Multi-task Learning for Left Atrial Segmentation on GE-MRI

Chen Chen; Wenjia Bai; Daniel Rueckert

doi:10.1007/978-3-030-12029-0_32

Multi-task Learning for Left Atrial Segmentation on GE-MRI

Chen Chen, Wenjia Bai, Daniel Rueckert

Imperial College London

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

45 Scopus citations

Abstract

Segmentation of the left atrium (LA) is crucial for assessing its anatomy in both pre-operative atrial fibrillation (AF) ablation planning and post-operative follow-up studies. In this paper, we present a fully automated framework for left atrial segmentation in gadolinium-enhanced magnetic resonance images (GE-MRI) based on deep learning. We propose a fully convolutional neural network and explore the benefits of multi-task learning for performing both atrial segmentation and pre/post ablation classification. Our results show that, by sharing features between related tasks, the network can gain additional anatomical information and achieve more accurate atrial segmentation, leading to a mean Dice score of 0.901 on a test set of 20 3D MRI images. Code of our proposed algorithm is available at https://github.com/cherise215/atria_segmentation_2018/.

Original language	English
Title of host publication	Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers
Editors	Kristin McLeod, Alistair Young, Kawal Rhode, Shuo Li, Jichao Zhao, Tommaso Mansi, Maxime Sermesant, Mihaela Pop
Publisher	Springer Verlag
Pages	292-301
Number of pages	10
ISBN (Print)	9783030120283
DOIs	https://doi.org/10.1007/978-3-030-12029-0_32
State	Published - 2019
Externally published	Yes
Event	9th International Workshop on Statistical Atlases and Computational Models of the Heart: Atrial Segmentation and LV Quantification Challenges, STACOM 2018, held in conjunction with Medical Image Computing and Computer-Assisted Intervention, MICCAI 2018 - Granada, Spain Duration: 16 Sep 2018 → 16 Sep 2018

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11395 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	9th International Workshop on Statistical Atlases and Computational Models of the Heart: Atrial Segmentation and LV Quantification Challenges, STACOM 2018, held in conjunction with Medical Image Computing and Computer-Assisted Intervention, MICCAI 2018
Country/Territory	Spain
City	Granada
Period	16/09/18 → 16/09/18

Keywords

Atrial segmentation
Fully convolution neural network
Multi-task learning

Access to Document

10.1007/978-3-030-12029-0_32

Cite this

Chen, C., Bai, W., & Rueckert, D. (2019). Multi-task Learning for Left Atrial Segmentation on GE-MRI. In K. McLeod, A. Young, K. Rhode, S. Li, J. Zhao, T. Mansi, M. Sermesant, & M. Pop (Eds.), Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers (pp. 292-301). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11395 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-12029-0_32

Chen, Chen ; Bai, Wenjia ; Rueckert, Daniel. / Multi-task Learning for Left Atrial Segmentation on GE-MRI. Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers. editor / Kristin McLeod ; Alistair Young ; Kawal Rhode ; Shuo Li ; Jichao Zhao ; Tommaso Mansi ; Maxime Sermesant ; Mihaela Pop. Springer Verlag, 2019. pp. 292-301 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{d54a6a0d9046448c8c228363d51f4ccf,

title = "Multi-task Learning for Left Atrial Segmentation on GE-MRI",

abstract = "Segmentation of the left atrium (LA) is crucial for assessing its anatomy in both pre-operative atrial fibrillation (AF) ablation planning and post-operative follow-up studies. In this paper, we present a fully automated framework for left atrial segmentation in gadolinium-enhanced magnetic resonance images (GE-MRI) based on deep learning. We propose a fully convolutional neural network and explore the benefits of multi-task learning for performing both atrial segmentation and pre/post ablation classification. Our results show that, by sharing features between related tasks, the network can gain additional anatomical information and achieve more accurate atrial segmentation, leading to a mean Dice score of 0.901 on a test set of 20 3D MRI images. Code of our proposed algorithm is available at https://github.com/cherise215/atria_segmentation_2018/.",

keywords = "Atrial segmentation, Fully convolution neural network, Multi-task learning",

author = "Chen Chen and Wenjia Bai and Daniel Rueckert",

note = "Publisher Copyright: {\textcopyright} 2019, Springer Nature Switzerland AG.; 9th International Workshop on Statistical Atlases and Computational Models of the Heart: Atrial Segmentation and LV Quantification Challenges, STACOM 2018, held in conjunction with Medical Image Computing and Computer-Assisted Intervention, MICCAI 2018 ; Conference date: 16-09-2018 Through 16-09-2018",

year = "2019",

doi = "10.1007/978-3-030-12029-0_32",

language = "English",

isbn = "9783030120283",

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

publisher = "Springer Verlag",

pages = "292--301",

editor = "Kristin McLeod and Alistair Young and Kawal Rhode and Shuo Li and Jichao Zhao and Tommaso Mansi and Maxime Sermesant and Mihaela Pop",

booktitle = "Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers",

}

Chen, C, Bai, W & Rueckert, D 2019, Multi-task Learning for Left Atrial Segmentation on GE-MRI. in K McLeod, A Young, K Rhode, S Li, J Zhao, T Mansi, M Sermesant & M Pop (eds), Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 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. 11395 LNCS, Springer Verlag, pp. 292-301, 9th International Workshop on Statistical Atlases and Computational Models of the Heart: Atrial Segmentation and LV Quantification Challenges, STACOM 2018, held in conjunction with Medical Image Computing and Computer-Assisted Intervention, MICCAI 2018, Granada, Spain, 16/09/18. https://doi.org/10.1007/978-3-030-12029-0_32

Multi-task Learning for Left Atrial Segmentation on GE-MRI. / Chen, Chen; Bai, Wenjia; Rueckert, Daniel.
Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers. ed. / Kristin McLeod; Alistair Young; Kawal Rhode; Shuo Li; Jichao Zhao; Tommaso Mansi; Maxime Sermesant; Mihaela Pop. Springer Verlag, 2019. p. 292-301 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11395 LNCS).

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

TY - GEN

T1 - Multi-task Learning for Left Atrial Segmentation on GE-MRI

AU - Chen, Chen

AU - Bai, Wenjia

AU - Rueckert, Daniel

PY - 2019

Y1 - 2019

N2 - Segmentation of the left atrium (LA) is crucial for assessing its anatomy in both pre-operative atrial fibrillation (AF) ablation planning and post-operative follow-up studies. In this paper, we present a fully automated framework for left atrial segmentation in gadolinium-enhanced magnetic resonance images (GE-MRI) based on deep learning. We propose a fully convolutional neural network and explore the benefits of multi-task learning for performing both atrial segmentation and pre/post ablation classification. Our results show that, by sharing features between related tasks, the network can gain additional anatomical information and achieve more accurate atrial segmentation, leading to a mean Dice score of 0.901 on a test set of 20 3D MRI images. Code of our proposed algorithm is available at https://github.com/cherise215/atria_segmentation_2018/.

AB - Segmentation of the left atrium (LA) is crucial for assessing its anatomy in both pre-operative atrial fibrillation (AF) ablation planning and post-operative follow-up studies. In this paper, we present a fully automated framework for left atrial segmentation in gadolinium-enhanced magnetic resonance images (GE-MRI) based on deep learning. We propose a fully convolutional neural network and explore the benefits of multi-task learning for performing both atrial segmentation and pre/post ablation classification. Our results show that, by sharing features between related tasks, the network can gain additional anatomical information and achieve more accurate atrial segmentation, leading to a mean Dice score of 0.901 on a test set of 20 3D MRI images. Code of our proposed algorithm is available at https://github.com/cherise215/atria_segmentation_2018/.

KW - Atrial segmentation

KW - Fully convolution neural network

KW - Multi-task learning

UR - http://www.scopus.com/inward/record.url?scp=85064053331&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-12029-0_32

DO - 10.1007/978-3-030-12029-0_32

M3 - Conference contribution

AN - SCOPUS:85064053331

SN - 9783030120283

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

SP - 292

EP - 301

BT - Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers

A2 - McLeod, Kristin

A2 - Young, Alistair

A2 - Rhode, Kawal

A2 - Li, Shuo

A2 - Zhao, Jichao

A2 - Mansi, Tommaso

A2 - Sermesant, Maxime

A2 - Pop, Mihaela

PB - Springer Verlag

T2 - 9th International Workshop on Statistical Atlases and Computational Models of the Heart: Atrial Segmentation and LV Quantification Challenges, STACOM 2018, held in conjunction with Medical Image Computing and Computer-Assisted Intervention, MICCAI 2018

Y2 - 16 September 2018 through 16 September 2018

ER -

Chen C, Bai W, Rueckert D. Multi-task Learning for Left Atrial Segmentation on GE-MRI. In McLeod K, Young A, Rhode K, Li S, Zhao J, Mansi T, Sermesant M, Pop M, editors, Statistical Atlases and Computational Models of the Heart. Atrial Segmentation and LV Quantification Challenges - 9th International Workshop, STACOM 2018, Held in Conjunction with MICCAI 2018, Revised Selected Papers. Springer Verlag. 2019. p. 292-301. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-12029-0_32

Multi-task Learning for Left Atrial Segmentation on GE-MRI

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this