Artificial intelligence to predict atheroma plaque vulnerability

Abstract

Cardiovascular diseases related to atherosclerosis are the first cause of death in the western world. This relevant fact has motivated the development of numerical models for arterial behavior in order to understand better cardiovascular pathologies. This chapter provides a parametric tool, using Machine Learning Techniques (MLTs), to assist the clinicians on decisions of the vulnerability of the atheroma plaque, especially when an instantaneous response is needed. The MLTs use an intelligent algorithm to model the atheroma plaque rupture in terms of four of the most influential geometrical factors in the plaque rupture: (i) fibrous cap thickness; (ii) stenosis ratio; (iii) lipid core width, and (iv) lipid core length. The output predicted is the maximum maximal principal stress occurred in an atherosclerotic coronary vessel with the input dimensions. For this purpose, an idealized and parametric coronary vessel model has been performed using finite element methods in order to train the machine learning.