Neural computing approach in simulating electrodynamics of magnetized blood enhanced with penta-hybrid nanoparticles in a multi-stenosed artery

Abstract

The dynamics of electro-osmotically induced flow are increasingly recognized for their vast applications in medical sciences and bioengineering. This research aims to simulate the electro-osmotically modulated flow of blood augmented with penta-hybrid nanoparticles through an artery exhibiting multiple stenoses. Utilizing the Casson fluid model, we trace the non-Newtonian characteristics of blood navigating through an artery with various stenotic obstructions. The model incorporates multiple physical aspects such as heat origination, Joule heating, interfacial nanolayers , and both thermodynamic and flow-slip constraints, simplifying the problem with Lubrication regime and Debye-Hückel linear expansion, and applying the perturbation-based homotopy approach to solve the dimensionless equations. Key hemodynamic features and quantities are visualized through graphs. Our findings reveal the pronounced influence of operational parameters on blood flow, with electro-osmotic factors significantly enhancing mobility through the stenosed artery. Increased electro-osmotic factor leads to a rise in blood temperature, while a larger stenosis shape factor decreases blood velocity. An expansion in the nanolayered interface thickness diminishes the thermal trait of blood, and a higher Hartmann number lowers the pressure gradient across the artery. The AI aspect of the study utilizes a regressor artificial neural network (ANN) and achieves remarkable predictive accuracies-99.61 % for arterial wall shear stress on the training database and 99.95 % on the validation data pool. The model results have significant potential to enhance treatments for arterial blockages and substantially impact cardiovascular healthcare.

Topics & Concepts