A finite element framework for high performance computer simulation of blood flow in the left ventricle of the human heart
2015 (English)Report (Other academic)
Progress in medical imaging, computational fluid dynamics and high performance computing enables computer simulations to evolve as a significant tool to enhance our understanding of the relationship between cardiovascular diseases and hemodynamics. The field of cardiac flow simulations is highly interdisciplinary and challenging. Therefore, the aim of our research is to build a simple and reliable framework for modeling and simulation of the blood flow in the heart that is both easy to modify and flexible to extend. In this paper, we present a patient-specific Arbitrary Lagrangian-Eulerian finite element framework for simulating the blood flow in the left ventricle of a human heart using high performance computing. The mathematical model is described together with the discretization method, mesh smoothing algorithms, and the parallel implementation in Unicorn which is part of the open source software framework FEniCS-HPC. The parallel performance is demonstrated, a convergence study is conducted and intraventricular flow patterns are visualized.The results capture essential features observed in other numerical models and imaging techniques, and thus indicate that our framework possesses the potential to provide relevant clinical information for diagnosis and medical treatment.Several studies have been conducted to calculate the three dimensional blood flow in the left ventricle of the human heart with prescribed wall movement. Our contribution to the field of cardiac research lies in combining patient-specific measurements and parallel computing in one flexible and robust open source software framework.
Place, publisher, year, edition, pages
KTH Royal Institute of Technology, 2015. , 17 p.
CTL Technical Report, 34
Finite element method, Arbitrary Lagrangian-Eulerian method, parallel algorithm, blood flow, left ventricle, patient-specific heart model
IdentifiersURN: urn:nbn:se:kth:diva-181110OAI: oai:DiVA.org:kth-181110DiVA: diva2:898808
FunderEU, European Research Council, 202984Swedish Research CouncilSwedish Foundation for Strategic Research
QC 201602122016-01-292016-01-292016-02-12Bibliographically approved