Session D4: Cardiovascular II

Volumetric velocimetry downstream of a percutaneous heart valve

Transcatheter aortic valve replacement has emerged as a safe and effective treatment for severe, symptomatic aortic stenosis in intermediate or greater surgical risk patients. However, despite excellent short-term outcomes, improved imaging and awareness has led to the identification of leaflet thrombosis on the aortic side of the prosthesis. Upon implantation, the transcatheter heart valve (THV) becomes enclosed in the native aortic valve leaflet tissue dividing the native sinus into two regions -- a smaller anatomical sinus and a neo-sinus. To understand the causes for thrombosis, plenoptic Particle Image Velocimetry (PIV) is used to investigate the pulsatile three-dimensional flow in the sinus and neo-sinus region of the THV. Experiments are conducted on both a real and a transparent THV model in a pulsatile flow loop capable of replicating physiological hemodynamics. Comparisons with planar PIV results demonstrate the feasibility of using Plenoptic PIV to study heart valve fluid dynamics. Large three-dimensional regions of low velocity magnitude and low viscous shear stress were observed near the heart valve which could increase particle residence time potentially leading to formation of clots the THV leaflet.   

An in vitro experimental study of flow past aortic valve under varied pulsatile conditions.

Flow past aortic valve represents a complex fluid-structure interaction phenomenon that involves pulsatile, vortical, and turbulent conditions. The flow characteristics immediately downstream of the valve, such as the variation of pulsatile flow velocity, formation of vortices, distribution of shear stresses, are of particular interest to further elucidate the role of hemodynamics in various aortic diseases. However, the fluid dynamics of a realistic aortic valve is not fully understood. Particularly, it is unclear how the flow fields downstream of the aortic valve would change under varied pulsatile inlet boundary conditions. In this study, an in vitro experiment has been conducted to investigate the flow fields downstream of a silicone aortic valve model within a cardiovascular flow simulator. Phased-locked Particle Image Velocimetry measurements were performed to map the velocity fields and Reynolds normal and shear stresses at different phases in a cardiac cycle. Temporal variations of pressure across the valve model were measured using high frequency transducers. Results have been compared for different pulsatile inlet conditions, including varied frequencies (heart rates), magnitudes (stroke volumes), and cardiac contractile functions (shapes of waveforms).   

Superhydrophobicity to minimize thrombogenic risk on mechanical heart valves

A large number of prosthetic heart valves are implanted each year to treat heart valve disease, where half of the surgically replaced valves are mechanical heart valves (MHV)s. MHVs are at high risk for thrombosis and therefore require lifelong antithrombotic therapies, causing an increased bleeding risk that can lead to death. To alleviate this need, we investigate the potential of superhydrophobic surfaces in reducing the thrombotic risk. Particle imaging velocimetry and computational fluid dynamics are used to quantify shear stress in the presence of potential slip on the surface. Coagulation and cell adhesion are quantified by incubating blood under static conditions. We further evaluate a dynamic blood response in polydimethylsiloxane channels under complex shear conditions that mimic the hinge region of bileaflet mechanical heart valves, a region known to exhibit thrombosis. Overall, Shear stress is not reduced on a superhydrophobic bileaflet MHV. However, superhydrophobic surfaces significantly reduce the potential for platelet responses under static and dynamic blood flow conditions, a counterintuitive result when considering that hydrophobic surfaces are prone to protein and cell adhesion.   

Experimental assessment of valve performance in healthy and diseased right ventricular outflow tracts using magnetic resonance velocimetry

Tetralogy of Fallot (ToF), the most common type of cyanotic congenital heart defect, affects 1 in every 2500 newborns annually and typically requires surgical repair of the right ventricular outflow tract (RVOT) and placement of an artificial pulmonary valve. All artificial valves are subject to dysfunction, but their longevity is highly variable. Clinical observation reveals large variations in RVOT anatomy in ToF patients, which may affect longevity. This work aims to experimentally assess the performance of artificial pulmonary valves in anatomically realistic healthy and diseased RVOT geometries using magnetic resonance velocimetry (MRV). With MRV, we can capture 3D, three-component, phase-averaged velocity fields in 3D printed RVOT geometries. The experiment is designed to ensure physiological flow rate and pressure waveforms, while the RVOT geometries are based on anatomies seen clinically in ToF patients. Two models are used in the current work: an idealized RVOT based on healthy subjects aged eleven to thirteen and a diseased geometry with a dilation of 150{\%} in vessel diameter downstream of the pulmonary valve. We will also present preliminary rigid-wall blood flow simulations in each model, towards the ultimate goal of experimental validation of valve simulations.   

Comparison of platelet activation through hinge vs bulk flow in mechanical heart valves

Bileaflet mechanical heart valves increase the risk of thrombus formation in patients which is believed to be initiated by platelet activation. Platelets can be activated by the elevated shear stresses in the bulk flow during the systole phase or the flow through the hinge during the diastole. However, the importance of platelet activation by the bulk flow vs the hinge in MHVs has yet to be studied. Here, we investigate the contribution of each of the above mechanisms to the activation of platelets in MHs by performing simulation of the flow through a 25mm St. Jude Medical valve placed in a straight aorta. Two different gap sizes (250 and 150 micrometer) are used in this study. The simulations are done using a sharp interface curvilinear immersed boundary method along with a strong-coupling algorithm for FSI solver on overset grids. The platelet activation through the hinge for different gap sizes is compared to the activation in the bulk flow using two platelet activation models to ensure the consistency of the results. Our results for all gap sizes using different activation models show that the integration of platelet activation caused by the bulk flow is several times higher in comparison to the activation through the hinge.   

Lagrangian coherent structures in the left ventricle in the presence of aortic valve regurgitation

Aortic valve regurgitation is a rather prevalent condition where the aortic valve improperly closes, allowing filling of the left ventricle of the heart to occur partly from backflow through the aortic valve. Although studies of intraventricular flow are rapidly gaining popularity in the fluid dynamics research community, much attention has been given to the left ventricular vortex and its potential for early detection of disease, particularly in the case of dilated cardiomyopathy. Notably, the subsequent flow in the left ventricle in the presence of aortic valve regurgitation ought to be appreciably disturbed and has yet to be described. Aortic valve regurgitation was simulated \textit{in vitro} in a double-activation left heart duplicator and the ensuing flow was captured using two-dimensional time-resolved particle image velocimetry. Further insight into the regurgitant flow is obtained by computing attracting and repelling Lagrangian coherent structures. An interesting interplay between the two inflowing jets and their shear layer roll-up is observed for various grades of regurgitation. This study highlights flow features which may find use in further assessing regurgitation severity.   

Flow-induced Flutter of Heart Valves: Experiments with Canonical Models

For the better understanding of hemodynamics associated with valvular function in health and disease, the flow-induced flutter of heart valve leaflets is studied using benchtop experiments with canonical valve models. A simple experimental model with flexible leaflets is constructed and a pulsatile pump drives the flow through the leaflets. We quantify the leaflet dynamics using digital image analysis and also characterize the dynamics of the flow around the leaflets using particle imaging velocimetry. Experiments are conducted over a wide range of flow and leaflet parameters and data curated for use as a benchmark for validation of computational fluid-structure interaction models.   

Flow-Induced Mitral Leaflet Motion in Hypertrophic Cardiomyopathy

Hypertrophic cardiomyopathy (HCM) is considered the cause of sudden cardiac death in developed countries. Clinically it is found to be related to the thickening of the intra-ventricular septum combined with elongated mitral leaflets. During systole the low pressure, induced by the abnormal velocities in the narrowed aortic channel, can attract one or both the mitral leaflets causing the aortic obstruction and sometimes instantaneous death. In this paper a fluid structure interaction model for the flow in the left ventricle with a native mitral valve is employed to investigate the physio-pathology of HCM. The problem is studied using direct numerical simulations of the Navier-Stokes equations with a two-way coupled structural solver based on interaction potential approach for the structure dynamics. Simulations are performed for two different degrees of hypertrophy, and two values of pumping efficiency. The leaflets dynamics and the ventricle deformation resulting from the echocardiography of patients affected by HCM are well captured by the simulations. Moreover, the procedures of leaflets plication and septum myectomy are simulated in order to get insights into the efficiency and reliability of such surgery.   

Vortex dynamics in Patient-Specific Stenotic Tricuspid and Bicuspid Aortic Valves pre- and post- Trans-catheter Aortic Valve Replacement

Understanding blood flow related adverse complications such as leaflet thrombosis post-transcatheter aortic valve implantation (TAVI) requires a deeper understanding of how patient-specific anatomic and hemodynamic factors, and relative valve positioning dictate sinus vortex flow and stasis regions. High resolution time-resolved particle image velocimetry measurements were conducted in compliant and transparent 3D printed patient-specific models of stenotic bicuspid and tricuspid aortic valve roots from patients who underwent TAVI. Using Lagrangian particle tracking analysis of sinus vortex flows and probability distributions of residence time and blood damage indices we show that (a) patient specific modeling provides a more realistic assessment of TAVI flows, (b) TAVI deployment alters sinus flow patterns by significantly decreasing sinus velocity and vorticity, and (c) relative valve positioning can control critical vortex structures that may explain preferential leaflet thrombosis corresponding to separated flow recirculation, secondary to valve jet vectoring relative to the aorta axis. This work provides new methods and understanding of the spatio-temporal aortic sinus vortex dynamics in post TAVI pathology.   

Multi-physics 3D computational study of leaflet thrombus formation following surgical and transcatheter aortic valve replacement

An increasingly recognized complication following surgical/transcatheter aortic valve replacement is thrombosis or blood clot formation on replacement valve leaflets. A predisposing factor in thrombus formation on biomaterial surfaces of a bioprosthetic heart valve is blood stasis. Longer residence time of blood provides an opportunity for platelets and agonists to accumulate to critical concentrations that leads to platelet activation and then thrombosis. In this study, we have developed a fluid-solid interaction (FSI) modeling approach, to quantify blood stasis on the leaflets of bioprosthetic aortic valves with different design operating in a patient-specific geometry. We have validated our FSI model against experimental measurements of valve opening/closing as well as in-vitro particle image velocimetry. We have also embedded in our method a model for transport of platelets and agonists (ADP, TxA2, and thrombin) and their interactions that result in platelets activation and adhesion to biomaterial bioprosthetic surfaces. We have provided quantitative evidence for the correlation between long residence of blood on bioprosthetic aortic valve leaflets and formation of high thrombogenicity risk regions on the leaflets that are characterized by accumulation of activated platelet.