Numerical studies on Palamodov and Generalized Feldkamp algorithm for general cone-beam scanning

Article type: Research Article

Authors: Zeng, Kai | Yu, Hengyong | Bai, Erwei | Wang, Ge

Affiliations: Biomedical Engineering Dept, University of Iowa, 132 Hawkeye court, Iowa city, IA 52246, USA | Biomedical imaging Division, VT-WFU School of Biomedical Engineering & Science, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA | Department of Electrical and Computer Engineering, University of Iowa, Iowa city, IA 52246, USA

Note: [] Corresponding author: Kai Zeng, Tel.: +1 319 353 4311; Fax: +1 540 231 0970; E-mail: kzeng@engineering.uiowa.edu

Abstract: Recently, the Palamodov algorithm, which was formulated for exact cone-beam reconstruction from data collected along a continuous locus, has been proven to be an excellent approximate algorithm for general cone beam reconstruction. The filtration-backprojection framework of the Palamodov algorithm is efficient for sequential and parallel implementation because its filtration step only involves a 1-D shift-invariant filtering along the tangent direction of the scanning trajectory. On the other hand, the generalized Feldkamp algorithm proposed by Wang et al.~also allows approximate general cone beam reconstruction. In this paper, we report a numerical study comparing these two approximate methods for the cases of helical and saddle curves. In this study, the image quality is evaluated in terms of mean square error (MSE), modulation transform function (MTF), etc. The results demonstrate that the Palamodov algorithm consistently performs similar or better than the generalized Feldkamp algorithm.

Keywords: Computed tomography (CT), palamodov algorithm, generalized feldkamp algorithm, approximate reconstruction, general cone-beam scanning

Journal: Journal of X-Ray Science and Technology, vol. 15, no. 3, pp. 131-146, 2007