Reconstruction of electrical impedance tomography images by simulated annealing using the quadratic global error of overdetermined systems as objective function

Abstract

Electrical impedance tomography is a non-invasive technique for obtaining images of the interior of a body. The image reconstruction can be approached as an optimization problem that looks for the maximization of the consistency between the behavior of a parameterized model of the body and physical measurements. This optimization problem is non-linear and non-convex, and its solution usually involves a great computational cost, as the evaluation of the objective function at each iteration involves the simulation of the parameterized model, usually through the Finite Element Method. The goal of this project is to create a new objective function that admits an iterative evaluation at a low computational cost.(AU)