Investigation of Rupture Risk of Thoracic Aortic Aneurysms via Fluid–Structure Interaction and Artificial Intelligence Method

Abstract

Patient-specific studies on vascular flows have significantly increased for hemodynamics due to the need for different observation techniques in clinical practice. In this study, we investigate aortic aneurysms in terms of deformation, stress, and rupture risk. The effect of Ascending Aortic Diameter (AAD) was investigated in different aortic arches (19.81 mm, 42.94 mm, and 48.01 mm) via Computational Fluid Dynamics (CFD), Two-way coupling Fluid–Structure Interactions (FSI) and deep learning. The non-newtonian Carreau viscosity model was utilized with patient-specific velocity waveform. Deformations, Wall Shear Stresses (WSSs), von Mises stress, and rupture risk were presented by safety factors. Results show that the WSS distribution is distinctly higher in rigid cases than the elastic cases. Although WSS values rise with the increase in AAD, aneurysm regions indicate low WSS values in both rigid and elastic artery solutions. For the given AADs, the deformations are 2.75 mm, 6. 82 mm, and 8.48 mm and Equivalent von Mises stresses are 0.16 MPa, 0.46 MPa, and 0.53 MPa. When the rupture risk was evaluated for the arteries, the results showed that the aneurysm with AAD of 48.01 mm poses a risk up to three times more than AAD of 19.81 mm. In addition, an Artificial neural network (ANN) method was developed to predict the rupture risk with a 98.6% accurate prediction by numerical data. As a result, FSI could indicate more accurately the level of rupture risk than the rigid artery assumptions to guide the clinical assessments and deep learning methods could decrease the computational costs according to CFD and FSI.