Comparing Analysis of Image Visualisation Accuracy of Electrical Capacitance Tomography and Gamma Tomography

R. Maad 1,2, D.Sankowski 3,, K Grudzien 3, R.Wajman 3, V.Mosorov 3, G. A. Johansen 2

1Sogn og Fjordane College, Faculty of Engineering, N-6800 Førde, Norway.

E-mail: rachid.maad@hisf.no

2University of Bergen, Department of Physics and Technology, Allégaten 55, N-5007 Bergen, Norway, E-mail: geir.johansen@ift.uib.no

3Computer Engineering Department, Technical University of Lodz, Poland, E-mail: dsan@kis.p.lodz.pl

ABSTRACT

The aim of this paper is to compare the visualisation quality of different kinds of tomography. The quality of reconstructed images using data from an 85-channel high-speed gamma-ray tomograph (GRT) and from a 12-electrode electrical capacitance tomograph (ECT have been compared. Based on this objective, a polypropylene phantom with two predrilled holes is used for both tomographs. For GRT several values of data acquisition time have been used, which is identical to the counting time allowing the effect of the Poisson counting statistics to be studied. The images have been reconstructed on square grids of 16 x 16, 32 x 32, and 64 x 64 pixels. The values of the attenuation coefficient and the permittivity of each pixel are calculated by projecting the grid onto the phantom cross section and comparing them to the true value. Several image reconstruction algorithms were used: Alternating Minimization (AM), Multiplicative ART (MART), Iterative Least Square Technique (ILST) for GRT, Iterative Back Projection (IBP), Landweber Iterations (LI) and Newton – Raphson (NR) for ECT. In all cases, the reconstruction error is calculated as the Root Mean Square Error (RMSE) with respect to the number of pixels (N).

Keywords Gamma-ray tomography, electrical capacitance tomography, image reconstruction