NOISE AND RESOLUTION OF BAYESIAN RECONSTRUCTION FOR MULTIPLE IMAGE CONFIGURATIONS

Images reconstructed by Bayesian and maximum-likelihood (ML) using a Gibbs prior with prior weight beta were compared with images produced by filtered backprojection (FBP) from sinogram data simulated with different counts and image configurations. Bayesian images were generated by the OSL algorithm accelerated by an overrelaxation parameter. For relatively low beta, Bayesian images can yield an overall improvement to the images compared to ML reconstruction. However, for larger beta, Bayesian images degrade from the standpoint of noise and quantitation. Compared to FBP, the ML images were superior in a mean-square error sense in regions of low activity level and for small structures. At a comparable noise level to FBP, Bayesian reconstruction can be used to effectively recover higher resolution images. The overall performance is dependent on the image structure and the weight of the Bayesian prior.