Estimation of Binary Markov Random Fields Using Markov chain Monte Carlo

This article compares three binary Markov random fields (MRFs) which are popular Bayesian priors for spatial smoothing. These are the Ising prior and two priors based on latent Gaussian MRFs. Concern is given to the selection of a suitable Markov chain Monte Carlo (MCMC) sampling scheme for each prior. The properties of the three priors and sampling schemes are investigated in the context of three empirical examples. The first is a simulated dataset, the second involves a confocal fluorescence microscopy dataset, while the third is based on the analysis of functional magnetic resonance imaging (fMRI) data. In the case of the Ising prior, single site and multi-site Swendsen-Wang sampling schemes are both considered. The single site scheme is shown to work consistently well, while it is shown that the Swendsen-Wang algorithm can have convergence problems. The sampling schemes for the priors are extended to generate the smoothing parameters, so that estimation becomes fully automatic. Although this works well, it is found that for highly contiguous images fixing smoothing parameters to very high values can improve results by injecting additional prior information concerning the level of contiguity in the image. The relative properties of the three binary MRFs are investigated, and it is shown how the Ising prior in particular defines sharp edges and encourages clustering. In addition, one of the latent Gaussian MRF priors is shown to be unable to distinguish between higher levels of smoothing. In the context of the fMRI example we also undertake a simulation study.

[1]  Irene A. Stegun,et al.  Handbook of Mathematical Functions. , 1966 .

[2]  J. Besag Spatial Interaction and the Statistical Analysis of Lattice Systems , 1974 .

[3]  Donald Geman,et al.  Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  F. Graybill,et al.  Matrices with Applications in Statistics. , 1984 .

[5]  B. Cipra An introduction to the Ising model , 1987 .

[6]  Wang,et al.  Nonuniversal critical dynamics in Monte Carlo simulations. , 1987, Physical review letters.

[7]  A. Baddeley An Error Metric for Binary Images , 1992 .

[8]  J. Besag,et al.  Spatial Statistics and Bayesian Computation , 1993 .

[9]  S. Chib,et al.  Bayesian analysis of binary and polychotomous response data , 1993 .

[10]  Walter R. Gilks,et al.  Hypothesis testing and model selection , 1995 .

[11]  J. Besag,et al.  On conditional and intrinsic autoregressions , 1995 .

[12]  Karl J. Friston,et al.  Spatial registration and normalization of images , 1995 .

[13]  P. Green Reversible jump Markov chain Monte Carlo computation and Bayesian model determination , 1995 .

[14]  Adrian E. Raftery,et al.  Hypothesis testing and model selection , 1996 .

[15]  Merrilee Hurn Difficulties in the use of auxiliary variables in Markov chain Monte Carlo methods , 1997, Stat. Comput..

[16]  Xavier Descombes,et al.  fMRI Signal Restoration Using a Spatio-Temporal Markov Random Field Preserving Transitions , 1998, NeuroImage.

[17]  Håvard Rue,et al.  Bayesian object recognition with baddeley's delta loss , 1998, Advances in Applied Probability.

[18]  D. Higdon Auxiliary Variable Methods for Markov Chain Monte Carlo with Applications , 1998 .

[19]  Xiao-Li Meng,et al.  Simulating Normalizing Constants: From Importance Sampling to Bridge Sampling to Path Sampling , 1998 .

[20]  Anthony N. Pettitt,et al.  Binary probability maps using a hidden conditional autoregressive Gaussian process with an application to Finnish common toad data , 2000 .

[21]  Christoff Gössl Bayesian Models in Functional Magnetic Resonance Imaging: Approaches for Human Brain Mapping , 2001 .

[22]  Robert Kohn,et al.  Nonparametric regression using linear combinations of basis functions , 2001, Stat. Comput..

[23]  H. Rue Fast sampling of Gaussian Markov random fields , 2000 .

[24]  Anthony N. Pettitt,et al.  A Conditional Autoregressive Gaussian Process for Irregularly Spaced Multivariate Data with Application to Modelling Large Sets of Binary Data , 2002, Stat. Comput..

[25]  Merrilee Hurn Confocal fluorescence microscopy of leaf cells: an application of Bayesian image analysis , 2002 .

[26]  P. Green,et al.  Hidden Markov Models and Disease Mapping , 2002 .

[27]  Ludwig Fahrmeir,et al.  Assessing brain activity through spatial bayesian variable selection , 2003, NeuroImage.

[28]  Gustav Zeuner,et al.  Markov chain Monte Carlo for Bayesian inference , 2007 .