FSI in Virtual Heart Surgery
Considering surgery, with the rapid advances in the fields of medical imaging, image reconstruction, grid generation and multi-physics computational tools, we are now quite close to construct high-fidelity, patient-specific computational models to devise the best surgical procedure for each individual. These advances can also help researchers to reach a better understanding of the causes of different diseases and potential remedies. ADINA’s powerful multiphysics capabilities play a major role in this overall endeavor.
In this News, we highlight a study performed by researchers using ADINA in which they investigated a computational approach to construct patient-specific ventricle models based on in vivo cardiac magnetic resonance (CMR) imaging for surgical planning and optimization. The study was motivated by the fact that the use of such computational models can help to assess different surgical hypotheses and replace empirical and perhaps risky clinical experimentation (see Ref.).
Figure 1 shows the reconstructed 3D geometry of the left and right ventricles (LV/RV), the computational grid and the location of the patch caused by surgery.
Figure 1 Reconstructed RV/LV geometry and computational grid
The fluid was assumed to be laminar, Newtonian, viscous and incompressible. The Navier-Stokes equations with ALE were used as the governing equations.
The right and left ventricles were assumed to behave as a hyperelastic, anisotropic, nearly incompressible material and were modeled using the Mooney-Rivlin material model. The anisotropic effect due to the presence of fibers in the ventricles was included in the model using the Holzapfel’s modified energy density function available in ADINA. Of course, the fiber orientations can be adjusted for different patients. All the material parameters were calibrated using the available experimental patient-specific data (see Ref.).
Heart expansion/contraction can be considered as a combination of passive elastic expansion/contraction caused by blood pressure and active contraction/relaxation caused by fiber stiffening/relaxation. In this study a time-dependent material stiffening in the fiber direction was used to model the active contraction.
To model the surgical procedure, the RV morphology was modified by trimming the scar and using different patches, see Figure 2.
In all cases the model predicted RV volume was compared with the CMR-measured data.
The coupled fluid-solid model was solved using direct FSI coupling since the response of the structure and the response of the fluid are strongly coupled. Large deformations and large strains of the RV/LV were taken into account. The ALE approach
was used to accommodate the large motions of the RV/LV. The time history response of the coupled fluid-solid system was calculated using the direct step-by-step time integration.
Figure 2 Schematic of the surgical patch and the corresponding computational models
The animation at the top of the page shows the contour plot of the 1st
principal strain in the RV/LV. The animation below shows the velocity vector field in the fluid inside the RV during the cardiac cycle.
For more information on ADINA FSI, refer to our fluid-structure interaction page where we have provided an overview of the features, many case studies and a list of more than 160 archival publications in which researchers have used ADINA FSI for solving a wide range of challenging fluid-structure interaction problems.
- D. Tang, C. Yang, T. Geva, and P.J. del Nido, "Image-based patient-specific ventricle models with fluid-structure interaction for cardiac function assessment and surgical design optimization", Progress in Pediatric Cardiology, 30:51–62, 2010
Virtual surgery, patient-specific, fluid structure interaction, large deformation, strongly coupled, direct FSI, hyperelastic, anisotropic, computer-aided surgical design, MRI, ventricle, Mooney-Rivlin, Holzapfel, pulmonary valve replacement procedure
Courtesy of D. Tang (Worcester Polytechnic Institute, USA), C. Yang (Worcester Polytechnic Institute, USA and Beijing Normal University, China), T. Geva (Children's Hospital Boston and Harvard Medical School, USA), and P. J. del Nido (Children's Hospital Boston and Harvard Medical School, USA)