ARBITRARY HYBRID PHYSICS/DATA DRIVEN AND MULTISCALE/MULTIPHYSICS SIMULATION METHODS FOR PATIENT-SPECIFIC STUDY OF CORONARY ARTERY DISEASE (MMPINN-CAD)
No Thumbnail Available
Files
Date
2024-11-20
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Nazarbayev University School of Engineering and Digital Sciences
Abstract
The study aims to develop novel hybrid physics/data-driven and multiscale/multiphysics simulation methods for patient-specific investigation of coronary artery disease (CAD), termed MMPINN-CAD. It introduces a Physiologically Based Algorithm (PBA) integrated into the OpenFOAM CFD solver for 3D CFD simulation of Fractional Flow Reserve (FFR) in coronary artery trees. The PBA simulates blood flow dynamics, accurately computing outlet boundary conditions based on Murray’s law and patient-specific inlet conditions. This non-invasive estimation of FFR offers a promising avenue for patient-specific CAD detection, overcoming challenges faced by traditional methods.
Additionally, the study presents the Hybrid CFD PINN FSI method, which combines deep learning and fundamental physics principles to replicate fluid flow patterns in coronary artery networks. The 1D Physics-Informed Neural Network (PINN) model demonstrates exceptional accuracy and efficiency, outperforming its Finite Element Method (FEM) counterparts. Further validation highlights the method's versatility and potential for non-invasive CAD diagnosis.
Furthermore, the research includes the development of a 3D PINN model for comprehensive fluid flow modelling in coronary artery trees. This model leverages deep learning techniques and fundamental physics principles to replicate fluid flow patterns and accurately identify potential stenotic regions. Through ongoing refinement and validation, the 3D PINN model aims to enhance patient-specific CAD diagnosis and treatment planning.
Overall, this research aligns with WHO's strategies for combating cardiovascular diseases globally, emphasizing the potential impact of the proposed simulation methods in improving patient care and outcomes.
Description
Keywords
Type of access: Gated
Citation
Alzhanov, N. (2024). Arbitrary hybrid physics/data driven and multiscale/multiphysics simulation methods for patient-specific study of coronary artery disease (MMPINN-CAD). Nazarbayev University School of Engineering and Digital Sciences.