Reconstruction of CT images by the Bayes- back projection method

In the course of research on quantitative assay of non-destructive measurement of radioactive waste, the have developed a unique program based on the Bayesian theory for reconstruction of transmission computed tomography (TCT) image. The reconstruction of cross-section images in the CT technology usually employs the Filtered Back Projection method. The new imaging reconstruction program reported here is based on the Bayesian Back Projection method, and it has a function of iterative improvement images by every step of measurement. Namely, this method has the capability of prompt display of a cross-section image corresponding to each angled projection data from every measurement. Hence, it is possible to observe an improved cross-section view by reflecting each projection data in almost real time. From the basic theory of Baysian Back Projection method, it can be not only applied to CT types of 1st, 2nd, and 3rd generation. This reported deals with a reconstruction program of cross-section images in the CT of 3rd generation type (fan beam projection). The performance tests of the program have been carried out using the projection data obtained with a virtual CT system (li.e,, simulation calculation with the MVP Monte Carlo code) and the real data obtained from the TCT instrument equipped with 1920ch array detectors. Finally, we have confirmed that sharp cross section images can be obtained with a spatial resolution of 0.1 mm. The present reconstruction program developed here for the gamma-ray CT can widely applied also to, for instance, the Neutron Imaging and the Positron Emission Tomography (PET).