Medizinische Universität Graz - Research portal
SHR
Neuro
Cancer
Cardio
Metab
Microb
Lipid
High-resolution In Silico Models of Total Heart Function
- Abstract
- Advances in tomographic imaging have enabled unprecedented ability to image cardiac anatomy. However, these technologies have had relatively modest clinical impact. Key challenge have been the use of imaging to extract important functional information and to understand how image-derived information couples with other physiological aspects that cannot be understood through imaging alone. In-silico models hold vast potential to address these challenges by enabling the integration of imaging data into quantitative frameworks that can aid in better understanding cardiac function in health and disease, and developing optimal treatment strategies.
Cardiac function arises from interactions between different physics, including electrophysiology, structural mechanics and fluid dynamics. While these physics are bidirectionally linked, virtually all reported modeling studies have considered them in isolation. In no small part this is due to the complexity of multiphysics models and the difficulties in devising advanced numerical methods and their efficient implementation on modern computing architectures. Addressing these challenges as we propose is key to keep model execution and analysis cycles tractable and, thus, to leverage modeling in a translational context. The proposed work goes beyond the state-of-the-art in many regards. Combined models of both heart and attached outflow vessels will be used to provide natural inflow boundary conditions for simulating blood flow and to facilitate studying feedback mechanisms between flow and pressure in the outflow vessels and pumping performance. This set of novel features and the parallel efficiency of the implemented methods will provide a unique platform for translational research. At the fundamental level, this model will be among the most advanced basic research tools for gaining mechanistic insight into cardiac pumping function. This challenging endeavor is feasible only as it builds upon combining the expertise of the applicant Christoph Augustin (Medical University of Graz) in modeling electrophysiology and soft tissue mechanics, and the host Shawn Shadden (UC Berkeley) in computational modeling of blood flow. Clinical datasets for model parametrization and validation are provided by Titus Kühne at the German Heart Centre Berlin, as well as by clinical collaborators of Prof. Shadden at the University of California San Francisco.Technical challenges to be addressed relate to multiphysics coupling methods which converge robustly and scale efficiently. While such methods have been developed recently for single physics cardiac simulations, this has not been the case for coupled multiphysics models of total heart function. This will be addressed by massively parallel implementations which are based on domain decomposition and strongly scalable iterative methods using appropriate preconditioning and stabilization techniques. In summary, the overall objective is to develop the most advanced in-silico model of total heart function, to facilitate biophysically detailed studies of electro-mechano-fluidic function of the heart. This model will be parametrized, verified and employed to study bidirectional cause-effect relationships between flow and pressure in the outflow vessels and pumping performance.
- Keywords
- Computermodell
- Hochaufgelöste, anatomisch detailierte Modelle
- MRI-basierte Validierung
- Multiphysiksimulationen
- Supercomputing
- Project Leader:
- Augustin Christoph
- Duration:
- 01.04.2017-01.04.2017
- Programme:
- MAX KADE
- Type of Research
- basic research
- MUG Research Units
- Division of Medical Physics and Biophysics
- Project partners
-
University of California, Berkeley, United States (USA)
Contact person: Prof. Shawn Shadden;