Modelo numérico de las trayectorias de las nanopartículas magnéticas en aproximación de un flujo no Newtoniano

Abstract

In this work, the biophysical variables present in the distribution of nanopharmaceuticals in the circulatory system were studied, using the Casson model with which the behavior of blood as a colloid was described, assuming viscosity as a variable that depends on the gradient of speed, shear stress and temperature. For this, the physical factors present during the supply of the magnetic nanoparticles (NPMs) were related, such as: gravitational field, magnetic field, Stokes force, thrust force and drag force. Thus, a numerical model of the trajectory of these NPMs was developed with the intention of estimating their distribution in specific regions of the organism. From the equations of motion, a numerical model was developed that was computationally solved by means of the Euler-Chrome algorithm, which allowed a detailed analysis of the trajectory of said particles, thus building a database that feeds a neural network, by means of from which the behavior of the NPMs was estimated, for which artificial neural networks were implemented, such as the multilayer perceptron, with optimization algorithms in which the Levenberg Marquadt algorithm stands out. From the above, different trajectories of the NPMs in coronary arteries were estimated, including parameters such as time, the position in X and Y, the speed that the nanoparticles can reach. The architecture obtained with the artificial neural network, which contains the optimization algorithm [5 4 3 2], presented the best performance with a training MSE of 1.763E-07, validation uRMSE of 0.0049 and trend probabilities at X 0, 62 and 0.57 in Y.