A generative model for image segmentation based on label fusion

We propose a nonparametric, probabilistic model for the automatic segmentation of medical images, given a training set of images and corresponding label maps. The resulting inference algorithms rely on pairwise registrations between the test image and individual training images. The training labels...

Full description

Bibliographic Details
Main Authors: Sabuncu, Mert R., Yeo, Boon Thye Thomas, Fischl, Bruce, Van Leemput, Koen, Golland, Polina
Other Authors: Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Format: Article
Language:en_US
Published: Institute of Electrical and Electronics Engineers 2011
Online Access:http://hdl.handle.net/1721.1/64791
https://orcid.org/0000-0002-5002-1227
https://orcid.org/0000-0003-2516-731X
_version_ 1826215550775197696
author Sabuncu, Mert R.
Yeo, Boon Thye Thomas
Fischl, Bruce
Van Leemput, Koen
Golland, Polina
author2 Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
author_facet Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Sabuncu, Mert R.
Yeo, Boon Thye Thomas
Fischl, Bruce
Van Leemput, Koen
Golland, Polina
author_sort Sabuncu, Mert R.
collection MIT
description We propose a nonparametric, probabilistic model for the automatic segmentation of medical images, given a training set of images and corresponding label maps. The resulting inference algorithms rely on pairwise registrations between the test image and individual training images. The training labels are then transferred to the test image and fused to compute the final segmentation of the test subject. Such label fusion methods have been shown to yield accurate segmentation, since the use of multiple registrations captures greater inter-subject anatomical variability and improves robustness against occasional registration failures. To the best of our knowledge, this manuscript presents the first comprehensive probabilistic framework that rigorously motivates label fusion as a segmentation approach. The proposed framework allows us to compare different label fusion algorithms theoretically and practically. In particular, recent label fusion or multiatlas segmentation algorithms are interpreted as special cases of our framework. We conduct two sets of experiments to validate the proposed methods. In the first set of experiments, we use 39 brain MRI scans - with manually segmented white matter, cerebral cortex, ventricles and subcortical structures - to compare different label fusion algorithms and the widely-used FreeSurfer whole-brain segmentation tool. Our results indicate that the proposed framework yields more accurate segmentation than FreeSurfer and previous label fusion algorithms. In a second experiment, we use brain MRI scans of 282 subjects to demonstrate that the proposed segmentation tool is sufficiently sensitive to robustly detect hippocampal volume changes in a study of aging and Alzheimer's Disease.
first_indexed 2024-09-23T16:34:44Z
format Article
id mit-1721.1/64791
institution Massachusetts Institute of Technology
language en_US
last_indexed 2024-09-23T16:34:44Z
publishDate 2011
publisher Institute of Electrical and Electronics Engineers
record_format dspace
spelling mit-1721.1/647912022-10-02T08:19:23Z A generative model for image segmentation based on label fusion Sabuncu, Mert R. Yeo, Boon Thye Thomas Fischl, Bruce Van Leemput, Koen Golland, Polina Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Golland, Polina Golland, Polina Van Leemput, Koen Sabuncu, Mert R. Yeo, Boon Thye Thomas Fischl, Bruce We propose a nonparametric, probabilistic model for the automatic segmentation of medical images, given a training set of images and corresponding label maps. The resulting inference algorithms rely on pairwise registrations between the test image and individual training images. The training labels are then transferred to the test image and fused to compute the final segmentation of the test subject. Such label fusion methods have been shown to yield accurate segmentation, since the use of multiple registrations captures greater inter-subject anatomical variability and improves robustness against occasional registration failures. To the best of our knowledge, this manuscript presents the first comprehensive probabilistic framework that rigorously motivates label fusion as a segmentation approach. The proposed framework allows us to compare different label fusion algorithms theoretically and practically. In particular, recent label fusion or multiatlas segmentation algorithms are interpreted as special cases of our framework. We conduct two sets of experiments to validate the proposed methods. In the first set of experiments, we use 39 brain MRI scans - with manually segmented white matter, cerebral cortex, ventricles and subcortical structures - to compare different label fusion algorithms and the widely-used FreeSurfer whole-brain segmentation tool. Our results indicate that the proposed framework yields more accurate segmentation than FreeSurfer and previous label fusion algorithms. In a second experiment, we use brain MRI scans of 282 subjects to demonstrate that the proposed segmentation tool is sufficiently sensitive to robustly detect hippocampal volume changes in a study of aging and Alzheimer's Disease. National Alliance for Medical Image Computing (U.S.) (NIH NIBIB NAMIC U54-EB005149) National Alliance for Medical Image Computing (U.S.) (NAC NIH NCRR NAC P41-RR13218) National Institutes of Health (U.S.) (mBIRN NIH NCRR mBIRN U24-RR021382) National Institutes of Health (U.S.) (NIH NINDS R01-NS051826 grant) National Science Foundation (U.S.) (NSF CAREER 0642971 grant) National Center for Research Resources (U.S.) (P41-RR14075, R01 RR16594-01A1) National Institute of Biomedical Imaging and Bioengineering (U.S.) (R01 EB001550, R01EB006758) National Institute of Neurological Disorders and Stroke (U.S.) (R01 NS052585-01) Mind Research Institute Ellison Medical Foundation (Autism and Dyslexia Project) Singapore. Agency for Science, Technology and Research Academy of Finland (grant number 133611) 2011-07-13T18:13:40Z 2011-07-13T18:13:40Z 2010-09 2010-04 Article http://purl.org/eprint/type/ConferencePaper 0278-0062 INSPEC Accession Number: 11534908 http://hdl.handle.net/1721.1/64791 Sabuncu, M.R. et al. “A Generative Model for Image Segmentation Based on Label Fusion.” Medical Imaging, IEEE Transactions On 29.10 (2010) : 1714-1729.© 2010 IEEE. https://orcid.org/0000-0002-5002-1227 https://orcid.org/0000-0003-2516-731X en_US http://dx.doi.org/ 10.1109/TMI.2010.2050897 IEEE Transactions on Medical Imaging, 2010 Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf Institute of Electrical and Electronics Engineers MIT web domain
spellingShingle Sabuncu, Mert R.
Yeo, Boon Thye Thomas
Fischl, Bruce
Van Leemput, Koen
Golland, Polina
A generative model for image segmentation based on label fusion
title A generative model for image segmentation based on label fusion
title_full A generative model for image segmentation based on label fusion
title_fullStr A generative model for image segmentation based on label fusion
title_full_unstemmed A generative model for image segmentation based on label fusion
title_short A generative model for image segmentation based on label fusion
title_sort generative model for image segmentation based on label fusion
url http://hdl.handle.net/1721.1/64791
https://orcid.org/0000-0002-5002-1227
https://orcid.org/0000-0003-2516-731X
work_keys_str_mv AT sabuncumertr agenerativemodelforimagesegmentationbasedonlabelfusion
AT yeoboonthyethomas agenerativemodelforimagesegmentationbasedonlabelfusion
AT fischlbruce agenerativemodelforimagesegmentationbasedonlabelfusion
AT vanleemputkoen agenerativemodelforimagesegmentationbasedonlabelfusion
AT gollandpolina agenerativemodelforimagesegmentationbasedonlabelfusion
AT sabuncumertr generativemodelforimagesegmentationbasedonlabelfusion
AT yeoboonthyethomas generativemodelforimagesegmentationbasedonlabelfusion
AT fischlbruce generativemodelforimagesegmentationbasedonlabelfusion
AT vanleemputkoen generativemodelforimagesegmentationbasedonlabelfusion
AT gollandpolina generativemodelforimagesegmentationbasedonlabelfusion