Predicting aortic regurgitation after transcatheter aortic valve replacement by finite element method

Abstract

Aortic regurgitation as a severe complication of transcatheter aortic valve replacement (TAVR) is usually due to the aortic valve leaflets that carry severity and inhomogeneous distribution of the calcification. However, it is difficult to precisely simulate the post-procedural biomechanical behavior on aortic tissue. This paper presents and validates a reliable system to predict which aortic stenosis patients may suffer aortic regurgitation after TAVR and to identify the best fit for TAVR valve. We randomly chose 22 patients (12 patients without regurgitation and 10 patients have regurgitation) who had been followed for at least 2 years after TAVR. An elastic model is designed to characterize the biomechanical behavior of the aortic tissue for each patient. After calculating the loading force on the tissue, the finite-element method (FEM) is applied to calculate the stresses of each tissue node. The support vector regression (SVR) method is used to model the relationship between the stress information and the risk of aortic regurgitation. Therefore, the risk of regurgitation and the optimal valve size can be predicted by this integrated model prior to the procedure. Leave-one-out cross-validation is implemented to assess the accuracy of our prediction. As a result, the mean prediction accuracy is 90.9% for all these cases, demonstrating the high value of this model as a decision-making assistant for pre-procedural planning of patients who are scheduled to undergo intervention. This method combines a bio-mechanical and machine learning approach to create a procedural planning tool that may support the clinical decision in the future.