Reconstruction algorithms for electrical impedance tomography

Abstract

Electrical Impedance Tomography (EIT) is a technique that is used for imaging the interior of a region, in a non-invasive fashion. Electrical current is passed across the object of interest and voltage measurements are taken around the periphery of the object. A mathematical model is then used to estimate the interior structure. EIT has applications in medicine and process tomography. This thesis studies two reconstruction algorithms for EIT. The first is based on a technique that is in wide use throughout the field and was pioneered by Yorkey. The second algorithm makes assumptions about the underlying geometry of the problem which allows the use of analytic methods. The approach of assuming the geometry of the object being imaged is used to reduce the dimension of the inverse problem in an attempt to make the problem better posed. The second algorithm presented was found to be more stable on test cases, but more limited in the applications for which it could be employed.