A generative probability model of joint label fusion for multi-atlas based brain segmentation

Guorong Wu, Qian Wang, Daoqiang Zhang, Feiping Nie, Heng Huang, Dinggang Shen

Research output: Contribution to journalArticle

80 Citations (Scopus)

Abstract

Automated labeling of anatomical structures in medical images is very important in many neuroscience studies. Recently, patch-based labeling has been widely investigated to alleviate the possible mis-alignment when registering atlases to the target image. However, the weights used for label fusion from the registered atlases are generally computed independently and thus lack the capability of preventing the ambiguous atlas patches from contributing to the label fusion. More critically, these weights are often calculated based only on the simple patch similarity, thus not necessarily providing optimal solution for label fusion. To address these limitations, we propose a generative probability model to describe the procedure of label fusion in a multi-atlas scenario, for the goal of labeling each point in the target image by the best representative atlas patches that also have the largest labeling unanimity in labeling the underlying point correctly. Specifically, sparsity constraint is imposed upon label fusion weights, in order to select a small number of atlas patches that best represent the underlying target patch, thus reducing the risks of including the misleading atlas patches. The labeling unanimity among atlas patches is achieved by exploring their dependencies, where we model these dependencies as the joint probability of each pair of atlas patches in correctly predicting the labels, by analyzing the correlation of their morphological error patterns and also the labeling consensus among atlases. The patch dependencies will be further recursively updated based on the latest labeling results to correct the possible labeling errors, which falls to the Expectation Maximization (EM) framework. To demonstrate the labeling performance, we have comprehensively evaluated our patch-based labeling method on the whole brain parcellation and hippocampus segmentation. Promising labeling results have been achieved with comparison to the conventional patch-based labeling method, indicating the potential application of the proposed method in the future clinical studies.

Original languageEnglish
Pages (from-to)881-890
Number of pages10
JournalMedical Image Analysis
Volume18
Issue number6
DOIs
Publication statusPublished - 2014 Jan 1
Externally publishedYes

Fingerprint

Atlases
Labeling
Labels
Brain
Fusion reactions
Joints
Weights and Measures
Neurosciences
Hippocampus

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Vision and Pattern Recognition
  • Radiology Nuclear Medicine and imaging
  • Health Informatics
  • Radiological and Ultrasound Technology

Cite this

A generative probability model of joint label fusion for multi-atlas based brain segmentation. / Wu, Guorong; Wang, Qian; Zhang, Daoqiang; Nie, Feiping; Huang, Heng; Shen, Dinggang.

In: Medical Image Analysis, Vol. 18, No. 6, 01.01.2014, p. 881-890.

Research output: Contribution to journalArticle

Wu, Guorong ; Wang, Qian ; Zhang, Daoqiang ; Nie, Feiping ; Huang, Heng ; Shen, Dinggang. / A generative probability model of joint label fusion for multi-atlas based brain segmentation. In: Medical Image Analysis. 2014 ; Vol. 18, No. 6. pp. 881-890.
@article{49e8105832aa4bdab1679ef024e6e384,
title = "A generative probability model of joint label fusion for multi-atlas based brain segmentation",
abstract = "Automated labeling of anatomical structures in medical images is very important in many neuroscience studies. Recently, patch-based labeling has been widely investigated to alleviate the possible mis-alignment when registering atlases to the target image. However, the weights used for label fusion from the registered atlases are generally computed independently and thus lack the capability of preventing the ambiguous atlas patches from contributing to the label fusion. More critically, these weights are often calculated based only on the simple patch similarity, thus not necessarily providing optimal solution for label fusion. To address these limitations, we propose a generative probability model to describe the procedure of label fusion in a multi-atlas scenario, for the goal of labeling each point in the target image by the best representative atlas patches that also have the largest labeling unanimity in labeling the underlying point correctly. Specifically, sparsity constraint is imposed upon label fusion weights, in order to select a small number of atlas patches that best represent the underlying target patch, thus reducing the risks of including the misleading atlas patches. The labeling unanimity among atlas patches is achieved by exploring their dependencies, where we model these dependencies as the joint probability of each pair of atlas patches in correctly predicting the labels, by analyzing the correlation of their morphological error patterns and also the labeling consensus among atlases. The patch dependencies will be further recursively updated based on the latest labeling results to correct the possible labeling errors, which falls to the Expectation Maximization (EM) framework. To demonstrate the labeling performance, we have comprehensively evaluated our patch-based labeling method on the whole brain parcellation and hippocampus segmentation. Promising labeling results have been achieved with comparison to the conventional patch-based labeling method, indicating the potential application of the proposed method in the future clinical studies.",
keywords = "Generative probability model, Multi-atlas based segmentation, Patch-based labeling, Sparse representation",
author = "Guorong Wu and Qian Wang and Daoqiang Zhang and Feiping Nie and Heng Huang and Dinggang Shen",
year = "2014",
month = "1",
day = "1",
doi = "10.1016/j.media.2013.10.013",
language = "English",
volume = "18",
pages = "881--890",
journal = "Medical Image Analysis",
issn = "1361-8415",
publisher = "Elsevier",
number = "6",

}

TY - JOUR

T1 - A generative probability model of joint label fusion for multi-atlas based brain segmentation

AU - Wu, Guorong

AU - Wang, Qian

AU - Zhang, Daoqiang

AU - Nie, Feiping

AU - Huang, Heng

AU - Shen, Dinggang

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Automated labeling of anatomical structures in medical images is very important in many neuroscience studies. Recently, patch-based labeling has been widely investigated to alleviate the possible mis-alignment when registering atlases to the target image. However, the weights used for label fusion from the registered atlases are generally computed independently and thus lack the capability of preventing the ambiguous atlas patches from contributing to the label fusion. More critically, these weights are often calculated based only on the simple patch similarity, thus not necessarily providing optimal solution for label fusion. To address these limitations, we propose a generative probability model to describe the procedure of label fusion in a multi-atlas scenario, for the goal of labeling each point in the target image by the best representative atlas patches that also have the largest labeling unanimity in labeling the underlying point correctly. Specifically, sparsity constraint is imposed upon label fusion weights, in order to select a small number of atlas patches that best represent the underlying target patch, thus reducing the risks of including the misleading atlas patches. The labeling unanimity among atlas patches is achieved by exploring their dependencies, where we model these dependencies as the joint probability of each pair of atlas patches in correctly predicting the labels, by analyzing the correlation of their morphological error patterns and also the labeling consensus among atlases. The patch dependencies will be further recursively updated based on the latest labeling results to correct the possible labeling errors, which falls to the Expectation Maximization (EM) framework. To demonstrate the labeling performance, we have comprehensively evaluated our patch-based labeling method on the whole brain parcellation and hippocampus segmentation. Promising labeling results have been achieved with comparison to the conventional patch-based labeling method, indicating the potential application of the proposed method in the future clinical studies.

AB - Automated labeling of anatomical structures in medical images is very important in many neuroscience studies. Recently, patch-based labeling has been widely investigated to alleviate the possible mis-alignment when registering atlases to the target image. However, the weights used for label fusion from the registered atlases are generally computed independently and thus lack the capability of preventing the ambiguous atlas patches from contributing to the label fusion. More critically, these weights are often calculated based only on the simple patch similarity, thus not necessarily providing optimal solution for label fusion. To address these limitations, we propose a generative probability model to describe the procedure of label fusion in a multi-atlas scenario, for the goal of labeling each point in the target image by the best representative atlas patches that also have the largest labeling unanimity in labeling the underlying point correctly. Specifically, sparsity constraint is imposed upon label fusion weights, in order to select a small number of atlas patches that best represent the underlying target patch, thus reducing the risks of including the misleading atlas patches. The labeling unanimity among atlas patches is achieved by exploring their dependencies, where we model these dependencies as the joint probability of each pair of atlas patches in correctly predicting the labels, by analyzing the correlation of their morphological error patterns and also the labeling consensus among atlases. The patch dependencies will be further recursively updated based on the latest labeling results to correct the possible labeling errors, which falls to the Expectation Maximization (EM) framework. To demonstrate the labeling performance, we have comprehensively evaluated our patch-based labeling method on the whole brain parcellation and hippocampus segmentation. Promising labeling results have been achieved with comparison to the conventional patch-based labeling method, indicating the potential application of the proposed method in the future clinical studies.

KW - Generative probability model

KW - Multi-atlas based segmentation

KW - Patch-based labeling

KW - Sparse representation

UR - http://www.scopus.com/inward/record.url?scp=84901691837&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84901691837&partnerID=8YFLogxK

U2 - 10.1016/j.media.2013.10.013

DO - 10.1016/j.media.2013.10.013

M3 - Article

C2 - 24315359

AN - SCOPUS:84901691837

VL - 18

SP - 881

EP - 890

JO - Medical Image Analysis

JF - Medical Image Analysis

SN - 1361-8415

IS - 6

ER -