Developing a reduced order model for pulsatile blood flow simulations using minimal three-dimensional simulation data

• Published in: Computer methods and programs in biomedicine Vol. 271; p. 108994
• Main Authors: Choi, Wonjin, Lee, Inpyo, Kim, Hyun Jin
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier B.V 01.11.2025
• Subjects: 1D blood flow equations; Aorta; Blood flow simulation; Blood Flow Velocity; Computational fluid dynamics; Computer Simulation; Coronary Vessels; Humans; Hydrodynamics; Imaging, Three-Dimensional; Models, Cardiovascular; Pulsatile Flow; Pulsatile simulation; Reduced-order model; 1D blood flow equations; Computational fluid dynamics; Blood flow simulation; Pulsatile simulation; Reduced-order model
• ISSN: 0169-2607; 1872-7565; 1872-7565
• DOI: 10.1016/j.cmpb.2025.108994

Pulsatile blood flow simulations are essential for understanding cardiovascular physiology but are often constrained by the computational cost of full 3D modeling. While reduced-order models (ROMs) offer efficiency, many depend on empirical parameters, limiting their accuracy in complex subject-specific geometries. This study introduces a 1D ROM that minimizes empirical assumptions by deriving parameters directly from 3D simulation data. The proposed ROM estimates pressure-flow relationships by fitting parameters of a simplified 1D blood flow equation to results from three 3D simulations under distinct flow conditions. Validation was performed against full 3D Computational Fluid Dynamics (CFD) simulations for three cases: an idealized stenotic cylinder, a patient-specific coarcted aorta, and a coronary artery model with serial lesions. The model was also compared with a widely used empirical stenosis model. The ROM achieved mean relative errors below 2.0 % in all cases and under 1.0 % in the coronary model with multiple lesions. It significantly outperformed the empirical model in complex geometries and delivered up to 3000 times faster computation on a desktop compared to 3D simulations performed on a 64-core high-performance computing system. By leveraging 3D simulation data, the proposed 1D ROM combines high accuracy with exceptional computational efficiency across various geometries. Its robustness and speed make it well suited for clinical applications, optimization tasks, and uncertainty quantification. •A 1D reduced-order model (ROM) with 3D simulation-based coefficients is developed for pulsatile blood flow simulations.•The proposed ROM shows high accuracy for various geometries and flow conditions.•The proposed ROM can replace pulsatile 3D simulations for computationally demanding studies such as optimization and uncertainty quantification.