The Heart Valve Society

Back to 2017 Program


Patient Specific Numerical Modeling of Diseased Aortic Valve Hemodynamics using Fluid-Structure Interaction Approach
Huseyin C. Yalcin1, Armin Amindari2, Kadir Kirkkopru2, Asma Althani1, Magdi Yacoub3.
1Qatar University Biomedical Research Center, Doha, Qatar, 2Istanbul Technical University, Faculty of Mechanical Engineering, Istanbul, Turkey, 3Faculty of Medicine, National Heart & Lung Institute, Imperial College, London, United Kingdom.

OBJECTIVE: A diseased aortic valve disturbs hemodynamics, which is associated with heart attack risk. Abnormal hemodynamics may also trigger mechano-biological mechanisms that may cause further complications. Therefore, it is very critical to qualitatively and quantitatively assess aortic valve hemodynamics for diagnosis, for decision making of the treatment options as well as for the biological investigation of the disease progression. In the current study, we aimed to develop numerical patient specific aortic valve models using commercial ANSYS solver. METHODS: Model geometries are generated in ANSYS workbench based on b-mode echo images from patients. In these geometries, flow domain is discretized into finite volumes and aortic root and leaflets are discretized into finite elements. Doppler measured velocities at aortic inlet are used as transient velocity boundary conditions. Fluid equations (i.e. Naiver Stokes and continuity) are solved via ANSYS Fluent whereas solid (i.e. deformation) equations are solved via ANSYS Mechanical. At each time step, fluid and solid equations are coupled using System Coupling Module and solution is converged in an iterative manner. The validity of the models were verified by comparing peak transvalvular pressures from the models with catheter pressure readings. RESULTS: Using this approach, we were able to investigate hemodynamics for a normal valve, a stiffened stenosed valve, and a congenital bicuspid valve. We identified specific flow patterns near the front ventricularis and back fibrosa surfaces of normal valve leaflets and showed how these patterns deteriorated during disease. Wall shear stress levels were calculated with higher stresses at ventricularis surfaces, and lower stresses on fibrosa surfaces for diseased valves.
CONCLUSIONS: We have developed a straightforward methodology to generate patient specific numerical models for aortic valve hemodynamics using a commercial numerical solver. We identified specific flow patterns and wall shear stress levels for normal and diseased cases implying significant contribution of hemodynamic stresses on disease progression. Our approach will be valuable and readily applicable for other researchers working on numerical modeling of aortic valve hemodynamics.


Back to 2017 Program