Bayesian reconstruction in emission computerized tomography

Iterative Bayesian image reconstruction algorithms for emission computerized tomography have been derived and applied using three general classes of a priori source distribution con-straints: 1) maximum entropy constraints on source elements; 2) source continuity and boundary information; 3) ideal and fuzzy pattern source distribution constraints. This work has recently been extended to include a combined Bernoulli process and entropy analysis on source element strength correlations and also to provide for more general a priori probability distributions of probable source element strengths in sub-sets of a source field. Improvement in image reconstruction over standard methods has been demonstrated by applying these Bayesian algorithms to computer simulation data and experimental nuclear isotope imaging data.