Effect of Mesh Elements and Hyper-parameter over 2-Dimensional Electrical Impedance Tomography Model

Abstract

Body surface electrodes are used in electrical impedance tomography (EIT) tests. These tests yields an image of the conductivity distribution. Image reconstruction mainly depends on data obtained from sensor electrodes through experiments or tests. Data contamination due to introduction of noise and poor conditioning of the system matrix makes the reconstruction techniques challenging. To give a stable and accurate solution to an ill conditioned EIT problem some form of prior knowledge is introduced by a scalar hyperparameter. Number of mesh elements used during reconstruction process plays a vital role. This paper proposes the comparative analysis of effect of number of mesh elements on reconstruction. Also the optimum number of mesh elements for reconstruction are proposed. Also the effect of change of hyperparameter on reconstruction is shown for above mentioned algorithms.