promise for future optimization of the Realheart (R) design and operation.