Bayesian segmentation of multislice brain magnetic resonance imaging using three-dimensional Gibbsian priors
We propose a maximum a posteriori probability method, which features 3D Gibbsian priors and the highest confidence first (HCF) optimization method, for segmentation of cross- sectional images through a 3D volume. The proposed algorithm has been successfully applied to segmentation of clinical magnetic resonance imaging (MRI) data on the human brain. Modeling the a priori probability of the segmentation by a 3D Gibbs random field imposes connectivity and smoothness constraints on the desired segmentation in all three directions. HCF is a recently proposed optimization method, which proves superior to other existing methods in this application. We discuss several implementation issues, including the effect of varying parameter values on algorithm performance. Experimental results with both phantoms and clinical MR data show that our proposed approach improves on existing methods in delineation of the cerebral cortex, deep nuclei and the cerebral white matter.