Article
Computer Science, Interdisciplinary Applications
Dongwei Ye, Valeria Krzhizhanovskaya, Alfons G. Hoekstra
Summary: This work presents a data-driven surrogate model for efficient prediction of blood flow simulations on similar but distinct domains. The proposed model utilizes group surface registration to parameterize shapes and uses geometry information for hemodynamics prediction. The results demonstrate that the surrogate model has accuracy and efficiency in hemodynamics prediction and can be applied to real-time simulation and uncertainty quantification for complex patient-specific scenarios.
JOURNAL OF COMPUTATIONAL PHYSICS
(2024)
Article
Computer Science, Information Systems
Eirini Kardampiki, Emanuele Vignali, Dorela Haxhiademi, Duccio Federici, Edoardo Ferrante, Stefano Porziani, Andrea Chiappa, Corrado Groth, Margherita Cioffi, Marco Evangelos Biancolini, Emiliano Costa, Simona Celi
Summary: A Medical Digital Twin pipeline based on reduced order modeling is presented for fast and interactive evaluation of the hemodynamic parameters of Modified Blalock Taussig Shunt (MBTS). The study demonstrates the potential of the proposed approach for real-time investigation of the effects of different MBTS morphologies on hemodynamic features.
Article
Thermodynamics
Mingyu Yang, Seongyoon Kim, Xiang Sun, Sanghyun Kim, Jiyong Choi, Tae Seon Park, Jung-Il Choi
Summary: This study optimized the operating conditions of a recuperative burner system using computational fluid dynamics (CFD) and reduced-order deep learning technique, resulting in improved performance and reduced environmental pollution. The study employed CFD simulation and a genetic algorithm to find the optimal conditions for various objectives.
APPLIED THERMAL ENGINEERING
(2024)
Article
Engineering, Marine
Zhe Sun, Lu-yu Sun, Li-xin Xu, Yu-long Hu, Gui-yong Zhang, Zhi Zong
Summary: A CFD-based data-driven reduced order modeling method was proposed for studying damaged ship motion in waves. It involved low-order modeling of the entire parameter range and high-order modeling for selected key scenarios. The difference between the low and high-order results was then modeled using machine learning or data regression methods. The proposed model showed comparable accuracy to pure high-order modeling while significantly reducing computational time for the studied cases.
JOURNAL OF MARINE SCIENCE AND ENGINEERING
(2023)
Article
Computer Science, Interdisciplinary Applications
Elizabeth H. Krath, Forrest L. Carpenter, Paul G. A. Cizmas, David A. Johnston
Summary: This study introduces a novel, more efficient reduced-order model for compressible flows based on proper orthogonal decomposition (POD). By using specific volume instead of density, the coefficients of the system of ODEs in the reduced-order model were pre-computed. Various methods were used to enhance ODE solver stability. Validation was done for two cases, showing a speedup exceeding four orders of magnitude compared to the full-order model.
JOURNAL OF COMPUTATIONAL PHYSICS
(2021)
Article
Computer Science, Information Systems
Pavlo Yevtushenko, Leonid Goubergrits, Lina Gundelwein, Arnaud Setio, Heiko Ramm, Heiko Lamecker, Tobias Heimann, Alexander Meyer, Titus Kuehne, Marie Schafstedde
Summary: Image-based patient-specific modelling of hemodynamics has the potential to improve diagnostic capabilities and clinical outcomes for cardiovascular diseases, but traditional numerical methods require significant computational resources. Machine learning approaches offer the advantage of quickly calculating patient-specific hemodynamic outcomes with high accuracy, showcasing their ability to perform tasks that previously required resource-intensive simulations.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2022)
Article
Engineering, Multidisciplinary
Emmanuel Menier, Michele Alessandro Bucci, Mouadh Yagoubi, Lionel Mathelin, Marc Schoenauer
Summary: This paper proposes a deep learning based closure modeling approach for improving classical POD-Galerkin reduced order models (ROM), which uses neural networks to approximate well studied operators. The approach is based on an interpretable continuous memory formulation, resulting in corrected models that can be simulated using classical time stepping schemes. The capabilities of the approach are demonstrated on two classical examples from Computational Fluid Dynamics and a parametric case.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2023)
Article
Computer Science, Interdisciplinary Applications
A. Garbo, P. Bekemeyer
Summary: A physics-based reduced order model is proposed to accurately predict unsteady flowfield solutions at a fraction of the computational time required by high fidelity computational fluid dynamics solvers. The results show that the proposed technique accurately predicts unsteady solutions with a computational cost that is around 10% of the time needed for equivalent high fidelity simulations, substantiating its viability in applications with constrained computational budgets.
COMPUTERS & FLUIDS
(2022)
Article
Engineering, Multidisciplinary
Luca Boscaglia, Aldo Boglietti, Shafigh Nategh, Fabio Bonsanto, Claudio Scema
Summary: This article introduces a numerical approach using reduced-order modeling to analyze the thermal behavior of electric traction motors, and the results show good agreement between measured and estimated values, indicating high accuracy of the reduced-order model in predicting motor thermal performance.
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS
(2021)
Article
Engineering, Biomedical
Nhien Tran-Nguyen, Francesca Condemi, Andrew Yan, Stephen Fremes, Piero Triverio, Laura Jimenez-Juan
Summary: This prospective study used computational fluid dynamics to investigate how graft hemodynamics may vary among different types of grafts one month after coronary artery bypass graft (CABG) surgery. The study found differences in abnormal wall shear stress area among different types of grafts, suggesting a possible influence on long-term graft failure rates.
ANNALS OF BIOMEDICAL ENGINEERING
(2022)
Article
Nuclear Science & Technology
Huilun Kang, Zhaofei Tian, Guangliang Chen, Lei Li, Tianhui Chu
Summary: This study proposes a multi-fidelity reduced-order model (MF-ROM) approach to address the high cost of performing high-fidelity computational fluid dynamics (HF-CFD) for predicting the flow and heat transfer state of coolant in a reactor core. The MF-ROM utilizes proper orthogonal decomposition (POD) to extract basis vectors and coefficients from both high-fidelity and low-fidelity CFD results, and trains a surrogate model to map the relationship between the extracted coefficients. The effectiveness of the MF-ROM is evaluated on a flow and heat transfer problem in PWR fuel rod bundles, and the results show good agreements between the MF-ROM and high-fidelity CFD simulation. The proposed MF-ROM offers a computationally efficient alternative for complex simulations.
NUCLEAR ENGINEERING AND TECHNOLOGY
(2022)
Article
Engineering, Aerospace
Massoud Tatar
Summary: This study presents a novel global reduced-order modeling and parameter estimation of a maneuvering aircraft using radial basis functions. A computational fluid dynamics approach accurately predicts the flow field, a neural network constructs a nonlinear aerodynamic model, and stability derivatives are analyzed for their dependency on reduced frequency and angle of attack, as well as the influence of angle of attack on moment coefficients.
JOURNAL OF AEROSPACE ENGINEERING
(2021)
Article
Engineering, Chemical
Jacob Johnston, Sarah M. Dischinger, Mostafa Nassr, Ji Yeon Lee, Pedram Bigdelou, Benny D. Freeman, Kristofer L. Gleason, Denis Martinand, Daniel J. Miller, Sergi Molins, Nicolas Spycher, William T. Stringfellow, Nils Tilton
Summary: Feed spacers in reverse osmosis systems limit computational fluid dynamics (CFD) simulations, but a reduced model with an analytical approach provides a 10,000-fold speedup and accurately reproduces CFD predictions for solute transport in membrane fouling phenomena. This model serves as a simple testbed for studying multispecies transport and membrane fouling for which simulating spacers is often impractical.
JOURNAL OF MEMBRANE SCIENCE
(2023)
Article
Mathematics, Applied
Francesco Romor, Giovanni Stabile, Gianluigi Rozza
Summary: Non-affine parametric dependencies, nonlinearities, and advection-dominated regimes can hinder the development of efficient reduced-order models based on linear subspace approximations. Data-driven methods utilizing autoencoders and their variants have shown promise, but there is a need for increased interpretability, especially in regions with limited data and outside the training range. Additionally, exploiting knowledge of the model's physics during the predictive phase is important. To address these challenges, we implement the non-linear manifold method introduced by Lee and Carlberg (J Comput Phys 404:108973, 2020) and combine it with hyper-reduction achieved through reduced over-collocation and teacher-student training of a reduced decoder. We evaluate the methodology on a 2D non-linear conservation law and a 2D shallow water model, comparing it to a purely data-driven approach using a long-short term memory network for time evolution.
JOURNAL OF SCIENTIFIC COMPUTING
(2023)
Article
Radiology, Nuclear Medicine & Medical Imaging
Allison Shields, Kyle Williams, Mohammad Mahdi Shiraz Bhurwani, Swetadri Vasan Setlur Nagesh, Venkat Keshav Chivukula, Daniel R. Bednarek, Stephen Rudin, Jason Davies, Adnan H. Siddiqui, Ciprian N. Ionita
Summary: This study explores the use of a pathlength-correction metric to convert 3D contrast flow to projected contrast flow in order to overcome the limitations of 2D angiographic parametric imaging (API). The results show that pathlength correction significantly improves the accuracy of API parameters and provides a more accurate representation of contrast distribution within each aneurysm.
Article
Computer Science, Interdisciplinary Applications
Adam Updegrove, Nathan M. Wilson, Shawn C. Shadden
ADVANCES IN ENGINEERING SOFTWARE
(2016)
Article
Oceanography
Siavash Ameli, Shawn C. Shadden
JOURNAL OF GEOPHYSICAL RESEARCH-OCEANS
(2019)
Article
Engineering, Biomedical
Kirk B. Hansen, Shawn C. Shadden
INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN BIOMEDICAL ENGINEERING
(2019)
Article
Computer Science, Interdisciplinary Applications
Fanwei Kong, Shawn C. Shadden
Summary: Patient-specific cardiac modeling using deep learning methods can efficiently generate accurate and consistent simulation-suitable models of the heart from medical images. This approach outperforms prior methods in terms of whole heart reconstruction and produces geometries that better satisfy requirements for cardiac flow simulations. The source code and pretrained networks for this method are publicly available for further development and application.
IEEE TRANSACTIONS ON MEDICAL IMAGING
(2023)
Article
Computer Science, Theory & Methods
Siavash Ameli, Shawn C. Shadden
Summary: In this paper, heuristic interpolation methods are developed for calculating two specific functions. By modifying sharp bounds, accurate computation is achieved. Experimental results validate the accuracy and performance of the proposed method.
STATISTICS AND COMPUTING
(2022)
Article
Mathematics, Applied
Siavash Ameli, Shawn C. Shadden
Summary: This paper studies a matrix derived from a singular form of the Woodbury matrix identity. Generalized inverse and pseudo-determinant identities for this matrix are presented, with direct applications for Gaussian process regression, especially in likelihood representation and precision matrix. The definition of precision matrix is extended to the Bott-Duffin inverse of the covariance matrix, preserving properties related to conditional independence, conditional precision, and marginal precision. An efficient algorithm and numerical analysis for the presented determinant identities are provided, demonstrating their advantages in computing log-determinant terms in likelihood functions of Gaussian process regression.
APPLIED MATHEMATICS AND COMPUTATION
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Fanwei Kong, Shawn C. Shadden
Summary: The study introduces a novel deep learning approach to reconstruct simulation-ready whole heart meshes from volumetric image data, aiming to efficiently create meshes for computational fluid dynamics simulations of cardiac flow.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION - MICCAI 2021, PT IV
(2021)
Article
Computer Science, Software Engineering
Miguel A. Rodriguez, Christoph M. Augustin, Shawn C. Shadden
Article
Clinical Neurology
Jaiyoung Ryu, Nerissa Ko, Xiao Hu, Shawn C. Shadden
CEREBROVASCULAR DISEASES
(2017)