Pediatric Cardiology
In collaboration with Dr. Piers Barker, Duke University, we are studying the hemodynamics in patients who undergo the Norwood procedure. This procedure is performed within a week of birth, and is generally considered among the most complex neonatal cardiac surgeries. Our long-term goal with this project is to improve the clinical outcome for patients with hypoplastic left heart syndrome (HLHS) by optimizing the balance between pulmonary and systemic blood flow.
Adult Cardiology
We are interested in understanding how risk factors like endothelial shear stress (ESS) trigger complex biomechanical events leading to atherosclerotic pathologies. Difficulties of measuring ESS in vivo made these simulations crucial for unraveling and predicting cardiovascular disease progression. Our lab is applying computational, algorithmic, and physics-based advances to model blood flow in large regions of patient-specific geometries over time. In collaboration with Dr. James Chen, University of Colorado Denver, Dr. Andrew Kahn, UCSD, Dr. Jane Leopold, Brigham and Women's Hospital, we are investigating the role of wall shear stress in the development and progression of cardiovascular disease.
Treatment Planning
Computational models are a potentially powerful tool for treatment planning. For each of the above-mentioned projects, we are determining how personalized hemodynamic models can effectively support clinical decision-making. We recently conducted a feasibility study in which post-operative vascular candidates were tested under both resting and exercise conditions. Our objective is to use simulation to study the impact different treatment plans would have on key hemodynamic properties or risk factors and provide the capability of stratifying potential patient risk before treatment takes place.
Validation
We are working with the Frakes Lab at ASU to validate our simulations through comparison with experimental results for a physiologically relevant flow in a vascular geometry. 3D-printing allows us to use the same patient-specific geometry for both the computational simulation and the in vitro experiment. An example looking at flow in a coarcted aorta is shown to the left. The model was connected to the flow loop with flexible Tygon tubing and placed on a custom stage to allow a laser light sheet to pass vertically through the model centerplane for optical imaging. The fluid flowing through the loop was a sodium iodide-based solution with a refractive index matched to the urethane block, to render the urethane wall invisible and eliminate optical distortion during particle image velocimetry (PIV).
Velocity fields on a plane passing through the center of the model were captured using PIV. A LaVision 3D Flowmaster PIV system was used (LaVision, Ypsilanti, MI, USA). The system operated in two frame cross-correlation mode, and standard stereo PIV procedures were followed to obtain low-noise measurements of particle displacement and velocity with a high percentage (>95%) of valid vectors. Acquisition for steady flow comprised of a minimum of three trials resulting in a minimum of 200 image pairs.
A constant flow rate was imposed at the inlet, and a zero pressure boundary condition governed the outlets. The magnitude of the velocity through the coarctation was recorded; corresponding slices from in vitro and HARVEY, our in-house developed computational fluid dynamics code, are shown in figure 2. With a RMSD=0.042 between average velocity magnitudes in simulation and experiment (figure 2, bottom), HARVEY successfully resolves the experimental flow pattern both coming from the aortic arch and through the CoA, in qualitative and quantitative respects.
