Radiation intensity image reconstruction in TGS using ART algorithm with geometrically-corrected system matrix and TV constraint

Abstract

Tomographic Gamma Scanning (TGS) is employed increasingly to perform the non-destructive assay for radioactive waste. In TGS, the radioactivity distribution within a target object is reconstructed based on the emission CT principles. However, due to the un-negligible detector size and the varying field of view with distances from detector, the ordinary line integral model for CT projection tends to be inapplicable for the radiation intensity image reconstruction in TGS system. The geometrical correction to system matrix and the implementation by iterative algorithm are proposed to solve this problem. In comparison to statistical iteration algorithms such as Maximum-Likelihood Expectation-Maximization (MLEM), the Algebraic Reconstruction Technique (ART) algorithm can probably achieve a higher convergence rate, whereas it accumulates reconstruction errors more significantly with iteration, particularly with the more complicated geometrically-corrected system matrix. In this paper, an ART algorithm combined with Total Variation (TV) minimization constraint is presented. By use of the TV constraint, the errors in reconstructed images are effectively eliminated. Monte-Carlo based numerical simulation is performed, and the results show that the ART-TV algorithm reduces the time for image reconstruction without loss of image quality compared with the MLEM algorithm.