Segmenting brain tumors from MRI using cascaded multi-modal U-Nets

Michal Marcinkiewicz; Jakub Nalepa; Pablo Ribalta Lorenzo; Wojciech Dudzik; Grzegorz Mrukwa

doi:10.1007/978-3-030-11726-9_2

Segmenting brain tumors from MRI using cascaded multi-modal U-Nets

Michal Marcinkiewicz
, Jakub Nalepa
, Pablo Ribalta Lorenzo
, Wojciech Dudzik
, Grzegorz Mrukwa

Department of Algorithmics and Software (RAU/RAU5)

Future Processing
Silesian University of Technology

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

26 Citations (Scopus)

Abstract

Gliomas are the most common primary brain tumors, and their accurate manual delineation is a time- consuming and very user-dependent process. Therefore, developing automated techniques for reproducible detection and segmentation of brain tumors from magnetic resonance imaging is a vital research topic. In this paper, we present a deep learning-powered approach for brain tumor segmentation which exploits multiple magnetic-resonance modalities and processes them in two cascaded stages. In both stages, we use multi-modal fully-convolutional neural nets inspired by U-Nets. The first stage detects regions of interests, whereas the second stage performs the multi-class classification. Our experimental study, performed over the newest release of the BraTS dataset (BraTS 2018) showed that our method delivers accurate brain-tumor delineation and offers very fast processing—the total time required to segment one study using our approach amounts to around 18 s.

Original language	English
Title of host publication	Brainlesion
Subtitle of host publication	Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 4th International Workshop, BrainLes 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers
Editors	Alessandro Crimi, Mauricio Reyes, Hugo Kuijf, Theo van Walsum, Spyridon Bakas, Farahani Keyvan
Publisher	Springer Verlag
Pages	13-24
Number of pages	12
ISBN (Print)	9783030117252
DOIs	https://doi.org/10.1007/978-3-030-11726-9_2
Publication status	Published - 2019
Event	4th International MICCAI Brainlesion Workshop, BrainLes 2018 held in conjunction with the Medical Image Computing for Computer Assisted Intervention Conference, MICCAI 2018 - Granada, Spain Duration: 16 Sept 2018 → 20 Sept 2018

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11384 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	4th International MICCAI Brainlesion Workshop, BrainLes 2018 held in conjunction with the Medical Image Computing for Computer Assisted Intervention Conference, MICCAI 2018
Country/Territory	Spain
City	Granada
Period	16/09/18 → 20/09/18

Keywords

Brain tumor
CNN
Deep learning
Segmentation

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-030-11726-9_2

Cite this

Marcinkiewicz, M., Nalepa, J., Lorenzo, P. R., Dudzik, W., & Mrukwa, G. (2019). Segmenting brain tumors from MRI using cascaded multi-modal U-Nets. In A. Crimi, M. Reyes, H. Kuijf, T. van Walsum, S. Bakas, & F. Keyvan (Eds.), Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 4th International Workshop, BrainLes 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers (pp. 13-24). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11384 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-11726-9_2

Marcinkiewicz, Michal ; Nalepa, Jakub ; Lorenzo, Pablo Ribalta et al. / Segmenting brain tumors from MRI using cascaded multi-modal U-Nets. Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 4th International Workshop, BrainLes 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers. editor / Alessandro Crimi ; Mauricio Reyes ; Hugo Kuijf ; Theo van Walsum ; Spyridon Bakas ; Farahani Keyvan. Springer Verlag, 2019. pp. 13-24 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{b55d4f5a07d64b77a70fdcd03b27a412,

title = "Segmenting brain tumors from MRI using cascaded multi-modal U-Nets",

abstract = "Gliomas are the most common primary brain tumors, and their accurate manual delineation is a time- consuming and very user-dependent process. Therefore, developing automated techniques for reproducible detection and segmentation of brain tumors from magnetic resonance imaging is a vital research topic. In this paper, we present a deep learning-powered approach for brain tumor segmentation which exploits multiple magnetic-resonance modalities and processes them in two cascaded stages. In both stages, we use multi-modal fully-convolutional neural nets inspired by U-Nets. The first stage detects regions of interests, whereas the second stage performs the multi-class classification. Our experimental study, performed over the newest release of the BraTS dataset (BraTS 2018) showed that our method delivers accurate brain-tumor delineation and offers very fast processing{\textemdash}the total time required to segment one study using our approach amounts to around 18 s.",

keywords = "Brain tumor, CNN, Deep learning, Segmentation",

author = "Michal Marcinkiewicz and Jakub Nalepa and Lorenzo, \{Pablo Ribalta\} and Wojciech Dudzik and Grzegorz Mrukwa",

note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2019.; 4th International MICCAI Brainlesion Workshop, BrainLes 2018 held in conjunction with the Medical Image Computing for Computer Assisted Intervention Conference, MICCAI 2018 ; Conference date: 16-09-2018 Through 20-09-2018",

year = "2019",

doi = "10.1007/978-3-030-11726-9\_2",

language = "English",

isbn = "9783030117252",

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

publisher = "Springer Verlag",

pages = "13--24",

editor = "Alessandro Crimi and Mauricio Reyes and Hugo Kuijf and \{van Walsum\}, Theo and Spyridon Bakas and Farahani Keyvan",

booktitle = "Brainlesion",

address = "Germany",

}

Marcinkiewicz, M, Nalepa, J, Lorenzo, PR, Dudzik, W & Mrukwa, G 2019, Segmenting brain tumors from MRI using cascaded multi-modal U-Nets. in A Crimi, M Reyes, H Kuijf, T van Walsum, S Bakas & F Keyvan (eds), Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 4th International Workshop, BrainLes 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11384 LNCS, Springer Verlag, pp. 13-24, 4th International MICCAI Brainlesion Workshop, BrainLes 2018 held in conjunction with the Medical Image Computing for Computer Assisted Intervention Conference, MICCAI 2018, Granada, Spain, 16/09/18. https://doi.org/10.1007/978-3-030-11726-9_2

Segmenting brain tumors from MRI using cascaded multi-modal U-Nets. / Marcinkiewicz, Michal; Nalepa, Jakub; Lorenzo, Pablo Ribalta et al.
Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 4th International Workshop, BrainLes 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers. ed. / Alessandro Crimi; Mauricio Reyes; Hugo Kuijf; Theo van Walsum; Spyridon Bakas; Farahani Keyvan. Springer Verlag, 2019. p. 13-24 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11384 LNCS).

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

TY - GEN

T1 - Segmenting brain tumors from MRI using cascaded multi-modal U-Nets

AU - Marcinkiewicz, Michal

AU - Nalepa, Jakub

AU - Lorenzo, Pablo Ribalta

AU - Dudzik, Wojciech

AU - Mrukwa, Grzegorz

N1 - Publisher Copyright: © Springer Nature Switzerland AG 2019.

PY - 2019

Y1 - 2019

N2 - Gliomas are the most common primary brain tumors, and their accurate manual delineation is a time- consuming and very user-dependent process. Therefore, developing automated techniques for reproducible detection and segmentation of brain tumors from magnetic resonance imaging is a vital research topic. In this paper, we present a deep learning-powered approach for brain tumor segmentation which exploits multiple magnetic-resonance modalities and processes them in two cascaded stages. In both stages, we use multi-modal fully-convolutional neural nets inspired by U-Nets. The first stage detects regions of interests, whereas the second stage performs the multi-class classification. Our experimental study, performed over the newest release of the BraTS dataset (BraTS 2018) showed that our method delivers accurate brain-tumor delineation and offers very fast processing—the total time required to segment one study using our approach amounts to around 18 s.

AB - Gliomas are the most common primary brain tumors, and their accurate manual delineation is a time- consuming and very user-dependent process. Therefore, developing automated techniques for reproducible detection and segmentation of brain tumors from magnetic resonance imaging is a vital research topic. In this paper, we present a deep learning-powered approach for brain tumor segmentation which exploits multiple magnetic-resonance modalities and processes them in two cascaded stages. In both stages, we use multi-modal fully-convolutional neural nets inspired by U-Nets. The first stage detects regions of interests, whereas the second stage performs the multi-class classification. Our experimental study, performed over the newest release of the BraTS dataset (BraTS 2018) showed that our method delivers accurate brain-tumor delineation and offers very fast processing—the total time required to segment one study using our approach amounts to around 18 s.

KW - Brain tumor

KW - CNN

KW - Deep learning

KW - Segmentation

UR - https://www.scopus.com/pages/publications/85063491524

U2 - 10.1007/978-3-030-11726-9_2

DO - 10.1007/978-3-030-11726-9_2

M3 - Conference contribution

AN - SCOPUS:85063491524

SN - 9783030117252

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

SP - 13

EP - 24

BT - Brainlesion

A2 - Crimi, Alessandro

A2 - Reyes, Mauricio

A2 - Kuijf, Hugo

A2 - van Walsum, Theo

A2 - Bakas, Spyridon

A2 - Keyvan, Farahani

PB - Springer Verlag

T2 - 4th International MICCAI Brainlesion Workshop, BrainLes 2018 held in conjunction with the Medical Image Computing for Computer Assisted Intervention Conference, MICCAI 2018

Y2 - 16 September 2018 through 20 September 2018

ER -

Marcinkiewicz M, Nalepa J, Lorenzo PR, Dudzik W, Mrukwa G. Segmenting brain tumors from MRI using cascaded multi-modal U-Nets. In Crimi A, Reyes M, Kuijf H, van Walsum T, Bakas S, Keyvan F, editors, Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 4th International Workshop, BrainLes 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers. Springer Verlag. 2019. p. 13-24. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-11726-9_2

Segmenting brain tumors from MRI using cascaded multi-modal U-Nets

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this