IEEE Transactions on Pattern Analysis and Machine Intelligence
Markov random field modeling in image analysis
Markov random field modeling in image analysis
A Framework for Incorporating Structural Prior Information into the Estimation of Medical Images
IPMI '93 Proceedings of the 13th International Conference on Information Processing in Medical Imaging
Hi-index | 0.00 |
The effectiveness of Bayesian reconstruction, or maximum a posteriori (MAP) method, has been proved in positron emission tomography. In this article, a novel convex MRF (Markov random fields) Membrane-Plate hybrid prior for Bayesian reconstruction, which combines quadratic smoothness prior of different orders, is proposed. The design of the new prior is based on the intrinsic properties of the two smoothness prior of different orders and aims to make an adaptive use of the two smoothness priors. The convexity of the new prior energy functional is ensured. Simulation experiments of their application in PET (Positron Emission Tomography) reconstruction are illustrated. Visional and quantitative comparisons showed the new hybrid prior’ good performance in lowering noise effect and preserving edges.