Improving arterial stiffness prediction with machine learning utilizing hemodynamics and biomechanical features derived from phase contrast magnetic resonance imaging.

Arterial stiffness has emerged as a prominent marker of risk for cardiovascular diseases. Few studies are interested in predicting symptomatic or asymptomatic arterial stiffness from hemodynamics and biomechanics parameters. Machine learning models can be used as an intelligent tool for arterial stiffness detection based on hemodynamic and biomechanical parameters. Indeed, in the case of arterial stiffness hemodynamics and biomechanics parameters present significant change, such as an increase in age, local wave velocity, arterial elastance, Young's modulus, reflected wave amplitude, decrease in arterial compliance, reflected wave arrival time, and reflection coefficient. This study aims to assess the impact of artificial intelligence using machine-learning algorithms for the detection of arterial stiffness. The ability of various machine-learning approaches can be investigated to predict wall stiffness in the carotid artery and to evaluate the risk of cardiovascular events. A mathematical model developed in previous work was used to determine hemodynamic and biomechanical parameters. Accuracy, sensitivity, and specificity are calculated to evaluate the performance of the proposed models. All used classifiers demonstrated high performance in predicting arterial stiffness, notably with the Support Vector Machine, Artificial Neural Network, and Decision Tree classifiers achieving exceptional accuracies of 100%. In this study, the potential of machine learning based on hemodynamic parameters for the prediction of symptomatic and asymptomatic arterial stiffness was demonstrated.