Automatic Fetal Brain Extraction Using Multi-stage U-Net with Deep Supervision

Jingjiao Lou; Dengwang Li; Toan Duc Bui; Fenqiang Zhao; Liang Sun; Gang Li; Dinggang Shen

doi:10.1007/978-3-030-32692-0_68

Automatic Fetal Brain Extraction Using Multi-stage U-Net with Deep Supervision

Jingjiao Lou, Dengwang Li, Toan Duc Bui, Fenqiang Zhao, Liang Sun, Gang Li, Dinggang Shen

Department of Brain and Cognitive Engineering

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

18 Citations (Scopus)

Abstract

Fetal brain extraction is one of the most essential steps for prenatal brain MRI reconstruction and analysis. However, due to the fetal movement within the womb, it is a challenging task to extract fetal brains from sparsely-acquired imaging stacks typically with motion artifacts. To address this problem, we propose an automatic brain extraction method for fetal magnetic resonance imaging (MRI) using multi-stage 2D U-Net with deep supervision (DS U-net). Specifically, we initially employ a coarse segmentation derived from DS U-net to define a 3D bounding box for localizing the position of the brain. The DS U-net is trained with deep supervision loss to acquire more powerful discrimination capability. Then, another DS U-net focuses on the extracted region to produce finer segmentation. The final segmentation results are obtained by performing refined segmentation. We validate the proposed method on 80 stacks of training images and 43 testing stacks. The experimental results demonstrate the precision and robustness of our method with the average Dice coefficient of 91.69%, outperforming the existing methods.

Original language	English
Title of host publication	Machine Learning in Medical Imaging - 10th International Workshop, MLMI 2019, Held in Conjunction with MICCAI 2019, Proceedings
Editors	Heung-Il Suk, Mingxia Liu, Chunfeng Lian, Pingkun Yan
Publisher	Springer
Pages	592-600
Number of pages	9
ISBN (Print)	9783030326913
DOIs	https://doi.org/10.1007/978-3-030-32692-0_68
Publication status	Published - 2019
Event	10th International Workshop on Machine Learning in Medical Imaging, MLMI 2019 held in conjunction with the 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019 - Shenzhen, China Duration: 2019 Oct 13 → 2019 Oct 13

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11861 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	10th International Workshop on Machine Learning in Medical Imaging, MLMI 2019 held in conjunction with the 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019
Country/Territory	China
City	Shenzhen
Period	19/10/13 → 19/10/13

Bibliographical note

Publisher Copyright:
© 2019, Springer Nature Switzerland AG.

Keywords

Brain extraction
Convolutional neural network
Fetal MRI

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-030-32692-0_68

Cite this

Lou, J., Li, D., Bui, T. D., Zhao, F., Sun, L., Li, G., & Shen, D. (2019). Automatic Fetal Brain Extraction Using Multi-stage U-Net with Deep Supervision. In H.-I. Suk, M. Liu, C. Lian, & P. Yan (Eds.), Machine Learning in Medical Imaging - 10th International Workshop, MLMI 2019, Held in Conjunction with MICCAI 2019, Proceedings (pp. 592-600). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11861 LNCS). Springer. https://doi.org/10.1007/978-3-030-32692-0_68

Automatic Fetal Brain Extraction Using Multi-stage U-Net with Deep Supervision. / Lou, Jingjiao; Li, Dengwang; Bui, Toan Duc et al.
Machine Learning in Medical Imaging - 10th International Workshop, MLMI 2019, Held in Conjunction with MICCAI 2019, Proceedings. ed. / Heung-Il Suk; Mingxia Liu; Chunfeng Lian; Pingkun Yan. Springer, 2019. p. 592-600 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11861 LNCS).

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

Lou, J, Li, D, Bui, TD, Zhao, F, Sun, L, Li, G & Shen, D 2019, Automatic Fetal Brain Extraction Using Multi-stage U-Net with Deep Supervision. in H-I Suk, M Liu, C Lian & P Yan (eds), Machine Learning in Medical Imaging - 10th International Workshop, MLMI 2019, Held in Conjunction with MICCAI 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11861 LNCS, Springer, pp. 592-600, 10th International Workshop on Machine Learning in Medical Imaging, MLMI 2019 held in conjunction with the 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019, Shenzhen, China, 19/10/13. https://doi.org/10.1007/978-3-030-32692-0_68

Lou J, Li D, Bui TD, Zhao F, Sun L, Li G et al. Automatic Fetal Brain Extraction Using Multi-stage U-Net with Deep Supervision. In Suk HI, Liu M, Lian C, Yan P, editors, Machine Learning in Medical Imaging - 10th International Workshop, MLMI 2019, Held in Conjunction with MICCAI 2019, Proceedings. Springer. 2019. p. 592-600. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-32692-0_68

Lou, Jingjiao ; Li, Dengwang ; Bui, Toan Duc et al. / Automatic Fetal Brain Extraction Using Multi-stage U-Net with Deep Supervision. Machine Learning in Medical Imaging - 10th International Workshop, MLMI 2019, Held in Conjunction with MICCAI 2019, Proceedings. editor / Heung-Il Suk ; Mingxia Liu ; Chunfeng Lian ; Pingkun Yan. Springer, 2019. pp. 592-600 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{2eac13503534445dadfe02c345d609a6,

title = "Automatic Fetal Brain Extraction Using Multi-stage U-Net with Deep Supervision",

abstract = "Fetal brain extraction is one of the most essential steps for prenatal brain MRI reconstruction and analysis. However, due to the fetal movement within the womb, it is a challenging task to extract fetal brains from sparsely-acquired imaging stacks typically with motion artifacts. To address this problem, we propose an automatic brain extraction method for fetal magnetic resonance imaging (MRI) using multi-stage 2D U-Net with deep supervision (DS U-net). Specifically, we initially employ a coarse segmentation derived from DS U-net to define a 3D bounding box for localizing the position of the brain. The DS U-net is trained with deep supervision loss to acquire more powerful discrimination capability. Then, another DS U-net focuses on the extracted region to produce finer segmentation. The final segmentation results are obtained by performing refined segmentation. We validate the proposed method on 80 stacks of training images and 43 testing stacks. The experimental results demonstrate the precision and robustness of our method with the average Dice coefficient of 91.69%, outperforming the existing methods.",

keywords = "Brain extraction, Convolutional neural network, Fetal MRI",

author = "Jingjiao Lou and Dengwang Li and Bui, {Toan Duc} and Fenqiang Zhao and Liang Sun and Gang Li and Dinggang Shen",

note = "Publisher Copyright: {\textcopyright} 2019, Springer Nature Switzerland AG.; 10th International Workshop on Machine Learning in Medical Imaging, MLMI 2019 held in conjunction with the 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019 ; Conference date: 13-10-2019 Through 13-10-2019",

year = "2019",

doi = "10.1007/978-3-030-32692-0_68",

language = "English",

isbn = "9783030326913",

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

publisher = "Springer",

pages = "592--600",

editor = "Heung-Il Suk and Mingxia Liu and Chunfeng Lian and Pingkun Yan",

booktitle = "Machine Learning in Medical Imaging - 10th International Workshop, MLMI 2019, Held in Conjunction with MICCAI 2019, Proceedings",

}

TY - GEN

T1 - Automatic Fetal Brain Extraction Using Multi-stage U-Net with Deep Supervision

AU - Lou, Jingjiao

AU - Li, Dengwang

AU - Bui, Toan Duc

AU - Zhao, Fenqiang

AU - Sun, Liang

AU - Li, Gang

AU - Shen, Dinggang

PY - 2019

Y1 - 2019

N2 - Fetal brain extraction is one of the most essential steps for prenatal brain MRI reconstruction and analysis. However, due to the fetal movement within the womb, it is a challenging task to extract fetal brains from sparsely-acquired imaging stacks typically with motion artifacts. To address this problem, we propose an automatic brain extraction method for fetal magnetic resonance imaging (MRI) using multi-stage 2D U-Net with deep supervision (DS U-net). Specifically, we initially employ a coarse segmentation derived from DS U-net to define a 3D bounding box for localizing the position of the brain. The DS U-net is trained with deep supervision loss to acquire more powerful discrimination capability. Then, another DS U-net focuses on the extracted region to produce finer segmentation. The final segmentation results are obtained by performing refined segmentation. We validate the proposed method on 80 stacks of training images and 43 testing stacks. The experimental results demonstrate the precision and robustness of our method with the average Dice coefficient of 91.69%, outperforming the existing methods.

AB - Fetal brain extraction is one of the most essential steps for prenatal brain MRI reconstruction and analysis. However, due to the fetal movement within the womb, it is a challenging task to extract fetal brains from sparsely-acquired imaging stacks typically with motion artifacts. To address this problem, we propose an automatic brain extraction method for fetal magnetic resonance imaging (MRI) using multi-stage 2D U-Net with deep supervision (DS U-net). Specifically, we initially employ a coarse segmentation derived from DS U-net to define a 3D bounding box for localizing the position of the brain. The DS U-net is trained with deep supervision loss to acquire more powerful discrimination capability. Then, another DS U-net focuses on the extracted region to produce finer segmentation. The final segmentation results are obtained by performing refined segmentation. We validate the proposed method on 80 stacks of training images and 43 testing stacks. The experimental results demonstrate the precision and robustness of our method with the average Dice coefficient of 91.69%, outperforming the existing methods.

KW - Brain extraction

KW - Convolutional neural network

KW - Fetal MRI

UR - http://www.scopus.com/inward/record.url?scp=85075686677&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-32692-0_68

DO - 10.1007/978-3-030-32692-0_68

M3 - Conference contribution

AN - SCOPUS:85075686677

SN - 9783030326913

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

SP - 592

EP - 600

BT - Machine Learning in Medical Imaging - 10th International Workshop, MLMI 2019, Held in Conjunction with MICCAI 2019, Proceedings

A2 - Suk, Heung-Il

A2 - Liu, Mingxia

A2 - Lian, Chunfeng

A2 - Yan, Pingkun

PB - Springer

T2 - 10th International Workshop on Machine Learning in Medical Imaging, MLMI 2019 held in conjunction with the 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019

Y2 - 13 October 2019 through 13 October 2019

ER -

Automatic Fetal Brain Extraction Using Multi-stage U-Net with Deep Supervision

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this