Abstract:
In this paper, we present a new derivation of a wide class of list-mode ordered-subsets (OS) algorithms for image reconstruction in PET. The derivation starts from the li...Show MoreMetadata
Abstract:
In this paper, we present a new derivation of a wide class of list-mode ordered-subsets (OS) algorithms for image reconstruction in PET. The derivation starts from the list-mode likelihood and follows the similar line to the block-gradient method for minimizing a general cost function. This derivation clarifies that there exist list-mode OS algorithms which behave significantly better than current standard list-mode OS-EM.
Published in: IEEE Nuclear Science Symposium Conference Record, 2005
Date of Conference: 23-29 October 2005
Date Added to IEEE Xplore: 27 February 2006
Print ISBN:0-7803-9221-3
Print ISSN: 1082-3654
Contents 