Predicting nanofluid behavior in inflamed stenotic arteries: a neural network and finite element-Based analysis.

Abstract

This study examines heat transfer and nanofluid-enhanced blood flow behaviour in stenotic arteries under inflammatory conditions, addressing critical challenges in cardiovascular health. The blood, treated as a Newtonian fluid, is augmented with gold nanoparticles to improve thermal conductivity and support drug delivery applications. A hybrid methodology combining finite element method (FEM) for numerical modelling and artificial neural networks (ANN) for stability prediction provides a robust analytical framework. Parametric analysis reveals that increasing stenosis severity (60% to 80%) results in a 45% enhancement in heat transfer, demonstrating the efficacy of nanoparticle integration. The results show that the size of the vortices decreases due to the position changing of the upper stenoses, whereas it rises with increasing stenosis peak. Higher nanoparticle volume fraction ($\varphi$) amplifies momentum diffusion, resulting in larger vortices, while improved thermal conductivity enhances heat transfer. Inflammation significantly affects flow patterns and heat transport with important implications in treating cardiovascular disorders and biological applications. The regression analysis confirms a close match between predicted and target data, showcasing the robustness of the FEM-ANN hybrid approach for modelling biofluid systems.