Segmentation of gliomas in pre-operative and post-operative multimodal magnetic resonance imaging volumes based on a hybrid generative-discriminative framework

Ke Zeng; Spyridon Bakas; Aristeidis Sotiras; Hamed Akbari; Martin Rozycki; Saima Rathore; Sarthak Pati; Christos Davatzikos

doi:10.1007/978-3-319-55524-9_18

Segmentation of gliomas in pre-operative and post-operative multimodal magnetic resonance imaging volumes based on a hybrid generative-discriminative framework

Ke Zeng, Spyridon Bakas, Aristeidis Sotiras, Hamed Akbari, Martin Rozycki, Saima Rathore, Sarthak Pati, Christos Davatzikos

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

26 Scopus citations

Abstract

We present an approach for segmenting both low- and highgrade gliomas in multimodal magnetic resonance imaging volumes. The proposed framework is an extension of our previous work [6,7], with an additional component for segmenting post-operative scans. The proposed approach is based on a hybrid generative-discriminative model. Firstly, a generative model based on a joint segmentation-registration framework is used to segment the brain scans into cancerous and healthy tissues. Secondly, a gradient boosting classification scheme is used to refine tumor segmentation based on information from multiple patients. We evaluated our approach in 218 cases during the training phase of the BRAin Tumor Segmentation (BRATS) 2016 challenge and report promising results. During the testing phase, the proposed approach was ranked among the top performing methods, after being additionally evaluated in 191 unseen cases.

Original language	English
Title of host publication	Brainlesion
Subtitle of host publication	Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - Second International Workshop, BrainLes 2016, with the Challenges on BRATS, ISLES and mTOP 2016 Held in Conjunction with MICCAI 2016, Revised Selected Papers
Editors	Bjoern Menze, Mauricio Reyes, Alessandro Crimi, Oskar Maier, Stefan Winzeck, Heinz Handels
Publisher	Springer Verlag
Pages	184-194
Number of pages	11
ISBN (Print)	9783319555232
DOIs	https://doi.org/10.1007/978-3-319-55524-9_18
State	Published - 2016
Event	2nd International Workshop on Brain Lesion, BrainLes 2016, with the challenges on Brain Tumor Segmentation BRATS, Ischemic Stroke Lesion Image Segmentation ISLES, and the Mild Traumatic Brain Injury Outcome Prediction mTOP held in conjunction with the International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016 - Athens, Greece Duration: Oct 17 2016 → Oct 17 2016

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10154 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	2nd International Workshop on Brain Lesion, BrainLes 2016, with the challenges on Brain Tumor Segmentation BRATS, Ischemic Stroke Lesion Image Segmentation ISLES, and the Mild Traumatic Brain Injury Outcome Prediction mTOP held in conjunction with the International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016
Country/Territory	Greece
City	Athens
Period	10/17/16 → 10/17/16

Keywords

BRATS challenge
Brain tumor
Expectation maximization
Glioma
Gradient boosting
Multimodal MRI
Probabilistic model
Segmentation

Access to Document

10.1007/978-3-319-55524-9_18

Cite this

Zeng, K., Bakas, S., Sotiras, A., Akbari, H., Rozycki, M., Rathore, S., Pati, S., & Davatzikos, C. (2016). Segmentation of gliomas in pre-operative and post-operative multimodal magnetic resonance imaging volumes based on a hybrid generative-discriminative framework. In B. Menze, M. Reyes, A. Crimi, O. Maier, S. Winzeck, & H. Handels (Eds.), Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - Second International Workshop, BrainLes 2016, with the Challenges on BRATS, ISLES and mTOP 2016 Held in Conjunction with MICCAI 2016, Revised Selected Papers (pp. 184-194). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10154 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-55524-9_18

Zeng, Ke ; Bakas, Spyridon ; Sotiras, Aristeidis et al. / Segmentation of gliomas in pre-operative and post-operative multimodal magnetic resonance imaging volumes based on a hybrid generative-discriminative framework. Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - Second International Workshop, BrainLes 2016, with the Challenges on BRATS, ISLES and mTOP 2016 Held in Conjunction with MICCAI 2016, Revised Selected Papers. editor / Bjoern Menze ; Mauricio Reyes ; Alessandro Crimi ; Oskar Maier ; Stefan Winzeck ; Heinz Handels. Springer Verlag, 2016. pp. 184-194 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{a3558b55fbca48f9afe79284db5d1fe6,

title = "Segmentation of gliomas in pre-operative and post-operative multimodal magnetic resonance imaging volumes based on a hybrid generative-discriminative framework",

abstract = "We present an approach for segmenting both low- and highgrade gliomas in multimodal magnetic resonance imaging volumes. The proposed framework is an extension of our previous work [6,7], with an additional component for segmenting post-operative scans. The proposed approach is based on a hybrid generative-discriminative model. Firstly, a generative model based on a joint segmentation-registration framework is used to segment the brain scans into cancerous and healthy tissues. Secondly, a gradient boosting classification scheme is used to refine tumor segmentation based on information from multiple patients. We evaluated our approach in 218 cases during the training phase of the BRAin Tumor Segmentation (BRATS) 2016 challenge and report promising results. During the testing phase, the proposed approach was ranked among the top performing methods, after being additionally evaluated in 191 unseen cases.",

keywords = "BRATS challenge, Brain tumor, Expectation maximization, Glioma, Gradient boosting, Multimodal MRI, Probabilistic model, Segmentation",

author = "Ke Zeng and Spyridon Bakas and Aristeidis Sotiras and Hamed Akbari and Martin Rozycki and Saima Rathore and Sarthak Pati and Christos Davatzikos",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2016.; 2nd International Workshop on Brain Lesion, BrainLes 2016, with the challenges on Brain Tumor Segmentation BRATS, Ischemic Stroke Lesion Image Segmentation ISLES, and the Mild Traumatic Brain Injury Outcome Prediction mTOP held in conjunction with the International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016 ; Conference date: 17-10-2016 Through 17-10-2016",

year = "2016",

doi = "10.1007/978-3-319-55524-9_18",

language = "English",

isbn = "9783319555232",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "184--194",

editor = "Bjoern Menze and Mauricio Reyes and Alessandro Crimi and Oskar Maier and Stefan Winzeck and Heinz Handels",

booktitle = "Brainlesion",

}

Zeng, K, Bakas, S, Sotiras, A, Akbari, H, Rozycki, M, Rathore, S, Pati, S & Davatzikos, C 2016, Segmentation of gliomas in pre-operative and post-operative multimodal magnetic resonance imaging volumes based on a hybrid generative-discriminative framework. in B Menze, M Reyes, A Crimi, O Maier, S Winzeck & H Handels (eds), Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - Second International Workshop, BrainLes 2016, with the Challenges on BRATS, ISLES and mTOP 2016 Held in Conjunction with MICCAI 2016, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10154 LNCS, Springer Verlag, pp. 184-194, 2nd International Workshop on Brain Lesion, BrainLes 2016, with the challenges on Brain Tumor Segmentation BRATS, Ischemic Stroke Lesion Image Segmentation ISLES, and the Mild Traumatic Brain Injury Outcome Prediction mTOP held in conjunction with the International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016, Athens, Greece, 10/17/16. https://doi.org/10.1007/978-3-319-55524-9_18

Segmentation of gliomas in pre-operative and post-operative multimodal magnetic resonance imaging volumes based on a hybrid generative-discriminative framework. / Zeng, Ke; Bakas, Spyridon; Sotiras, Aristeidis et al.
Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - Second International Workshop, BrainLes 2016, with the Challenges on BRATS, ISLES and mTOP 2016 Held in Conjunction with MICCAI 2016, Revised Selected Papers. ed. / Bjoern Menze; Mauricio Reyes; Alessandro Crimi; Oskar Maier; Stefan Winzeck; Heinz Handels. Springer Verlag, 2016. p. 184-194 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10154 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Segmentation of gliomas in pre-operative and post-operative multimodal magnetic resonance imaging volumes based on a hybrid generative-discriminative framework

AU - Zeng, Ke

AU - Bakas, Spyridon

AU - Sotiras, Aristeidis

AU - Akbari, Hamed

AU - Rozycki, Martin

AU - Rathore, Saima

AU - Pati, Sarthak

AU - Davatzikos, Christos

N1 - Publisher Copyright: © Springer International Publishing AG 2016.

PY - 2016

Y1 - 2016

N2 - We present an approach for segmenting both low- and highgrade gliomas in multimodal magnetic resonance imaging volumes. The proposed framework is an extension of our previous work [6,7], with an additional component for segmenting post-operative scans. The proposed approach is based on a hybrid generative-discriminative model. Firstly, a generative model based on a joint segmentation-registration framework is used to segment the brain scans into cancerous and healthy tissues. Secondly, a gradient boosting classification scheme is used to refine tumor segmentation based on information from multiple patients. We evaluated our approach in 218 cases during the training phase of the BRAin Tumor Segmentation (BRATS) 2016 challenge and report promising results. During the testing phase, the proposed approach was ranked among the top performing methods, after being additionally evaluated in 191 unseen cases.

AB - We present an approach for segmenting both low- and highgrade gliomas in multimodal magnetic resonance imaging volumes. The proposed framework is an extension of our previous work [6,7], with an additional component for segmenting post-operative scans. The proposed approach is based on a hybrid generative-discriminative model. Firstly, a generative model based on a joint segmentation-registration framework is used to segment the brain scans into cancerous and healthy tissues. Secondly, a gradient boosting classification scheme is used to refine tumor segmentation based on information from multiple patients. We evaluated our approach in 218 cases during the training phase of the BRAin Tumor Segmentation (BRATS) 2016 challenge and report promising results. During the testing phase, the proposed approach was ranked among the top performing methods, after being additionally evaluated in 191 unseen cases.

KW - BRATS challenge

KW - Brain tumor

KW - Expectation maximization

KW - Glioma

KW - Gradient boosting

KW - Multimodal MRI

KW - Probabilistic model

KW - Segmentation

UR - http://www.scopus.com/inward/record.url?scp=85018696985&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-55524-9_18

DO - 10.1007/978-3-319-55524-9_18

M3 - Conference contribution

C2 - 28725878

AN - SCOPUS:85018696985

SN - 9783319555232

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

SP - 184

EP - 194

BT - Brainlesion

A2 - Menze, Bjoern

A2 - Reyes, Mauricio

A2 - Crimi, Alessandro

A2 - Maier, Oskar

A2 - Winzeck, Stefan

A2 - Handels, Heinz

PB - Springer Verlag

T2 - 2nd International Workshop on Brain Lesion, BrainLes 2016, with the challenges on Brain Tumor Segmentation BRATS, Ischemic Stroke Lesion Image Segmentation ISLES, and the Mild Traumatic Brain Injury Outcome Prediction mTOP held in conjunction with the International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016

Y2 - 17 October 2016 through 17 October 2016

ER -

Zeng K, Bakas S, Sotiras A, Akbari H, Rozycki M, Rathore S et al. Segmentation of gliomas in pre-operative and post-operative multimodal magnetic resonance imaging volumes based on a hybrid generative-discriminative framework. In Menze B, Reyes M, Crimi A, Maier O, Winzeck S, Handels H, editors, Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - Second International Workshop, BrainLes 2016, with the Challenges on BRATS, ISLES and mTOP 2016 Held in Conjunction with MICCAI 2016, Revised Selected Papers. Springer Verlag. 2016. p. 184-194. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-55524-9_18

Segmentation of gliomas in pre-operative and post-operative multimodal magnetic resonance imaging volumes based on a hybrid generative-discriminative framework

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this