Bulletin of the American Physical Society
75th Annual Meeting of the Division of Fluid Dynamics
Volume 67, Number 19
Sunday–Tuesday, November 20–22, 2022; Indiana Convention Center, Indianapolis, Indiana.
Session Q06: Cardiovascular Flows: Methods |
Hide Abstracts |
Chair: Jeffery Tithof, U Minnesota Room: 133 |
Monday, November 21, 2022 1:25PM - 1:38PM |
Q06.00001: Molecular Tagging Velocimetry (MTV) to Measure Wall Shear Stress in Model Cardiovascular Flows Kartik V Bulusu, Charles Fort, Anton Yanovich, Philippe M Bardet, Michael W Plesniak WSS measurements in cardiovascular flows have been challenging when using traditional non-intrusive and tracer particle-based optical diagnostic techniques. The goal of this study is to perform high-fidelity, experimental wall shear stress (WSS) measurements in straight, rigid, cylindrical tubes representing canonical artery phanta with Newtonian blood analog fluids. Molecular Tagging Velocimetry (MTV) is used to determine near-wall velocity gradients under physiological flow pertinent to the cardiovasculature. The traditional limitations of spatio-temporal resolution of MTV techniques for liquids are overcome by a novel photobleaching approach with Talbot-effect structured illumination using a pair of common neodymium-doped pulsed lasers (355 and 527 nm). The results are benchmarked using Womersley’s analytical solution for pulsatile flow in a tube, wherein an analytical expression for near-wall, pulsatile drag force was obtained and the dimensionless Womersley number was introduced. This exploratory study aims to evaluate the utility of MTV techniques for extensive, experimental WSS-data analysis pertinent to in vitro models of cardiovascular flows. |
Monday, November 21, 2022 1:38PM - 1:51PM |
Q06.00002: Quantifying particle residence time and blood damage using time-resolved 3D particle tracking Huang Chen, Roya Kamali, Satheesh Kumar Harikrishnan, Thangam Natarajan, Lakshmi P Dasi Quantifying particle residence time and blood damage is essential in developing prosthetic heart valves and other blood-contacting medical devices. Given the high cost of the in vitro blood loop experiments, past studies have mainly relied on computational fluid dynamics (CFD) with hemolysis models for such tasks. However, CFD simulation is challenging to perform when resolving the complex fluid-structure-interaction problem is required. In this study, a method to quantify blood damage experimentally using 3D time-resolved particle tracking is developed, and the results are compared with CFD simulations. Experiments are performed in a fully refractive index-matched setup, and particle tracks are obtained using the shake-the-box algorithm. Direct quantifications of the particle residence time and blood damage only become possible after reconnecting the broken tracks with an extension algorithm that extrapolates the particle path forward and backward in time. The blood damage of an individual particle is calculated by a numerical hemolysis model based on the shear stress and exposure time. Preliminary results show that 3D particle tracking can be a very useful and accurate tool to quantify blood damage, especially when CFD simulation is difficult to perform. |
Monday, November 21, 2022 1:51PM - 2:04PM |
Q06.00003: Reducing the computational burden of coagulation cascade models in cardiovascular simulations Manuel Guerrero-Hurtado, Manuel Garcia-Villalba, Alejandro Gonzalo, Clarissa Bargellini, Pablo Martínez-Legazpi, Andrew M Kahn, Javier Bermejo, Juan Carlos del Alamo, Oscar Flores Thrombosis is a complex process that begins with the coagulation cascade, a series of biochemical reactions involving more than 40 species. Simulating the coagulation cascade requires solving tens of 3D unsteady advection-reaction-diffusion (ADR) equations, which is challenging. Here we present a novel approach to drastically reduce the computational burden of these simulations. Based on the observation that diffusive transport in arteries is low compared to advection, we model the Lagrangian evolution of the coagulation cascade following each fluid particle using the blood residence time as the independent variable. The ADR eqs. are then reduced to a system of ODEs. We test this methodology in an idealized aneurysm (a pulsating 2D cavity flow), using a simple coagulation model with three biochemical species (thrombin, factor XIa, and protein Ca). We show that the reduced model is cost-effective, accurately reproducing the spatio-temporal development of the coagulation cascade in the ADR system for up to ~15 cardiac cycles. |
Monday, November 21, 2022 2:04PM - 2:17PM |
Q06.00004: Vascular velocity reconstruction using Color Doppler flow field velocity Reza Babakhani Galangashi, Brett A. Meyers, Sayantan Bhattacharya, Javad Eshraghi, Pavlos Vlachos Understanding the complex hemodynamics within the vasculature resulting from blood flow patterns is of great importance and has been used clinically and in research applications. Color Doppler imaging (CDI) has been used to assess these hemodynamic patterns by producing 2D velocity fields of the flow within the blood vessels. However, CDI measures only one velocity component and in order to resolve the flow patterns, a flow field reconstruction is needed. Current methods to reconstruct velocity components from CDI have high errors, due to irrotational flow assumptions, oversimplification of the wall and bulk fluid motion effects on local velocity estimates, or non-realistic and invalid boundary conditions. Besides, Doppler velocity reconstruction method known as DoVeR, which uses the relation between stream function and vorticity of the flow to increase the robustness of velocity vector field reconstruction, has been introduced for LV flow reconstruction and fails to work for vascular flows. Therefore, an improved DoVeR method is introduced here to reconstruct velocity fields from Color Doppler images in the vasculature. In this work, synthetic data is used to validate the algorithm. Moreover, steady and pulsatile flow conditions with various Reynolds numbers are used to assess DoVeR reliability in reconstructing velocity vectors experimentally, and the DoVeR reconstructed velocity vectors were compared with Particle Image Velocimetry (PIV) data in a round tube mimicking the vasculature. Error analysis (EA) will be reported for the DoVeR method, and we will compare the results of the EA with other methods to determine robustness and reliability for accurately measuring hemodynamic parameters in the vasculature. |
Monday, November 21, 2022 2:17PM - 2:30PM |
Q06.00005: Geometric Deep Neural Differentiable Modeling for Efficient 3-D Patient-Specific Aortic Flow Simulations Pan Du, Jian-Xun Wang Computational modeling of hemodynamics plays an increasing role in the diagnosis and treatment planning of cardiovascular diseases. However, traditional CFD-based patient-specific models suffer from high computational costs and large modeling uncertainties. With the rapid development in AI and GPU computing, there has been a growing interest in developing deep learning (DL)-based surrogate models due to their higher efficiency and scalability. Nonetheless, most existing DL models are subject to approximation error, poor generalizability, and high dependency on training labels. To address these issues, we propose to integrate the governing physics into the learning structures by leveraging both advantages of numerical solvers and DL techniques within a differentiable programming framework. Specifically, a hybrid neural differentiable model based on graph neural networks and differentiable CFD solvers is developed for fast predictions of transient aortic flows given various input flow boundary conditions. Multi-resolution data and physics will be integrated for optimized efficiency and accuracy. The proposed model is compared with state-of-the-art DL-based surrogate models and pure CFD models, showing significant superiority in terms of efficiency, accuracy, and generalizability. |
Monday, November 21, 2022 2:30PM - 2:43PM |
Q06.00006: A scalable Euler-Lagrange strategy for particle-laden anatomical flows in subject-specific geometries Abhilash Reddy Malipeddi, C. Alberto Figueroa, Jesse Capecelatro We develop an Euler-Lagrange method for large scale particle-laden biological flows in subject-specific geometries. Towards that end, the CRIMSON cardiovascular flow modeling framework is extended to include Lagrangian particles that interact with each other and are coupled with the fluid. CRIMSON solves the incompressible Navier-Stokes equations using a stabilized finite-element method with equal order interpolation for velocity and pressure on unstructured grids. The complex morphology of the vasculature presents unique challenges in implementing scalable particle tracking and collision algorithms. Due to the significant unoccupied space in a bounding box, purely Cartesian-based particle collision acceleration schemes are not memory efficient. Here, we propose an efficient hash-table based cell-list to accelerate particle collision detection. Additionally, efficient procedures to initialize and inject non-overlapping particles in arbitrary geometries are developed. The initialization procedure uses a fast-marching level set method to generate a signed distance field to accelerate particle seeding. Poisson disc sampling with linear complexity is employed to initialize the particles. |
Monday, November 21, 2022 2:43PM - 2:56PM |
Q06.00007: Blood flow predictions in data-poor regimes: A physics-informed Bayesian approach Shaghayegh Zamani Ashtiani, Mohammad Sarabian, Kaveh Laksari, Hessam Babaee Computation modeling blood flow properties can aid diagnosis and treatment of cardiovascular and cerebrovascular diseases. However, high-fidelity predictions are computationally expansive, and blood flow measurement via transcranial Doppler ultrasound or imaging alone lack the sufficient resolution to be used directly or else used to train a machine learning surrogate model. Such limitations make it vital to develop a computationally inexpensive model that provides prediction based on sparse computational/clinical data. We present a physics-informed Gaussian process regression technique to predict the blood flow properties from a very few sparse measurements. The presented algorithm is computationally inexpensive, and it has the potential to be used in clinical settings. We demonstrate our methodology on examples such as a Y-shaped bifurcation, abdominal aorta, and brain vasculature. |
Monday, November 21, 2022 2:56PM - 3:09PM |
Q06.00008: Bayesian Intraventricular Vector Flow Mapping (BVFM) Cathleen M Nguyen, Bahetihazi Maidu, Darrin J Wong, Sachiyo Igata, Christian Chazo Paz, Pablo Martinez-Legazpi, Javier Bermejo, Andrew M Kahn, Anthony DeMaria, Juan Carlos del Alamo Intracardiac blood flow analysis is quickly growing and allows for assessing flow dynamics, vortex formation, cardiac dysfunction, energetic efficiency, and other quantitative biomarkers of cardiovascular health. However, the lack of uncertainty quantification in medical imaging modalities makes it challenging to translate intracardiac blood flow analysis to guide medical decisions. Additionally, current methods used to quantify blood flow in the heart are limited due to assumptions (e.g., flow planarity), and post-processing algorithms rely on input data (e.g., wall segmentations) for boundary conditions that are also prone to uncertainty. Here we present a general flow mapping method rooted in Bayesian inference that allows us to fuse data from different imaging modalities and propagate their respective uncertainties to reconstruct intracardiac flow fields. We use an echocardiographic simulator (MATLAB Ultrasound Toolbox and SIMUS) to explore image quality and the parameter space to test our Bayesian framework. We apply our general framework to synthetic ultrasound data and two different clinical cases: 1) echo-PIV and echo-Doppler fusion and 2) Doppler multiscale fusion. |
Monday, November 21, 2022 3:09PM - 3:22PM |
Q06.00009: Machine learning for predicting microvascular network hemodynamics Saman Ebrahimi, Prosenjit Bagchi We investigate the applicability of the machine learning (ML) techniques for the prediction of time-averaged and time-dependent blood flow rate and red cell distributions in physiologically realistic and large capillary vessel networks. To train and test the ML models, we acquire data from high-fidelity simulations of the flow of deformable red blood cells suspension in the networks. We use two networks that are geometrically quite different, with one used for the model building and the other for prediction. For the prediction of time-averaged blood flow rate, a regression-type ANN model is used, while for the prediction of time-average RBC distribution, a classification-type ANN model is used. With the flow rate and hematocrit specified at an inlet of a vasculature, the models predict time-averaged flow rate and RBC distributions in the entire network. For the prediction of time-dependent quantities, we use LSTM models. With these, we first predict the time-dependent flow rate and hematocrit in each vessel in isolation over a long time, as well as such quantities in isolated vascular bifurcations. These models constitute a model bank which is then used to predict simultaneous spatially and temporally evolving quantities through the vessel hierarchy in the networks. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700