On complete-data spaces for PET reconstruction algorithms
- Univ. of Michigan, Ann Arbor, MI (United States)
As investigators consider more comprehensive measurement models for emission tomography, there will be more choices for the complete-data spaces of the associated expectation-maximization (EM) algorithms for maximum-likelihood (ML) estimation. In this paper, the authors show that EM algorithms based on smaller complete-data spaces will typically converge faster. They discuss two practical applications of those concepts: (1) the ML-IA and ML-IB image reconstruction algorithms of Politte and Snyder which are based on measurement models that account for attenuation and accidental coincidences in positron-emissions tomography (PET), and (2) the problem of simultaneous estimation of emission and transmission parameters. Although the PET applications may often violate the necessary regularity conditions, their analysis predicts heuristically that the ML-IB algorithm, which has a smaller complete-data space, should converge faster than ML-IA. This is corroborated by the empirical findings.
- DOE Contract Number:
- FG02-87ER60561
- OSTI ID:
- 6011464
- Journal Information:
- IEEE Transactions on Nuclear Science (Institute of Electrical and Electronics Engineers); (United States), Journal Name: IEEE Transactions on Nuclear Science (Institute of Electrical and Electronics Engineers); (United States) Vol. 40:4 part 1; ISSN 0018-9499; ISSN IETNAE
- Country of Publication:
- United States
- Language:
- English
Similar Records
A multigrid expectation maximization reconstruction algorithm for positron emission tomography
Maximum likelihood reconstruction in fully 3D PET via the SAGE algorithm