Efficient Reduced Order Modeling Based on HODMD to Predict Intraventricular Flow Dynamics Journal: Journal of L a T e X Templates

E. Lazpita, J. Garicano-Mena, S. Le Clainche · 2025

Read Efficient Reduced Order Modeling Based on HODMD to Predict Intraventricular Flow Dynamics Journal: Journal of L a T e X Templates by E. Lazpita, J. Garicano-Mena, S. Le Clainche (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-21716.txt.

Introduction

Eneko Lazpita, Jesus Garicano-Mena, and Soledad Le Clainche presented a fast model in two thousand twenty-five for blood flow inside the heart's left chamber. Full computer models of this flow are exact but slow and costly. Their model uses a method called HODMD, which keeps only the main repeating patterns in the flow.

The model was tested on two simple chamber shapes with distinct flow styles. It rebuilt full heart cycles well and forecast flow past its training window, with errors below five and ten percent for the two shapes. It ran at least a hundred thousand times faster than full models, opening the way to quick clinical tools.

Catalog record

CREATORE. Lazpita, J. Garicano-Mena, S. Le Clainche
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE48 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-21716

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Efficient Reduced Order Modeling Based on HODMD to Predict Intraventricular Flow Dynamics Journal: Journal of L a T e X Templates E. Lazpita, J. Garicano-Mena, S. Le Clainche Abstract Accurate and efficient modeling of cardiac blood flow is crucial for advancing data-driven tools in cardiovascular research and clinical applications. Recently, the accuracy and availability of computational fluid dynamics (CFD) methodologies for simulating intraventricular flow have increased. However, these methods remain complex and computationally costly. This study presents a reduced order model (ROM) based on higher order dynamic mode decomposition (HODMD). The proposed approach enables accurate reconstruction and long term prediction of left ventricle flow fields. The method is tested on two idealized ventricular geometries exhibiting distinct flow regimes to assess its robustness under different hemodynamic conditions. By leveraging a small number of training snapshots and focusing on the dominant periodic components representing the physics of the system, the HODMD-based model accurately reconstructs the flow field over entire cardiac cycles and provides reliable long-term predictions beyond the training window. The reconstruction and prediction errors remain below 5% for the first geometry and below 10% for the second, even when using as few as the first 3 cycles of simulated data, representing the transitory regime. Additionally, the approach reduces computational costs with a speed-up factor of at least 10 5 10^{5} compared to full-order simulations, enabling fast surrogate modeling of complex cardiac flows. These results highlight the potential of spectrally-constrained HODMD as a robust and interpretable ROM for simulating intraventricular hemodynamics. This approach shows promise for integration in real-time analysis and patient specific models. 1 Introduction Cardiovascular diseases remain the leading cause of mortality worldwide, with heart failure representing a major source of morbidity and a significant reduction in quality of life [ 1 ] . In this context, understanding and predicting intraventricular blood flow dynamics is essential to improving diagnosis, prognosis, and treatment planning in patients with cardiac dysfunction (see Ref. [ 2 , 3 ] ). The left ventricle (LV), as the main pumping chamber of the heart, plays a central role in cardiac output, and its hemodynamics are intimately linked to both structural integrity and functional performance [ 4 , 5 ] . Computational fluid dynamics (CFD) has emerged as a powerful tool for investigating intraventricular flows, complementing in vivo measurements by enabling detailed, controllable exploration of flow patterns. Numerous studies have applied CFD to both idealized and patient-specific LV geometries, providing valuable insights into vortex dynamics, energy dissipation, and wall shear stresses. For instance, Spandan et al. [ 6 ] introduced a parallel interaction potential approach coupled with the immersed boundary method for fully resolved fluid–structure interaction simulations of deformable cardiac structures. This methodology has been extended to simulate healthy and pathological ventricles with natural and prosthetic mitral valves [ 7 ] , as well as to investigate systolic anterior motion in hypertrophic cardiomyopathy [ 8 ] . Domenichini et al. [ 9 ] conducted early numerical studies on the three-dimensional inflow into a ventricular-shaped cavity. They later complemented this with experimental validation [ 10 ] , providing deeper insights into the mechanisms driving intraventricular flow. More recently, they explored the interaction between myocardial motion and fluid dynamics, demonstrating how alterations in tissue displacement are reflected in hemodynamic force distributions [ 11 ] . Quarteroni and collaborators have developed advanced frameworks for cardiac modeling that combine electromechanics, fluid dynamics, and patient-specific data [ 12 , 13 ] . These contributions reinforce the value of CFD as a predictive tool in cardiac research. These works represent only a small sample of the growing body of literature. A comprehensive overview of CFD-based cardiac flow simulations, including their application to vortex ring (VR) analysis, is provided in the review by Mahesh et al. [ 14 ] . The VR forms during early diastolic filling and plays a critical role in the efficient reorientation of the blood influx towards the aortic valve throughout the cardiac cycle. Its formation and stability are closely tied to ventricular geometry, wall motion, and pathological changes, making it a relevant biomarker for cardiovascular health. In a recent study (see Ref. [ 15 ] ), we analyzed the physical mechanisms underlying VR dynamics in two idealized LV geometries, identifying the VR as the dominant flow structure driving the system’s temporal evolution and highlighting its potential as a diagnostic indicator. Despite these advancements, generating high-fidelity CFD simulations of cardiac flows remains computationally demanding. The need to account for large deformations, moving boundaries, and physiologically realistic conditions results in considerable computational cost. Moreover, many CFD solvers suffer from numerical instability when applied to complex ventricular geometries or pathological flow regimes, often requiring extensive tuning or simplification of the physical model. These challenges restrict the scope of CFD applications in both clinical and research contexts. As a result, many studies are constrained to a limited number of cardiac cycles, often fewer than five, impeding the statistical robustness and generalizability of the resulting data [ 16 , 17 ] . This limitation poses a significant barrier to the development of data-driven models, which require temporally rich datasets to achieve reliable performance across varying cardiac conditions. In order to address these challenges, a novel reduced order model (ROM) is proposed. This model is designed to accelerate the obtention of intraventricular flow databases as an alternative to high-fidelity CFD simulations. Although ROMs are widely used in fluid mechanics to reduce computational cost, their application to cardiac hemodynamics is still limited. This is mainly due to the complexity of the problem, which involves large deformations, moving boundaries, strong nonlinearities and fluid-structure interaction [ 18 , 19 ] . Our method extends the capabilities of existing modal decomposition frameworks by leveraging the higher order dynamic mode decomposition (HODMD) algorithm [ 20 ] . The HODMD technique extracts the dominant coherent structures of the flow and models their temporal evolution. In doing so, it captures the underlying physics, most notably the VR dynamics that govern early diastolic filling in the left ventricle. Although HODMD has been successfully applied in simplified fluid scenarios, such as bluff body wakes or laminar shear flows [ 21 ] , this work presents its first application, to the authors’ knowledge, to cardiac flows. In this context, we address additional challenges inherent to cardiac flow simulations, such as significant domain deformation driven by myocardial motion ( i.e. involving moving meshes). These complexities are further augmented by the influence of the ventricular wall on VR dynamics, as well as by the substantial temporal variability in velocity and pressure fields throughout the cardiac cycle, driven by the strong coupling between geometry deformation and flow evolution. By adapting the ROM to the spatiotemporal characteristics of the data, we achieve significant computational savings while preserving the essential dynamics of the system. The following sections are structured as follows: Section 2 introduces the ROM methodology based on HODMD and its implementation. Section 3 presents the two idealized left ventricle databases used in this study. In Section 4 , we provide a detailed analysis of the ROM configuration using two distinct idealized LV geometries. Finally, Section 5 summarizes the main findings and discusses future directions. 2 Reduced Order Modelling As mentioned, the ROM employed in this study is based

Read the complete text