Bayesian Parameter Estimation and Segmentation in the Multi-Atlas Random Orbit Model

Research output: Contribution to journalArticle

Abstract

This paper examines the multiple atlas random diffeomorphic orbit model in Computational Anatomy (CA) for parameter estimation and segmentation of subcortical and ventricular neuroanatomy in magnetic resonance imagery. We assume that there exist multiple magnetic resonance image (MRI) atlases, each atlas containing a collection of locally-defined charts in the brain generated via manual delineation of the structures of interest. We focus on maximum a posteriori estimation of high dimensional segmentations of MR within the class of generative models representing the observed MRI as a conditionally Gaussian random field, conditioned on the atlas charts and the diffeomorphic change of coordinates of each chart that generates it. The charts and their diffeomorphic correspondences are unknown and viewed as latent or hidden variables. We demonstrate that the expectation-maximization (EM) algorithm arises naturally, yielding the likelihood-fusion equation which the a posteriori estimator of the segmentation labels maximizes. The likelihoods being fused are modeled as conditionally Gaussian random fields with mean fields a function of each atlas chart under its diffeomorphic change of coordinates onto the target. The conditional-mean in the EM algorithm specifies the convex weights with which the chart-specific likelihoods are fused. The multiple atlases with the associated convex weights imply that the posterior distribution is a multi-modal representation of the measured MRI. Segmentation results for subcortical and ventricular structures of subjects, within populations of demented subjects, are demonstrated, including the use of multiple atlases across multiple diseased groups.

Original languageEnglish (US)
Article numbere65591
JournalPLoS One
Volume8
Issue number6
DOIs
StatePublished - Jun 18 2013

Fingerprint

Atlases
orbits
Orbit
Magnetic resonance
Parameter estimation
Orbits
magnetic resonance imaging
Magnetic Resonance Spectroscopy
brain
Labels
Brain
Fusion reactions
Weights and Measures
Neuroanatomy
Imagery (Psychotherapy)
Anatomy
Population

ASJC Scopus subject areas

  • Agricultural and Biological Sciences(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Medicine(all)

Cite this

Bayesian Parameter Estimation and Segmentation in the Multi-Atlas Random Orbit Model. / Tang, Xiaoying; Oishi, Kenichi; Faria, Andreia Vasconcellos; Hillis-Trupe, Argye; Albert, Marilyn; Mori, Susumu; Miller, Michael I.

In: PLoS One, Vol. 8, No. 6, e65591, 18.06.2013.

Research output: Contribution to journalArticle

@article{2bd4f847680743fd9300c49dd0dc0454,
title = "Bayesian Parameter Estimation and Segmentation in the Multi-Atlas Random Orbit Model",
abstract = "This paper examines the multiple atlas random diffeomorphic orbit model in Computational Anatomy (CA) for parameter estimation and segmentation of subcortical and ventricular neuroanatomy in magnetic resonance imagery. We assume that there exist multiple magnetic resonance image (MRI) atlases, each atlas containing a collection of locally-defined charts in the brain generated via manual delineation of the structures of interest. We focus on maximum a posteriori estimation of high dimensional segmentations of MR within the class of generative models representing the observed MRI as a conditionally Gaussian random field, conditioned on the atlas charts and the diffeomorphic change of coordinates of each chart that generates it. The charts and their diffeomorphic correspondences are unknown and viewed as latent or hidden variables. We demonstrate that the expectation-maximization (EM) algorithm arises naturally, yielding the likelihood-fusion equation which the a posteriori estimator of the segmentation labels maximizes. The likelihoods being fused are modeled as conditionally Gaussian random fields with mean fields a function of each atlas chart under its diffeomorphic change of coordinates onto the target. The conditional-mean in the EM algorithm specifies the convex weights with which the chart-specific likelihoods are fused. The multiple atlases with the associated convex weights imply that the posterior distribution is a multi-modal representation of the measured MRI. Segmentation results for subcortical and ventricular structures of subjects, within populations of demented subjects, are demonstrated, including the use of multiple atlases across multiple diseased groups.",
author = "Xiaoying Tang and Kenichi Oishi and Faria, {Andreia Vasconcellos} and Argye Hillis-Trupe and Marilyn Albert and Susumu Mori and Miller, {Michael I.}",
year = "2013",
month = "6",
day = "18",
doi = "10.1371/journal.pone.0065591",
language = "English (US)",
volume = "8",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "6",

}

TY - JOUR

T1 - Bayesian Parameter Estimation and Segmentation in the Multi-Atlas Random Orbit Model

AU - Tang, Xiaoying

AU - Oishi, Kenichi

AU - Faria, Andreia Vasconcellos

AU - Hillis-Trupe, Argye

AU - Albert, Marilyn

AU - Mori, Susumu

AU - Miller, Michael I.

PY - 2013/6/18

Y1 - 2013/6/18

N2 - This paper examines the multiple atlas random diffeomorphic orbit model in Computational Anatomy (CA) for parameter estimation and segmentation of subcortical and ventricular neuroanatomy in magnetic resonance imagery. We assume that there exist multiple magnetic resonance image (MRI) atlases, each atlas containing a collection of locally-defined charts in the brain generated via manual delineation of the structures of interest. We focus on maximum a posteriori estimation of high dimensional segmentations of MR within the class of generative models representing the observed MRI as a conditionally Gaussian random field, conditioned on the atlas charts and the diffeomorphic change of coordinates of each chart that generates it. The charts and their diffeomorphic correspondences are unknown and viewed as latent or hidden variables. We demonstrate that the expectation-maximization (EM) algorithm arises naturally, yielding the likelihood-fusion equation which the a posteriori estimator of the segmentation labels maximizes. The likelihoods being fused are modeled as conditionally Gaussian random fields with mean fields a function of each atlas chart under its diffeomorphic change of coordinates onto the target. The conditional-mean in the EM algorithm specifies the convex weights with which the chart-specific likelihoods are fused. The multiple atlases with the associated convex weights imply that the posterior distribution is a multi-modal representation of the measured MRI. Segmentation results for subcortical and ventricular structures of subjects, within populations of demented subjects, are demonstrated, including the use of multiple atlases across multiple diseased groups.

AB - This paper examines the multiple atlas random diffeomorphic orbit model in Computational Anatomy (CA) for parameter estimation and segmentation of subcortical and ventricular neuroanatomy in magnetic resonance imagery. We assume that there exist multiple magnetic resonance image (MRI) atlases, each atlas containing a collection of locally-defined charts in the brain generated via manual delineation of the structures of interest. We focus on maximum a posteriori estimation of high dimensional segmentations of MR within the class of generative models representing the observed MRI as a conditionally Gaussian random field, conditioned on the atlas charts and the diffeomorphic change of coordinates of each chart that generates it. The charts and their diffeomorphic correspondences are unknown and viewed as latent or hidden variables. We demonstrate that the expectation-maximization (EM) algorithm arises naturally, yielding the likelihood-fusion equation which the a posteriori estimator of the segmentation labels maximizes. The likelihoods being fused are modeled as conditionally Gaussian random fields with mean fields a function of each atlas chart under its diffeomorphic change of coordinates onto the target. The conditional-mean in the EM algorithm specifies the convex weights with which the chart-specific likelihoods are fused. The multiple atlases with the associated convex weights imply that the posterior distribution is a multi-modal representation of the measured MRI. Segmentation results for subcortical and ventricular structures of subjects, within populations of demented subjects, are demonstrated, including the use of multiple atlases across multiple diseased groups.

UR - http://www.scopus.com/inward/record.url?scp=84879131763&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84879131763&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0065591

DO - 10.1371/journal.pone.0065591

M3 - Article

C2 - 23824159

AN - SCOPUS:84879131763

VL - 8

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 6

M1 - e65591

ER -