Machine learning analysis of shear stress over a symmetrically stenosed arterial wall under the impact of magnetic field

This study provides a framework to predict the shear stress over a symmetrically stenosed arterial wall under the impact of magnetic field using machine learning algorithms. We investigated the magnetohydrodynamics blood flow with the help of machine learning techniques. The considered model is governed by the principles of MHD. To predict the flow patterns and velocity of the blood flow it is important to study the relationship between the blood shear stress rate with the six blood flow input variables: flow rate (q), viscosity, radius (r), height(h), pressure (P) and velocity (v). We presented the correlation graph of shear stress with the six blood flow input variables. The machine learning models (CatBoost regressor, Random Forest regressor, Regressor tree, MLP regressor and Lasso model) were used to validate the dataset. Radar graphs have been plotted to visualize and compare the accuracy measured for all the five machine learning models. Our conclusion pretends that the random forest regressor analyze this problem accurately with noise free data.