Multimodal 3D Brain Tumor Segmentation with Adversarial Training and Conditional Random Field

Lan  Jiang; Yuchao Zheng; Miao Yu; Haiqing Zhang; Fatemah Aladwani; Alessandro Perelli

doi:10.1007/978-3-031-66955-2_5

Multimodal 3D Brain Tumor Segmentation with Adversarial Training and Conditional Random Field

Lan Jiang, Yuchao Zheng, Miao Yu, Haiqing Zhang, Fatemah Aladwani, Alessandro Perelli (Lead / Corresponding author)

Biomedical Engineering

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

Abstract

Accurate brain tumor segmentation remains a challenging task due to structural complexity and great individual differences of gliomas. Leveraging the pre-eminent detail resilience of CRF and spatial feature extraction capacity of V-net, we propose a multimodal 3D Volume Generative Adversarial Network (3D-vGAN) for precise segmentation. The model utilizes Pseudo-3D for V-net improvement, adds conditional random field after generator and use original image as supplemental guidance. Results, using the BraTS-2018 dataset, show that 3D-vGAN outperforms classical segmentation models, including U-net, Gan, FCN and 3D V-net, reaching specificity over 99.8%.

Original language	English
Title of host publication	Medical Image Understanding and Analysis
Subtitle of host publication	28th Annual Conference, MIUA 2024, Proceedings
Editors	Moi Hoon Yap, Connah Kendrick, Ardhendu Behera, Timothy Cootes, Reyer Zwiggelaar
Publisher	Springer
Pages	68-80
Number of pages	13
Volume	14859
ISBN (Electronic)	9783031669552
ISBN (Print)	9783031669545
DOIs	https://doi.org/10.1007/978-3-031-66955-2_5
Publication status	Published - 24 Jul 2024
Event	28th Conference on Medical Image Understanding and Analysis (MIUA) - Manchester Metropolitan University, Manchester, United Kingdom Duration: 24 Jul 2024 → 26 Jul 2024 https://miua2024.github.io/

Publication series

Name	Lecture Notes in Computer Science
Volume	14859 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	28th Conference on Medical Image Understanding and Analysis (MIUA)
Abbreviated title	MIUA
Country/Territory	United Kingdom
City	Manchester
Period	24/07/24 → 26/07/24
Internet address	https://miua2024.github.io/

Keywords

Multimodal Segmentation
Generative Adversarial Network (GAN)
Brain tumor
Generative Adversarial Network

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-031-66955-2_5

Cite this

Jiang, L., Zheng, Y., Yu, M., Zhang, H., Aladwani, F., & Perelli, A. (2024). Multimodal 3D Brain Tumor Segmentation with Adversarial Training and Conditional Random Field. In M. H. Yap, C. Kendrick, A. Behera, T. Cootes, & R. Zwiggelaar (Eds.), Medical Image Understanding and Analysis: 28th Annual Conference, MIUA 2024, Proceedings (Vol. 14859, pp. 68-80). (Lecture Notes in Computer Science; Vol. 14859 LNCS). Springer . https://doi.org/10.1007/978-3-031-66955-2_5

Jiang, Lan ; Zheng, Yuchao ; Yu, Miao et al. / Multimodal 3D Brain Tumor Segmentation with Adversarial Training and Conditional Random Field. Medical Image Understanding and Analysis: 28th Annual Conference, MIUA 2024, Proceedings. editor / Moi Hoon Yap ; Connah Kendrick ; Ardhendu Behera ; Timothy Cootes ; Reyer Zwiggelaar. Vol. 14859 Springer , 2024. pp. 68-80 (Lecture Notes in Computer Science).

@inproceedings{9c6a2f23abe94c50b2917c0e8d629c41,

title = "Multimodal 3D Brain Tumor Segmentation with Adversarial Training and Conditional Random Field",

abstract = "Accurate brain tumor segmentation remains a challenging task due to structural complexity and great individual differences of gliomas. Leveraging the pre-eminent detail resilience of CRF and spatial feature extraction capacity of V-net, we propose a multimodal 3D Volume Generative Adversarial Network (3D-vGAN) for precise segmentation. The model utilizes Pseudo-3D for V-net improvement, adds conditional random field after generator and use original image as supplemental guidance. Results, using the BraTS-2018 dataset, show that 3D-vGAN outperforms classical segmentation models, including U-net, Gan, FCN and 3D V-net, reaching specificity over 99.8%.",

keywords = "Multimodal Segmentation, Generative Adversarial Network (GAN), Brain tumor, Generative Adversarial Network",

author = "Lan Jiang and Yuchao Zheng and Miao Yu and Haiqing Zhang and Fatemah Aladwani and Alessandro Perelli",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.; 28th Conference on Medical Image Understanding and Analysis (MIUA), MIUA ; Conference date: 24-07-2024 Through 26-07-2024",

year = "2024",

month = jul,

day = "24",

doi = "10.1007/978-3-031-66955-2_5",

language = "English",

isbn = "9783031669545",

volume = "14859",

series = "Lecture Notes in Computer Science",

publisher = "Springer ",

pages = "68--80",

editor = "Yap, {Moi Hoon} and Connah Kendrick and Ardhendu Behera and Timothy Cootes and Reyer Zwiggelaar",

booktitle = "Medical Image Understanding and Analysis",

url = "https://miua2024.github.io/",

}

Jiang, L, Zheng, Y, Yu, M, Zhang, H, Aladwani, F & Perelli, A 2024, Multimodal 3D Brain Tumor Segmentation with Adversarial Training and Conditional Random Field. in MH Yap, C Kendrick, A Behera, T Cootes & R Zwiggelaar (eds), Medical Image Understanding and Analysis: 28th Annual Conference, MIUA 2024, Proceedings. vol. 14859, Lecture Notes in Computer Science, vol. 14859 LNCS, Springer , pp. 68-80, 28th Conference on Medical Image Understanding and Analysis (MIUA), Manchester, United Kingdom, 24/07/24. https://doi.org/10.1007/978-3-031-66955-2_5

Multimodal 3D Brain Tumor Segmentation with Adversarial Training and Conditional Random Field. / Jiang, Lan ; Zheng, Yuchao; Yu, Miao et al.
Medical Image Understanding and Analysis: 28th Annual Conference, MIUA 2024, Proceedings. ed. / Moi Hoon Yap; Connah Kendrick; Ardhendu Behera; Timothy Cootes; Reyer Zwiggelaar. Vol. 14859 Springer , 2024. p. 68-80 (Lecture Notes in Computer Science; Vol. 14859 LNCS).

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

TY - GEN

T1 - Multimodal 3D Brain Tumor Segmentation with Adversarial Training and Conditional Random Field

AU - Jiang, Lan

AU - Zheng, Yuchao

AU - Yu, Miao

AU - Zhang, Haiqing

AU - Aladwani, Fatemah

AU - Perelli, Alessandro

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

PY - 2024/7/24

Y1 - 2024/7/24

N2 - Accurate brain tumor segmentation remains a challenging task due to structural complexity and great individual differences of gliomas. Leveraging the pre-eminent detail resilience of CRF and spatial feature extraction capacity of V-net, we propose a multimodal 3D Volume Generative Adversarial Network (3D-vGAN) for precise segmentation. The model utilizes Pseudo-3D for V-net improvement, adds conditional random field after generator and use original image as supplemental guidance. Results, using the BraTS-2018 dataset, show that 3D-vGAN outperforms classical segmentation models, including U-net, Gan, FCN and 3D V-net, reaching specificity over 99.8%.

AB - Accurate brain tumor segmentation remains a challenging task due to structural complexity and great individual differences of gliomas. Leveraging the pre-eminent detail resilience of CRF and spatial feature extraction capacity of V-net, we propose a multimodal 3D Volume Generative Adversarial Network (3D-vGAN) for precise segmentation. The model utilizes Pseudo-3D for V-net improvement, adds conditional random field after generator and use original image as supplemental guidance. Results, using the BraTS-2018 dataset, show that 3D-vGAN outperforms classical segmentation models, including U-net, Gan, FCN and 3D V-net, reaching specificity over 99.8%.

KW - Multimodal Segmentation

KW - Generative Adversarial Network (GAN)

KW - Brain tumor

KW - Generative Adversarial Network

UR - http://www.scopus.com/inward/record.url?scp=85200690660&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-66955-2_5

DO - 10.1007/978-3-031-66955-2_5

M3 - Conference contribution

SN - 9783031669545

VL - 14859

T3 - Lecture Notes in Computer Science

SP - 68

EP - 80

BT - Medical Image Understanding and Analysis

A2 - Yap, Moi Hoon

A2 - Kendrick, Connah

A2 - Behera, Ardhendu

A2 - Cootes, Timothy

A2 - Zwiggelaar, Reyer

PB - Springer

T2 - 28th Conference on Medical Image Understanding and Analysis (MIUA)

Y2 - 24 July 2024 through 26 July 2024

ER -

Jiang L, Zheng Y, Yu M, Zhang H, Aladwani F , Perelli A. Multimodal 3D Brain Tumor Segmentation with Adversarial Training and Conditional Random Field. In Yap MH, Kendrick C, Behera A, Cootes T, Zwiggelaar R, editors, Medical Image Understanding and Analysis: 28th Annual Conference, MIUA 2024, Proceedings. Vol. 14859. Springer . 2024. p. 68-80. (Lecture Notes in Computer Science). doi: 10.1007/978-3-031-66955-2_5

Multimodal 3D Brain Tumor Segmentation with Adversarial Training and Conditional Random Field

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Multimodal 3D Brain Tumor Segmentation with Adversarial Training and Conditional Random Field

Cite this