Proper analysis of finite element results for predicting abdominal aortic aneurysm rupture

Abstract

Investigation of the predictability of finite element (FE) models regarding rupture risk assessment of abdominal aortic aneurysms (AAAs).Peak wall stress (PWS) and peak wall rupture risk (PWRR) of ruptured (n = 20) and non-ruptured (n = 30) AAAs were predicted by four FE models of different complexities derived from computed tomography (CT) data. Two matching sub-groups of ruptured and non-ruptured aneurysms were used to investigate the usability of different FE models to discriminate amongst them.All FE models exhibited a strong positive correlation between PWS and PWRR with the maximum diameter. FE models, which excluded the intra-luminal thrombus (ILT) failed to discriminate between ruptured and non-ruptured aneurysms. The predictability of all applied FE models was strengthened by including wall strength data, that is, computing the PWRR. The most sophisticated FE model applied in this study predicted PWS and PWRR 1.17 (p = 0.021) and 1.43 (p = 0.016) times higher in ruptured than diameter-matched non-ruptured aneurysms, respectively.PWRR reinforces PWS as a biomechanical rupture risk index. The ILT has a major impact on AAA biomechanics and rupture risk, and hence, needs to be considered in meaningful FE simulations. The applied FE models, however, could not explain rupture in all analysed aneurysms.