Shape optimization of an artificial bypass graft using genetic algorithm

Abstract

Shape optimization has been applied in medical engineering. It is our goal to establish an optimization framework suited to perform shape optimization of complex geometries in compute-intensive applications. In medical practice, bypass grafts provide a way to restore blood flow in critically stenosed arteries, and, together with stenting, constitute the most typical treatment in such cases. In this work, the numerical analysis of the blood flow phenomena uses the finite element method approach applied to a geometrical model of the artery and bypass. Blood flow is described by the incompressible Navier-Stokes equations and the simulation is carried out under steady flow conditions. The optimization of the bypass geometry of an idealized artificial graft is searched aiming to minimize the shear rate integral. The shapes are described either using elliptical or sinusoidal functions. An evolutionary genetic algorithm is considered in order to calculate the optimal shape for the graft geometry. Numerical results show the benefits of numerical shape optimization in achieving design improvements.