Attenuation and Scattering Correction in Passive Gamma Emission Tomography Reconstruction

The International Atomic Energy Agency has developed a novel non-destruction assay technique, Passive Gamma Emission Tomography, to inspect spent nuclear fuel assemblies. Accurate physical modeling of the tomography system is required for quantitative tomography reconstruction of the fuel assembly. We have implemented an accelerated Monte Carlo algorithm to compute the imaging system response matrix, applying the attenuation and scattering correction. We performed iterative reconstruction of four simulated fuel assemblies by solving the corresponding imaging inverse problem. The images yielded by the inverse approach were then processed by a convolutional neural network for pin identification. Perfect identification of fuel pins was achieved for medium and high activity level pins. The accuracy of fuel pin activity estimation was significantly improved, compared to the standard filtered back projection (FBP) approach.