Dirichlet Markov Random Field Segmentation of Brain MR Images
Wentao Wang, Cong Chen · International Conference on Bioinformatics and Biomedical Engineering · 2010
Accurate segmentation of magnetic resonance images according to tissue type is widely studded by many researcher, Recently Markov Random Field (MRF) has been used in this area. However the original MRF is supervised. So we introduce a novel approach called Dirichlet Markov Random Field for Magnetic Resonance Image (MRI) brain tissue classification. The approach uses Dirchilet Process Mixture (DPM) to get local information of MRF's energy function. Instead of finite component, DPM use infinite component, in which the prior distribution is defined on the space of all possible distribution. But efficient implementations of the DP mixture model can be slowly to converge and their convergence can be difficult to diagnose with the Markov Chain Monte Carlo (MCMC) methods for sampling from the posterior distribution of the parameters. So this algorithm uses variational Bayesian (VB) approximations to the DP mixture model. Experiment result proved this algorithm can segment the MRI smoothly and accurately.