Multiple Sclerosis Lesion Segmentation Using Joint Label Fusion

Mengjin Dong, Ipek Oguz, Nagesh Subbana, Peter Calabresi, Russell T. Shinohara, Paul Yushkevich

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper adapts the joint label fusion (JLF) multi-atlas image segmentation algorithm to the problem of multiple sclerosis (MS) lesion segmentation in multi-modal MRI. Conventionally, JLF requires a set of atlas images to be co-registered to the target image using deformable registration. However, given the variable spatial distribution of lesions in the brain, whole-brain deformable registration is unlikely to line up lesions between atlases and the target image. As a solution, we propose to first pre-segment the target image using an intensity regression based technique, yielding a set of “candidate” lesions. Each “candidate” lesion is then matched to a set of similar lesions in the atlas based on location and size; and deformable registration and JLF are applied at the level of the “candidate” lesion. The approach is evaluated on a dataset of 74 subjects with MS and shown to improve Dice similarity coefficient with reference manual segmentation by 12% over intensity regression technique.

Original languageEnglish (US)
Title of host publicationPatch-Based Techniques in Medical Imaging - 3rd International Workshop, Patch-MI 2017 Held in Conjunction with MICCAI 2017, Proceedings
PublisherSpringer Verlag
Pages138-145
Number of pages8
Volume10530 LNCS
ISBN (Print)9783319674339
DOIs
StatePublished - 2017
Event3rd International Workshop on Patch-Based Techniques in Medical Imaging, Patch-MI 2017 held in conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017 - Quebec City, Canada
Duration: Sep 14 2017Sep 14 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10530 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other3rd International Workshop on Patch-Based Techniques in Medical Imaging, Patch-MI 2017 held in conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017
CountryCanada
CityQuebec City
Period9/14/179/14/17

Fingerprint

Multiple Sclerosis
Atlas
Labels
Fusion
Fusion reactions
Segmentation
Registration
Brain
Target
Regression
Similarity Coefficient
Image segmentation
Dice
Magnetic resonance imaging
Spatial distribution
Spatial Distribution
Image Segmentation
Line

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Dong, M., Oguz, I., Subbana, N., Calabresi, P., Shinohara, R. T., & Yushkevich, P. (2017). Multiple Sclerosis Lesion Segmentation Using Joint Label Fusion. In Patch-Based Techniques in Medical Imaging - 3rd International Workshop, Patch-MI 2017 Held in Conjunction with MICCAI 2017, Proceedings (Vol. 10530 LNCS, pp. 138-145). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10530 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-67434-6_16

Multiple Sclerosis Lesion Segmentation Using Joint Label Fusion. / Dong, Mengjin; Oguz, Ipek; Subbana, Nagesh; Calabresi, Peter; Shinohara, Russell T.; Yushkevich, Paul.

Patch-Based Techniques in Medical Imaging - 3rd International Workshop, Patch-MI 2017 Held in Conjunction with MICCAI 2017, Proceedings. Vol. 10530 LNCS Springer Verlag, 2017. p. 138-145 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10530 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Dong, M, Oguz, I, Subbana, N, Calabresi, P, Shinohara, RT & Yushkevich, P 2017, Multiple Sclerosis Lesion Segmentation Using Joint Label Fusion. in Patch-Based Techniques in Medical Imaging - 3rd International Workshop, Patch-MI 2017 Held in Conjunction with MICCAI 2017, Proceedings. vol. 10530 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10530 LNCS, Springer Verlag, pp. 138-145, 3rd International Workshop on Patch-Based Techniques in Medical Imaging, Patch-MI 2017 held in conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017, Quebec City, Canada, 9/14/17. https://doi.org/10.1007/978-3-319-67434-6_16
Dong M, Oguz I, Subbana N, Calabresi P, Shinohara RT, Yushkevich P. Multiple Sclerosis Lesion Segmentation Using Joint Label Fusion. In Patch-Based Techniques in Medical Imaging - 3rd International Workshop, Patch-MI 2017 Held in Conjunction with MICCAI 2017, Proceedings. Vol. 10530 LNCS. Springer Verlag. 2017. p. 138-145. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-67434-6_16
Dong, Mengjin ; Oguz, Ipek ; Subbana, Nagesh ; Calabresi, Peter ; Shinohara, Russell T. ; Yushkevich, Paul. / Multiple Sclerosis Lesion Segmentation Using Joint Label Fusion. Patch-Based Techniques in Medical Imaging - 3rd International Workshop, Patch-MI 2017 Held in Conjunction with MICCAI 2017, Proceedings. Vol. 10530 LNCS Springer Verlag, 2017. pp. 138-145 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{673e2bd167b74ab180a0a4367941faf7,
title = "Multiple Sclerosis Lesion Segmentation Using Joint Label Fusion",
abstract = "This paper adapts the joint label fusion (JLF) multi-atlas image segmentation algorithm to the problem of multiple sclerosis (MS) lesion segmentation in multi-modal MRI. Conventionally, JLF requires a set of atlas images to be co-registered to the target image using deformable registration. However, given the variable spatial distribution of lesions in the brain, whole-brain deformable registration is unlikely to line up lesions between atlases and the target image. As a solution, we propose to first pre-segment the target image using an intensity regression based technique, yielding a set of “candidate” lesions. Each “candidate” lesion is then matched to a set of similar lesions in the atlas based on location and size; and deformable registration and JLF are applied at the level of the “candidate” lesion. The approach is evaluated on a dataset of 74 subjects with MS and shown to improve Dice similarity coefficient with reference manual segmentation by 12{\%} over intensity regression technique.",
author = "Mengjin Dong and Ipek Oguz and Nagesh Subbana and Peter Calabresi and Shinohara, {Russell T.} and Paul Yushkevich",
year = "2017",
doi = "10.1007/978-3-319-67434-6_16",
language = "English (US)",
isbn = "9783319674339",
volume = "10530 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "138--145",
booktitle = "Patch-Based Techniques in Medical Imaging - 3rd International Workshop, Patch-MI 2017 Held in Conjunction with MICCAI 2017, Proceedings",

}

TY - GEN

T1 - Multiple Sclerosis Lesion Segmentation Using Joint Label Fusion

AU - Dong, Mengjin

AU - Oguz, Ipek

AU - Subbana, Nagesh

AU - Calabresi, Peter

AU - Shinohara, Russell T.

AU - Yushkevich, Paul

PY - 2017

Y1 - 2017

N2 - This paper adapts the joint label fusion (JLF) multi-atlas image segmentation algorithm to the problem of multiple sclerosis (MS) lesion segmentation in multi-modal MRI. Conventionally, JLF requires a set of atlas images to be co-registered to the target image using deformable registration. However, given the variable spatial distribution of lesions in the brain, whole-brain deformable registration is unlikely to line up lesions between atlases and the target image. As a solution, we propose to first pre-segment the target image using an intensity regression based technique, yielding a set of “candidate” lesions. Each “candidate” lesion is then matched to a set of similar lesions in the atlas based on location and size; and deformable registration and JLF are applied at the level of the “candidate” lesion. The approach is evaluated on a dataset of 74 subjects with MS and shown to improve Dice similarity coefficient with reference manual segmentation by 12% over intensity regression technique.

AB - This paper adapts the joint label fusion (JLF) multi-atlas image segmentation algorithm to the problem of multiple sclerosis (MS) lesion segmentation in multi-modal MRI. Conventionally, JLF requires a set of atlas images to be co-registered to the target image using deformable registration. However, given the variable spatial distribution of lesions in the brain, whole-brain deformable registration is unlikely to line up lesions between atlases and the target image. As a solution, we propose to first pre-segment the target image using an intensity regression based technique, yielding a set of “candidate” lesions. Each “candidate” lesion is then matched to a set of similar lesions in the atlas based on location and size; and deformable registration and JLF are applied at the level of the “candidate” lesion. The approach is evaluated on a dataset of 74 subjects with MS and shown to improve Dice similarity coefficient with reference manual segmentation by 12% over intensity regression technique.

UR - http://www.scopus.com/inward/record.url?scp=85029410629&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029410629&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-67434-6_16

DO - 10.1007/978-3-319-67434-6_16

M3 - Conference contribution

C2 - 29707700

AN - SCOPUS:85029410629

SN - 9783319674339

VL - 10530 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 138

EP - 145

BT - Patch-Based Techniques in Medical Imaging - 3rd International Workshop, Patch-MI 2017 Held in Conjunction with MICCAI 2017, Proceedings

PB - Springer Verlag

ER -