Failure prediction and robust design of grafts for aortic aneurysms

The development of an abdominal aortic aneurysm (AAA) is a common condition that can lead to fatal rupture of the aorta. Endovascular repair (EVAR) is a recently implemented AAA treatment option that involves the catheterized insertion of an endoluminal graft (ELG) into the aneurysm sac to relieve pressure. Though EVAR is relatively noninvasive, it is subject to failure mechanisms including re-pressurization of the aneurysm sac as a result of leakage between the ELG and aorta (i.e. endoleak). In order to understand the influence of clinically relevant variables on the occurrence of failure, a parametric study was conducted using a finite element model capable of simulating the fluid-structure interaction inherent to the initiation of endoleak. The parameter space was explicitly decomposed into failure and non-failure regions using Support Vector Machines, a novel approach in the field of design that facilitates calculation of the probability of failure as well as robust graft design.