The Annals of Applied Statistics

Quantitative magnetic resonance image analysis via the EM algorithm with stochastic variation

Xiaoxi Zhang, Timothy D. Johnson, Roderick J. A. Little, and Yue Cao

Full-text: Open access


Quantitative Magnetic Resonance Imaging (qMRI) provides researchers insight into pathological and physiological alterations of living tissue, with the help of which researchers hope to predict (local) therapeutic efficacy early and determine optimal treatment schedule. However, the analysis of qMRI has been limited to ad-hoc heuristic methods. Our research provides a powerful statistical framework for image analysis and sheds light on future localized adaptive treatment regimes tailored to the individual’s response. We assume in an imperfect world we only observe a blurred and noisy version of the underlying pathological/physiological changes via qMRI, due to measurement errors or unpredictable influences. We use a hidden Markov random field to model the spatial dependence in the data and develop a maximum likelihood approach via the Expectation–Maximization algorithm with stochastic variation. An important improvement over previous work is the assessment of variability in parameter estimation, which is the valid basis for statistical inference. More importantly, we focus on the expected changes rather than image segmentation. Our research has shown that the approach is powerful in both simulation studies and on a real dataset, while quite robust in the presence of some model assumption violations.

Article information

Ann. Appl. Stat., Volume 2, Number 2 (2008), 736-755.

First available in Project Euclid: 3 July 2008

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

EM algorithm hidden Markov random field missing data model selection quantitative MRI stochastic variation


Zhang, Xiaoxi; Johnson, Timothy D.; Little, Roderick J. A.; Cao, Yue. Quantitative magnetic resonance image analysis via the EM algorithm with stochastic variation. Ann. Appl. Stat. 2 (2008), no. 2, 736--755. doi:10.1214/07-AOAS157.

Export citation


  • Akaike, H. (1973). Information theory and an extension of the maximum likelihood principle. In, Second International Symposium on Information Theory (Tsahkasdor, 1971) 267–281. Akadémiai Kiadó, Budapest.
  • Besag, J. E. (1974). Spatial interaction and the statistical analysis of lattice systems (with discussion)., J. Roy. Statist. Soc. Ser. B 36 192–236.
  • Besag, J. E. (1986). On the statistical analysis of dirty pictures., J. Roy. Statist. Soc. Ser. B 48 259–302.
  • Buckland, S. T., Burnham, K. P. and Augustin, N. H. (1997). Model selection: An integral part of inference., Biometrics 53 603–618.
  • Burnham, K. P. and Anderson, D. R. (2002)., Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach, 2nd ed. Springer, New York.
  • Cao, Y., Tsien, C. I., Shen, Z., Tatro, D. S., Ten Haken, R., Kessler, M. L., Chenevert, T. L. and Lawrence, T. S. (2005). Use of magnetic resonance imaging to assess blood–brain/blood–glioma barrier opening during conformal radiotherapy., J. Clinical Oncology 23 4127–4136.
  • Cappe, O., Moulines, E. and Ryden, T. (2005)., Inference in Hidden Markov Models. Springer, New York.
  • Celeux, G. and Diebolt, J. (1985). The SEM algorithm: A probabilistic teacher algorithm derived from the EM algorithm for the mixture problem., Comput. Statist. Quarterly 2 73–82.
  • Chalmond, B. (1989). An iterative Gibbsian technique for reconstruction of, M-ary images. Pattern Recognition 22 747–761.
  • Claeskens, G. and Hjort, N. L. (2003). The focused information criterion (with discussion)., J. Amer. Statist. Assoc. 98 900–945.
  • Delyon, B., Lavielle, M. and Moulines, E. (1999). Convergence of a stochastic approximation version of the EM algorithm., Ann. Statist. 27 94–128.
  • Dempster, A. P., Laird, N. M. and Rubin, D. B. (1977). Maximum likelihood from incomplete data via the EM algorithm., J. Roy. Statist. Soc. Ser. B 39 1–38.
  • Descombes, X., Kruggel, F. and von Cramon, D. Y. (1998a). Spatio-temporal fMRI analysis using Markov random fields., IEEE Trans. Medical Imaging 17 1028–1039.
  • Descombes, X., Kruggel, F. and von Cramon, D. Y. (1998b). fMRI signal restoration using a spatio-temporal Markov random field preserving transitions., NeuroImage 8 340–349.
  • Deng, H. and Clausi, D. A. (2004). Unsupervised image segmentation using a simple MRF model with a new implementation scheme., Pattern Recognition 37 2323–2335.
  • Destrempes, F., Mignotte, M. and Angers, J. F. (2005). A stochastic method for Bayesian estimation of hidden Markov random field models with application to a color model., IEEE Trans. Image Process. 14 1096–1108.
  • Francois, O. (2002). Global optimization with exploration/selection algorithms and simulated annealing., Ann. Appl. Probab. 12 248–271.
  • Hamstra, D. A., Chenevert, T. L., Moffat, B. A., Johnson, T. D., Meyer, C. R., Mukherji, S. K., Quint, D. J., Gebarski, S. S., Fan, X., Tsien, C. I., Lawrence, T. S., Junck, L., Rehemtulla, A. and Ross, B. D. (2005). Evaluation of the functional diffusion map as an early biomarker of time-to-progression and overall survival in high-grade glioma., Proc. Natl. Acad. Sci. 46 16759–16764.
  • Hjort, N. L. and Claeskens, G. (2003). Frequentist model average estimators., J. Amer. Statist. Assoc. 98 879–899.
  • Green, P. J. (1995). Reversible jump Markov chain Monte Carlo computation and Bayesian model determination., Biometrica 8 711–732.
  • Lakshmanan, S. and Derin, H. (1989). Simultaneous parameter estimation and segmentation of Gibbs random fields using simulated annealing., IEEE Trans. Pattern Analysis and Machines Intelligence 2 799–813.
  • Little, R. J. A. and Rubin, D. B. (1983). On jointly estimating parameters and missing data by maximizing the complete-data likelihood., American Statistician 37 218–220.
  • Lei, T. and Udupa, J. K. (2002). Statistical properties of X-ray CT and MRI from imaging physics to image statistics., Proceedings of the SPIE, Medical Imaging 4682 82–93.
  • Lei, T. and Udupa, J. K. (2003). Image analysis: Recovering the true scene from noisy images., Proceedings of the SPIE, Image Processing 5032 1719–1727.
  • Liang, Z. P. and Lauterbur, P. C. (1999)., Principles of Magnetic Resonance Imaging: A Signal Processing Perspective. Wiley, New York.
  • Louis, T. (1982). Finding the observed information matrix when using the EM algorithm., J. Roy. Statist. Soc. Ser. B 44 226–233.
  • Melas, D. E. and Wilson, S. P. (2002). Double Markov random fields and Bayesian image segmentation., IEEE Trans. Signal Process. 50 357–365.
  • Moffat, B. A., Chenevert, T. L., Lawrence, T. S., Meyer, C. R., Johnson, T. D., Dong, Q., Tsien, C., Mukherji, S., Quint, D. J., Gebarski, S. S., Robertson, P. L., Junck, L. R., Rehemtulla, A. and Ross, B. D. (2005). Functional diffusion map: A noninvasive MRI biomarker for early stratification of clinical brain tumor response., Proc. Natl. Acad. Sci. USA 102 5524–5529.
  • Panjwani, D. K. and Healey, G. (1995). Markov random field models for unsupervised segmentation of textures color images., IEEE Trans. on Pattern Analysis and Machines Intelligence 17 939–954.
  • Potts, R. B. (1952). Some generalized order-disorder transformations., Proc. Cambridge Philos. Soc. 48 106–109.
  • Rajapakse, J. C. and Piyaratna, J. (2001). Bayesian approach to segmentation of statistical parametric maps., IEEE Trans. Biomedical Engineering 48 1186–1194.
  • Rubin, D. B. (1987)., Multiple Imputation for Nonresponse in Sample Surveys. Wiley, New York.
  • Schwarz, G. (1978). Estimating the dimension of a model., Ann. Statist. 6 461–464.
  • Sengur, A., Turkoglu, I. and Ince, C. M. (2006). Unsupervised image segmentation using Markov random fields., Lecture Notes in Artificial Intelligence 3949 158–167. Springer, Berlin.
  • Svensen, M., Kruggel, F. and Cramon, D. Y. (2000). Probabilistic modeling of single-trial fMRI data., IEEE Trans. Medical Imaging 19 25–35.
  • Swendsen, R. H. and Wang, J. S. (1987). Nonuniversal critical dynamics in Monte Carlo simulations., Phys. Rev. Lett. 58 86–88.
  • Wang, Y. and Rajapakse, J. (2006). Contextual modeling of function MR images with conditional random fields., IEEE Trans. Medical Imaging 25 804–812.
  • Wei, G. C. G. and Tanner, M. A. (1990). A Monte Carlo implementation of the EM algorithm and the poor man’s data augmentation algorithms., J. Amer. Statist. Assoc. 85 699–704.
  • Won, C. S. and Derin, H. (1992). Unsupervised segmentation of noisy and textured images using Markov random fields., Graphical Models and Image Processing 54 308–328.
  • Woolrich, M. W., Behrens, T. E., Beckmann, C. F. and Smith, S. M. (2005). Mixture models with adaptive spatial regularization for segmentation with an application to fMRI data., IEEE Trans. Medical Imaging 24 1–11.
  • Worsley, K. J. and Friston, K. J. (1995). Analysis of fMRI time-series revisited—Again., NeuroImage 2 173–181.
  • Zhang, J., Modestino, J. W. and Langan, D. A. (1994). Maximum-likelihood parameter estimation for unsupervised stochastic model-based image segmentation., IEEE Trans. Image Process. 3 404–420.
  • Zhang, Y., Brady, M. and Smith, S. (2001). Segmentation of brain MR images through a hidden Markov random field model and the expectation–maximization algorithm., IEEE Trans. Medical Imaging 20 45–57.