Introduction of ordered subsets algorithm to maximum a posteriori expectation maximization method

Iterative reconstruction methods such as the maximum likelihood (ML)-expectation maximization (EM) method can be accelerated by introducing an ordered subsets (OS) algorithm, in which the projection data are grouped into subsets and a pixel in a reconstructed image is updated by using projections in each subset. In this paper we introduced the OS algorithm to the maximum a posteriori (MAP)-EM method and named this method OS-Bayesian reconstruction (BR). The performance of OS-BR was compared with ML-EM, MAP-EM and OS-EM on reconstructed images. The results showed that OS-BR with suitable parameters yielded higher quality images than the other methods at earlier iterations.