TwoPath U-Net for automatic brain tumor segmentation from multimodal MRI data

Keerati Kaewrak; John Soraghan; Gaetano Di Caterina; Derek Grose

doi:10.1007/978-3-030-72087-2_26

TwoPath U-Net for automatic brain tumor segmentation from multimodal MRI data

Keerati Kaewrak, John Soraghan, Gaetano Di Caterina, Derek Grose

Electronic And Electrical Engineering

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

2 Citations (Scopus)

49 Downloads (Pure)

Abstract

A novel encoder-decoder deep learning network called TwoPath U-Net for multi-class automatic brain tumor segmentation task is presented. The network uses cascaded local and global feature extraction paths in the down-sampling path of the network which allows the network to learn different aspects of both the low-level feature and high-level features. The proposed network architecture using a full image and patches input technique was used on the BraTS2020 training dataset. We tested the network performance using the BraTS2019 validation dataset and obtained the mean dice score of 0.76, 0.64, and 0.58 and the Hausdorff distance 95% of 25.05, 32.83, and 37.57 for the whole tumor, tumor core and enhancing tumor regions.

Original language	English
Title of host publication	Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries
Subtitle of host publication	6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020, Lima, Peru, October 4, 2020, Revised Selected Papers, Part II
Editors	Alessandro Crimi, Spyridon Bakas
Place of Publication	Cham
Publisher	Springer International Publishing AG
Pages	300–309
Number of pages	10
Volume	12659
ISBN (Electronic)	9783030720872
ISBN (Print)	9783030720865
DOIs	https://doi.org/10.1007/978-3-030-72087-2_26
Publication status	Published - 26 Mar 2021
Event	6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020 - Online Duration: 4 Oct 2020 → 4 Oct 2020 http://www.brainlesion-workshop.org/

Conference

Conference	6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020
Abbreviated title	MICCAI BrainLes 2020
Period	4/10/20 → 4/10/20
Internet address	http://www.brainlesion-workshop.org/

Keywords

brain tumor
deep learning
segmentation
TwoPath U-Net
multimodal MRI

Access to Document

10.1007/978-3-030-72087-2_26

Kaewrak-etal-Springer-2021-TwoPath-U-Net-for-automatic-brain-tumor-segmentationAccepted author manuscript, 362 KB

Cite this

Kaewrak, K., Soraghan, J., Di Caterina, G., & Grose, D. (2021). TwoPath U-Net for automatic brain tumor segmentation from multimodal MRI data. In A. Crimi, & S. Bakas (Eds.), Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries: 6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020, Lima, Peru, October 4, 2020, Revised Selected Papers, Part II (Vol. 12659, pp. 300–309). Springer International Publishing AG. https://doi.org/10.1007/978-3-030-72087-2_26

Kaewrak, Keerati ; Soraghan, John ; Di Caterina, Gaetano et al. / TwoPath U-Net for automatic brain tumor segmentation from multimodal MRI data. Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries: 6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020, Lima, Peru, October 4, 2020, Revised Selected Papers, Part II. editor / Alessandro Crimi ; Spyridon Bakas. Vol. 12659 Cham : Springer International Publishing AG, 2021. pp. 300–309

@inproceedings{3b1731c64ab743508e8b7f4656c39097,

title = "TwoPath U-Net for automatic brain tumor segmentation from multimodal MRI data",

abstract = "A novel encoder-decoder deep learning network called TwoPath U-Net for multi-class automatic brain tumor segmentation task is presented. The network uses cascaded local and global feature extraction paths in the down-sampling path of the network which allows the network to learn different aspects of both the low-level feature and high-level features. The proposed network architecture using a full image and patches input technique was used on the BraTS2020 training dataset. We tested the network performance using the BraTS2019 validation dataset and obtained the mean dice score of 0.76, 0.64, and 0.58 and the Hausdorff distance 95% of 25.05, 32.83, and 37.57 for the whole tumor, tumor core and enhancing tumor regions.",

keywords = "brain tumor, deep learning, segmentation, TwoPath U-Net, multimodal MRI",

author = "Keerati Kaewrak and John Soraghan and {Di Caterina}, Gaetano and Derek Grose",

year = "2021",

month = mar,

day = "26",

doi = "10.1007/978-3-030-72087-2_26",

language = "English",

isbn = "9783030720865",

volume = "12659",

pages = "300–309",

editor = "Alessandro Crimi and Spyridon Bakas",

booktitle = "Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries",

publisher = "Springer International Publishing AG",

address = "Switzerland",

note = "6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020, MICCAI BrainLes 2020 ; Conference date: 04-10-2020 Through 04-10-2020",

url = "http://www.brainlesion-workshop.org/",

}

Kaewrak, K , Soraghan, J , Di Caterina, G & Grose, D 2021, TwoPath U-Net for automatic brain tumor segmentation from multimodal MRI data. in A Crimi & S Bakas (eds), Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries: 6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020, Lima, Peru, October 4, 2020, Revised Selected Papers, Part II. vol. 12659, Springer International Publishing AG, Cham, pp. 300–309, 6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020, 4/10/20. https://doi.org/10.1007/978-3-030-72087-2_26

TwoPath U-Net for automatic brain tumor segmentation from multimodal MRI data. / Kaewrak, Keerati ; Soraghan, John ; Di Caterina, Gaetano et al.
Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries: 6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020, Lima, Peru, October 4, 2020, Revised Selected Papers, Part II. ed. / Alessandro Crimi; Spyridon Bakas. Vol. 12659 Cham: Springer International Publishing AG, 2021. p. 300–309.

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

TY - GEN

T1 - TwoPath U-Net for automatic brain tumor segmentation from multimodal MRI data

AU - Kaewrak, Keerati

AU - Soraghan, John

AU - Di Caterina, Gaetano

AU - Grose, Derek

PY - 2021/3/26

Y1 - 2021/3/26

N2 - A novel encoder-decoder deep learning network called TwoPath U-Net for multi-class automatic brain tumor segmentation task is presented. The network uses cascaded local and global feature extraction paths in the down-sampling path of the network which allows the network to learn different aspects of both the low-level feature and high-level features. The proposed network architecture using a full image and patches input technique was used on the BraTS2020 training dataset. We tested the network performance using the BraTS2019 validation dataset and obtained the mean dice score of 0.76, 0.64, and 0.58 and the Hausdorff distance 95% of 25.05, 32.83, and 37.57 for the whole tumor, tumor core and enhancing tumor regions.

AB - A novel encoder-decoder deep learning network called TwoPath U-Net for multi-class automatic brain tumor segmentation task is presented. The network uses cascaded local and global feature extraction paths in the down-sampling path of the network which allows the network to learn different aspects of both the low-level feature and high-level features. The proposed network architecture using a full image and patches input technique was used on the BraTS2020 training dataset. We tested the network performance using the BraTS2019 validation dataset and obtained the mean dice score of 0.76, 0.64, and 0.58 and the Hausdorff distance 95% of 25.05, 32.83, and 37.57 for the whole tumor, tumor core and enhancing tumor regions.

KW - brain tumor

KW - deep learning

KW - segmentation

KW - TwoPath U-Net

KW - multimodal MRI

UR - http://www.brainlesion-workshop.org/

U2 - 10.1007/978-3-030-72087-2_26

DO - 10.1007/978-3-030-72087-2_26

M3 - Conference contribution book

SN - 9783030720865

VL - 12659

SP - 300

EP - 309

BT - Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries

A2 - Crimi, Alessandro

A2 - Bakas, Spyridon

PB - Springer International Publishing AG

CY - Cham

T2 - 6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020

Y2 - 4 October 2020 through 4 October 2020

ER -

Kaewrak K , Soraghan J , Di Caterina G, Grose D. TwoPath U-Net for automatic brain tumor segmentation from multimodal MRI data. In Crimi A, Bakas S, editors, Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries: 6th International Workshop, BrainLes 2020, Held in Conjunction with MICCAI 2020, Lima, Peru, October 4, 2020, Revised Selected Papers, Part II. Vol. 12659. Cham: Springer International Publishing AG. 2021. p. 300–309 doi: 10.1007/978-3-030-72087-2_26

TwoPath U-Net for automatic brain tumor segmentation from multimodal MRI data

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this