Enhancing Arterial Blood Flow Simulations through Physics-Informed Neural Networks

Shivam Bhargava,Nagaiah Chamakuri

2024-04-26

Abstract:This study introduces a computational approach leveraging Physics-Informed Neural Networks (PINNs) for the efficient computation of arterial blood flows, particularly focusing on solving the incompressible Navier-Stokes equations by using the domain decomposition technique. Unlike conventional computational fluid dynamics methods, PINNs offer advantages by eliminating the need for discretized meshes and enabling the direct solution of partial differential equations (PDEs). In this paper, we propose the weighted Extended Physics-Informed Neural Networks (WXPINNs) and weighted Conservative Physics-Informed Neural Networks (WCPINNs), tailored for detailed hemodynamic simulations based on generalized space-time domain decomposition techniques. The inclusion of multiple neural networks enhances the representation capacity of the weighted PINN methods. Furthermore, the weighted PINNs can be efficiently trained in parallel computing frameworks by employing separate neural networks for each sub-domain. We show that PINNs simulation results circumvent backflow instabilities, underscoring a notable advantage of employing PINNs over traditional numerical methods to solve such complex blood flow models. They naturally address such challenges within their formulations. The presented numerical results demonstrate that the proposed weighted PINNs outperform traditional PINNs settings, where sub-PINNs are applied to each subdomain separately. This study contributes to the integration of deep learning methodologies with fluid mechanics, paving the way for accurate and efficient high-fidelity simulations in biomedical applications, particularly in modeling arterial blood flow.

Numerical Analysis,Fluid Dynamics

What problem does this paper attempt to address?

### Problems the Paper Aims to Solve

This paper aims to demonstrate the effectiveness of Physics-Informed Neural Networks (PINNs) in simulating arterial blood flow, overcoming existing computational barriers while ensuring high-fidelity simulation results. Specifically, the study addresses the following points:

1. **Efficient Computation Using PINNs**: Efficient computation of arterial blood flow using Physics-Informed Neural Networks (PINNs), particularly by leveraging domain decomposition techniques, focusing on solving the incompressible Navier-Stokes equations.

2. **Improving Computational Efficiency**: Proposing Weighted Extended Physics-Informed Neural Networks (WXPINNs) and Weighted Conservative Physics-Informed Neural Networks (WCPINNs) for detailed hemodynamic simulations based on generalized spatiotemporal domain decomposition techniques. These methods can significantly improve computational efficiency and can be efficiently trained within a parallel computing framework.

3. **Addressing Reflux Instability Issues**: The PINNs method can naturally handle reflux instability issues without the need for traditional stabilization terms, thus outperforming conventional numerical methods.

4. **Parallel Computation for Large-Scale Problems**: By decomposing the computational domain into multiple subdomains, parallel computation for large-scale problems can be achieved, thereby enhancing overall computational efficiency.

5. **Improving Convergence**: Enhancing the scalability and robustness of PINNs through optimization techniques and domain decomposition strategies.

In summary, this study aims to provide a new paradigm for high-fidelity simulations in biomedical engineering and other fields by combining machine learning with fluid mechanics.