Fluid-electro-mechanical model of the human heart for supercomputers
The heart is a complex system. From the transmembrane cell activity to the spatial organization in helicoidal fibers, it includes several spatial and temporal scales. The heart muscle is surrounded by two main tissues that modulate how it deforms: the pericardium and the blood. The former constrains...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis doctoral |
| Estado: | Versión publicada |
| Fecha de publicación: | 2018 |
| País: | España |
| Institución: | CBUC, CESCA |
| Repositorio: | TDR. Tesis Doctorales en Red |
| OAI Identifier: | oai:www.tdx.cat:10803/481980 |
| Acceso en línea: | http://hdl.handle.net/10803/481980 https://dx.doi.org/10.5821/dissertation-2117-116302 |
| Access Level: | acceso abierto |
| Palabra clave: | Àrees temàtiques de la UPC::Enginyeria biomèdica 004 615 616.1 |
| Sumario: | The heart is a complex system. From the transmembrane cell activity to the spatial organization in helicoidal fibers, it includes several spatial and temporal scales. The heart muscle is surrounded by two main tissues that modulate how it deforms: the pericardium and the blood. The former constrains the epicardial surface and the latter exerts a force in the endocardium. The main function of this peculiar muscle is to pump blood to the pulmonary and systemic circulations. In this way, solid dynamics of the heart is as important as the induced fluid dynamics. Despite the work done in computational research of multiphysics heart modelling, there is no reference of a tightly-coupled scheme that includes electrophysiology, solid and fluid mechanics in a whole human heart. In this work, we propose, develop and test a fluid-electro-mechanical model of the human heart. To start, the heartbeat phenomenon is disassembled in the different composing problems. The first building block is the electrical activity of the myocytes, that induces the mechanical deformation of the myocardium. The contraction of the muscle reduces the intracavitary space, that pushes out the contained blood. At the same time, the inertia, pressure and viscous stresses in this fluid exerts a force on the solid wall. In this way, we can understand the heart as a fluid-electro-mechanical problem. All the models are implemented in Alya, the Barcelona Supercomputing Center simulation software. A multi-code approach is used, splitting the problem in a solid and a fluid domain. In the former, electrophysiology coupled with solid mechanics are solved. In the later, fluid dynamics in an arbitrary Lagrangian-Eulerian domain are computed. The equations are spatially discretized using the finite element method and temporally discretized using finite differences. Facilitated by the multi-code approach, a novel high performance quasi-Newton method is developed to deal with the intrinsic issues of fluid-structure interaction problems in iomechanics. All the schemes are optimized to run in massively parallel computers. A wide range of experiments are shown to validate, test and tune the numerical model. The different hypothesis proposed — as the critical effect of the atrium or the presence of pericardium — are also tested in these experiments. Finally, a normal heartbeat is simulated and deeply analyzed. This healthy computational heart is first diseased with a left bundle branch block. After this, its function is restored simulating a cardiac resynchronization therapy. Then, a third grade atrioventricular block is simulated in the healthy heart. In this case, the pathologic model is treated with a minimally invasive leadless intracardiac pacemaker. This requires to include the device in the geometrical description of the problem, solve the structural problem with the tissue, and the fluid-structure interaction problem with the blood. As final experiment, we test the parallel performance of the coupled solver. In the cases mentioned above, the results are qualitatively compared against experimental measurements, when possible. Finally, a first glance in a coupled fluid-electro-mechanical cardiovascular system is shown. This model is build adding a one dimensional model of the arterial network created by the Laboratório Nacional de Computação Científica in Petropolis, Brasil. Despite the artificial geometries used, the outflow curves are comparable with physiological observations. The model presented in this thesis is a step towards the virtual human heart. In a near future computational models like the presented in this thesis will change how pathologies are understood and treated, and the way biomedical devices are designed. |
|---|