EDART, a discrete algebraic reconstructing technique for experimental data obtained with high resolution computed tomography

Article type: Research Article

Authors: Brabant, L. | Dierick, M. | Pauwels, E. | Boone, M.N. | Van Hoorebeke, L.

Affiliations: Department of Physics and Astronomy, Ghent University, Gent, Belgium

Note: [] Corresponding author: L. Brabant, Department of Physics and Astronomy, Ghent University, Proeftuinstraat 86, 9000 Gent, Belgium. Tel.: +32 9264 6611; Fax: +32 9264 6697; E-mail: Loes.Brabant@ugent.be

Abstract: A novel reconstruction method is presented to improve the image quality of three dimensional (3D) datasets for samples consisting of only one material and surrounding air, obtained with high resolution X-ray Computed Tomography (μCT). It combines discrete tomography with iterative reconstruction algorithms, it is applicable for routine μCT applications and is referred to as the Experimental Discrete Algebraic Reconstruction Technique (EDART). A fast and intuitive method to estimate the attenuation coefficient and segmentation threshold, in case these are unknown, is included. Experimental results illustrate that EDART allows the improvement of the reconstruction quality as compared to standard iterative reconstruction when few projections are available, without significantly increasing the reconstruction time.

Keywords: Discrete tomography, high resolution CT, iterative reconstruction

DOI: 10.3233/XST-130408

Journal: Journal of X-Ray Science and Technology, vol. 22, no. 1, pp. 47-61, 2014