Bayesian Consistent Estimation in Deformable Models using Stochastic Algorithms: Applications to Medical Images

Allassonnière, Stéphanie; Kuhn, Estelle; Trouvé, Alain

Journal de la société française de statistique, Tome 151 (2010) no. 1, pp. 1-16.

Cet article vise à résumer et valider sur données réelles la méthode proposée dans ( 2 , 3 , 4 ) pour l’estimation d’atlas appelé Bayesian Mixed Effect (BME) atlas. Un tel atlas est composé d’une image de référence et d’une métrique pour chaque sous-groupe d’une population ainsi que du poids de ce sous-groupe. L’estimation est consistante sur un échantillon d’images données non labellisées. Nous rappelons ici le modèle statistique génératif qui permet l’estimation simultanée des sous-groupes, de leurs poids, des images de référence et des variabilités géométriques (liées aux métriques). Comme proposé en ( 2 , 3 , 4 ), nous travaillons dans un cadre bayésien, utilisons l’estimateur de Maximum A Posteriori et approchons sa valeur par une variante stochastique de l’algorithme EM (Expectation Maximisation). Cette méthode est validée sur deux ensembles de données d’images médicales : une partie du cortex humain et des excroissances de dendrites de souris liées à la maladies de Parkinson. Nous présentons les performances de cette méthode sur l’estimation de l’image de référence, la variabilité géométrique et le label.

This paper aims at summarising and validating a methodology proposed in [ 2 , 3 , 4 ] for estimating a Bayesian Mixed Effect (BME) atlas, i.e. coupled templates and geometrical metrics for estimated clusters, in a statistically consistent way given a sample of images. We recall the generative statistical model applied to the observations which enables the simultaneous estimation of the clusters, the templates and geometrical variabilities (related to the metric) in the population. Following [ 2 , 3 , 4 ], we work in a Bayesian framework, use a Maximum A Posteriori estimator and approach its value using a stochastic variant of the Expectation Maximisation (EM) algorithm. The method is validated with two data set consisting of medical images of part of the human cortex and dendrite spines from a mouse model of Parkinson’s disease. We present the performances of the method on the estimation of the template, the geometrical variability and the clustering.

[1] Aldridge, G.M.; Ratnanather, J.T.; Martone, M.E.; Terada, M.; Beg, M.F.; Fong, L.; Ceyhan, E.; Kolasny, A.E.; Brown, T.J.A.; Cochran, E.L.; Tang, S.J.; Pisano, D.V.; Vaillant, M.; Hurdal, M.K.; Churchill, J.D; Greenough, W.T.; Miller, M.I.; Ellisman, M.H. Semi-automated shape analysis of dendrite spines from animal models of FragileX and Parkinson’s disease using Large Deformation Diffeomorphic Metric Mapping, Society for Neuroscience Annual Meeting, Washington DC (2005)

[2] Allassonnière, Stéphanie; Amit, Yali; Trouvé, Alain Toward a coherent statistical framework for dense deformable template estimation, JRSS, Volume 69 (2007), pp. 3-29 | MR

[2] Amit, Y.; Grenander, U.; Piccioni, M. Structural image restoration through deformable templates, Journal of the American Statistical Association, Volume 86 (1989), pp. 376-387

[3] Allassonnière, Stéphanie; Kuhn, Estelle; Trouvé, Alain Bayesian Deformable Models Bulding via Stochastic Approximation Algorithm: A convergence Study, In Press in Bernoulli J. | MR | Zbl

[4] Allassonnière, Stéphanie; Kuhn, Estelle Stochastic Algorithm For Bayesian Mixture Effect Template Estimation, In press in ESAIM Probab.Stat. | MR | Zbl

[5] Andrieu, Christophe; Moulines, Éric; Priouret, Pierre Stability of stochastic approximation under verifiable conditions, SIAM J. Control Optim., Volume 44 (2005) no. 1, p. 283-312 (electronic) | MR | Zbl

[7] Ashburner, John A fast diffeomorphic image registration algorithm, NeuroImage, Volume 38 (2007), pp. 95-113

[8] Cootes, T.F.; Edwards, G.J.; Taylor, C.J. Active Appearance Model, IEEE Transactions on Pattern Analysis and Machine Intelligence, Volume 23(6) (2001), pp. 681-685

[9] Ceyhan, E.; Ölken, R.Ç.; Fong, L.; Tasky, T.N.; Hurdal, M.K.; Beg, M.F.; Martone, M.E.; Ratnanather, J.T. Modeling Metric Distances of Dendrite Spines of Mice Based on Morphometric Measures, Int. Symp on Health Informatics and Bioinformatics (2007)

[10] Ceyhan, E.; Fong, L.; Tasky, T.N.; Hurdal, M.K.; Beg, M.E. M.F.and Martone; Ratnanather, J.T. Type-Specific Analysis of Morphometry of Dendrite Spines of Mice, 5th Int. Symp. Image Signal Proc. Analysis, ISPA (2007), pp. 7-12

[11] Delyon, Bernard; Lavielle, Marc; Moulines, Eric Convergence of a stochastic approximation version of the EM algorithm, Ann. Statist., Volume 27 (1999) no. 1, pp. 94-128 | MR | Zbl

[12] Dempster, A. P.; Laird, N. M.; Rubin, D. B. Maximum likelihood from incomplete data via the EM algorithm, Journal of the Royal Statistical Society, Volume 1 (1977), pp. 1-22 | MR | Zbl

[13] Glaunès, Joan; Joshi, Sarang Template estimation form unlabeled point set data and surfaces for Computational Anatomy, Proc. of the International Workshop on the Mathematical Foundations of Computational Anatomy (MFCA) (Pennec, X.; Joshi, S., eds.) (2006), pp. 29-39

[14] Glasbey, C. A.; Mardia, K. V. A penalised likelihood approach to image warping, Journal of the Royal Statistical Society, Series B, Volume 63 (2001), pp. 465-492 | MR | Zbl

[15] Kuhn, Estelle; Lavielle, Marc Coupling a stochastic approximation version of EM with an MCMC procedure, ESAIM Probab. Stat., Volume 8 (2004), p. 115-131 (electronic) | Numdam | MR | Zbl

[16] Miller, M. I.; Trouvé, A.; Younes, L. On the Metrics and Euler-Lagrange Equations of Computational Anatomy, Annual Review of biomedical Engineering, Volume 4 (2002)

[17] Richard, Frédéric.; Samson, Adeline.; Cuenod, Charles A. A SAEM algorithm for the estimation of template and deformation parameters in medical image sequences, Statistics and Computing, Volume 19 (2009), pp. 465-478 | MR

[18] Sabuncu, Mert; Balci, Serdar K.; Golland, Polina Discovering Modes of an Image Population through Mixture Modeling, Proceeding of the MICCAI conference, Volume LNCS (2008) no. 5242, pp. 381-389

[19] Twining, Carole; Cootes, Tim; Marsland, Stephen; Petrovic, Vladimir; Schestowitz, Roy; Taylor, Chris Information-Theoretic Unification of Groupwise Non-Rigid Registration and Model Building, Proceedings of Medical Image Understanding and Analysis (MIUA), Volume 2 (2006), pp. 226-230

[20] Vaida, Florin Parameter convergence for EM and MM algorithms, Statist. Sinica, Volume 15 (2005) no. 3, pp. 831-840 | MR | Zbl

[21] Vercauteren, Tom; Pennec, Xavier; Perchant, Aymeric; Ayache, Nicholas Diffeomorphic demons: Efficient non-parametric image registration., Neuroimage, Volume 45 (2009), pp. 61-72