ITERATIVE RECONSTRUCTION AND STATISTICAL OPTIMIZATION FOR THE NONDESTRUCTIVE ASSAY OF DISTRIBUTED GAMMA-SOURCE IN A LARGE NUCLEAR WASTE CONTAINER

An iterative reconstruction and statistical optimization method for th e nondestructive assay of total source activity of distributed gamma s ource in a large nuclear waste container is presented. Multiple detect ors positioned as closely as possible to the waste barrel are used to measure the emerging radiation field from the distributed radiation so urce. The source distribution is reconstructed by using the conjugate gradient with nonnegative constraint method or maximum likelihood expe ctation method based on measured detector responses. Total source acti vity is estimated by examining the error bond and its associated confi dence level. These are determined statistically by performing a large number of numerical experiments involving the counting statistics, the nonuniformity of source distribution, and the heterogeneous density o f the self-absorbing medium. Using the conjugate gradient with nonnega tive constraint method (CGNN) and maximum likelihood expectation maxim um method (MLEM) methods the authors compare results of total activity estimation.