Convergent stochastic expectation maximization algorithm with efficient sampling in high dimension. Application to deformable template model estimation

DOI10.1016/J.CSDA.2015.04.011zbMATH Open1468.62016arXiv1207.5938OpenAlexW1951747706MaRDI QIDQ1663188FDOQ1663188

Estelle Kuhn, Stéphanie Allassonnière

Publication date: 21 August 2018

Published in: Computational Statistics and Data Analysis (Search for Journal in Brave)

Abstract: Estimation in the deformable template model is a big challenge in image analysis. The issue is to estimate an atlas of a population. This atlas contains a template and the corresponding geometrical variability of the observed shapes. The goal is to propose an accurate algorithm with low computational cost and with theoretical guaranties of relevance. This becomes very demanding when dealing with high dimensional data which is particularly the case of medical images. We propose to use an optimized Monte Carlo Markov Chain method into a stochastic Expectation Maximization algorithm in order to estimate the model parameters by maximizing the likelihood. In this paper, we present a new Anisotropic Metropolis Adjusted Langevin Algorithm which we use as transition in the MCMC method. We first prove that this new sampler leads to a geometrically uniformly ergodic Markov chain. We prove also that under mild conditions, the estimated parameters converge almost surely and are asymptotically Gaussian distributed. The methodology developed is then tested on handwritten digits and some 2D and 3D medical images for the deformable model estimation. More widely, the proposed algorithm can be used for a large range of models in many fields of applications such as pharmacology or genetic.

Full work available at URL: https://arxiv.org/abs/1207.5938

Recommendations

• Construction of Bayesian deformable models via a stochastic approximation algorithm: a convergence study
• A new class of stochastic EM algorithms. Escaping local maxima and handling intractable sampling
• Towards a Coherent Statistical Framework for Dense Deformable Template Estimation
• Convergence of a stochastic approximation version of the EM algorithm
• scientific article; zbMATH DE number 6458272
• Stochastic algorithm for Bayesian mixture effect template estimation
• scientific article; zbMATH DE number 788229

zbMATH Keywords

maximum likelihood estimationMCMCstochastic EM algorithmhigh dimensiondeformable templategeometric variabilityanisotropic MALAmissing variable

Mathematics Subject Classification ID

Computational methods for problems pertaining to statistics (62-08) Applications of statistics to biology and medical sciences; meta analysis (62P10)

Cites Work

• Title not available (Why is that?)
• Exponential convergence of Langevin distributions and their discrete approximations
• Title not available (Why is that?)
• Convergence of a stochastic approximation version of the EM algorithm
• Geometric ergodicity of Metropolis algorithms
• Stability of Stochastic Approximation under Verifiable Conditions
• Markov chains and stochastic stability
• On the ergodicity properties of some adaptive MCMC algorithms
• Maximum likelihood estimation in nonlinear mixed effects models
• Riemann Manifold Langevin and Hamiltonian Monte Carlo Methods
• An adaptive version for the Metropolis adjusted Langevin algorithm with a truncated drift
• A penalized likelihood approach to image warping. (With discussion).
• Langevin-type models II: Self-targeting candidates for MCMC algorithms
• On the consistency of Fréchet means in deformable models for curve and image analysis
• Stochastic algorithm for Bayesian mixture effect template estimation
• Computational anatomy: an emerging discipline
• Coupling a stochastic approximation version of EM with an MCMC procedure
• Construction of Bayesian deformable models via a stochastic approximation algorithm: a convergence study
• Langevin-type models. I: Diffusions with given stationary distributions and their discretizations
• Title not available (Why is that?)
• Detection, classification and estimation of individual shapes in 2D and 3D point clouds
• An adaptive approach to Langevin MCMC
• Towards a Coherent Statistical Framework for Dense Deformable Template Estimation

Cited In (6)

• Estimating blood vessel areas in ultrasound images using a deformable template model
• Online EM for functional data
• Construction of Bayesian deformable models via a stochastic approximation algorithm: a convergence study
• Statistical models for deformable templates in image and shape analysis
• Template Shape Estimation: Correcting an Asymptotic Bias
• Bayesian Mixed Effect Atlas Estimation with a Diffeomorphic Deformation Model

Uses Software

• LDDMM

This page was built for publication: Convergent stochastic expectation maximization algorithm with efficient sampling in high dimension. Application to deformable template model estimation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1663188)