Investigating the Feasibility of a Convolutional Neural Network Surrogate Model Employing an Overlapping Domain Decomposition Approach for Steady-State Blood Flow Predictions

This study builds on the work of Takashi Shimokawabe et al., who developed a CNN-based surrogate model using an overlapping domain decomposition approach to predict steady-state flow in large geometries with embedded obstacles. The goal of the current research is to investigate the feasibility of applying this approach to steady-state blood flow predictions in arteries with varying diameters. This presentation addresses the challenges of this approach and presents preliminary results.