Latest revision as of 14:07, 21 April 2020

Abstract

Combination of numerical modeling and artificial intelligence (AI) in bioengineering processes are a promising pathway for the further development of bioengineering sciences. The objective of this work is to use Artificial Neural Networks (ANN) to reduce the long computational times needed in the analysis of shear stress in the Abdominal Aortic Aneurysm (AAA) by finite element methods (FEM). For that purpose two different neural networks are created. The first neural network (Mesh Neural Network, MNN) creates the aneurysm geometry in terms of four geometrical factors (asymmetry factor, aneurism diameter, aneurism thickness, aneurism length). The second neural network (Tension Neural Network, TNN) combines the results of the first neural network with the arterial pressure (new factor) to obtain the maximum stress distribution (output variable) in the aneurysm wall. The use of FEM for the analysis and design of bioengineering processes often requires high computational costs, but if this technique is combined with artificial intelligence, such as neural networks, the simulation time is significantly reduced. The shear stress obtained by the artificial neural models developed in this work achieved 95% of accuracy respect to the wall stress obtained by the FEM. On the other hand, the computational time is significantly reduced compared to the FEM.