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

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

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 languageEnglish
Title of host publicationMachine Learning in Medical Imaging - 10th International Workshop, MLMI 2019, Held in Conjunction with MICCAI 2019, Proceedings
EditorsHeung-Il Suk, Mingxia Liu, Chunfeng Lian, Pingkun Yan
PublisherSpringer
Pages592-600
Number of pages9
ISBN (Print)9783030326913
DOIs
Publication statusPublished - 2019 Jan 1
Externally publishedYes
Event10th 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 132019 Oct 13

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11861 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference10th 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
CountryChina
CityShenzhen
Period19/10/1319/10/13

Fingerprint

Brain
Segmentation
Magnetic Resonance Imaging
Magnetic resonance
Imaging techniques
Dice
Discrimination
Imaging
Robustness
Testing
Motion
Experimental Results
Coefficient
Demonstrate

Keywords

  • Brain extraction
  • Convolutional neural network
  • Fetal MRI

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

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; Zhao, Fenqiang; Sun, Liang; Li, Gang; Shen, Dinggang.

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 proceedingConference 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 H-I, 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)). https://doi.org/10.1007/978-3-030-32692-0_68
Lou, Jingjiao ; Li, Dengwang ; Bui, Toan Duc ; Zhao, Fenqiang ; Sun, Liang ; Li, Gang ; Shen, Dinggang. / 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",
year = "2019",
month = "1",
day = "1",
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 - 2019/1/1

Y1 - 2019/1/1

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

UR - http://www.scopus.com/inward/citedby.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

ER -