Improving Three-Dimensional Electrical Capacitance Tomography Imaging Using Approximation Error Model Theory

Abstract

—Electrical capacitance tomography (ECT) is a noninvasive technique that aims to reconstruct images of internal permittivity values of a volume of interest, based on measurements taken on the external boundary. Since most reconstruction algorithms rely on model-based approximations, it is important to ensure numerical accuracy for the model being used. Finite element methods (FEM) are the most commonly used modelling technique for the ECT forward problem. Due to the characteristic of the ECT technique, the size of the electrodes needs to be built as big as possible so as to cover the entire image area. To accurately solve the forward model a large number of elements are required in and around electrodes. Notably complete sensor model (CSM) will provide the best solution of the forward model. It is also known that denser mesh provides more accurate solution; however, it also requires longer computation time. This paper demonstrates the advantages of the incorporating discretisation and modelling error in 3D ECT problem. The approximation error model (AEM) has been used to account for discretisation error, and this allows using a coarse mesh for the forward model and achieving the reconstruction accuracy close to that of denser mesh.