Segmentation of MR image based on maximum a posterior

Brain MR image segmentation takes an important role in research and clinical application. Statistical method is effective in the segmentation, which is usually based on maximum a posterior (MAP). The key of MAP method is to estimate a prior probability of the segmentation. Multilevel logistic (MILL) model has been used in practice for the estimation. To further improve the performance of the segmentation, a weighted MLL (WMLL) model is proposed in this paper. The simulated results show that the WMILL model is effective.