Using a Conditional Generative Adversarial Network (cGAN) for Prostate Segmentation

Amélie Grall; Azam Hamidinekoo; Paul Malcolm; Reyer Zwiggelaar

doi:10.1007/978-3-030-39343-4_2

Using a Conditional Generative Adversarial Network (cGAN) for Prostate Segmentation

Amélie Grall, Azam Hamidinekoo, Paul Malcolm, Reyer Zwiggelaar

Department of Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference Proceeding (Non-Journal item)

5 Citations (Scopus)

Abstract

Prostate cancer is the second most commonly diagnosed cancer among men and currently multi-parametric MRI is a promising imaging technique used for clinical workup of prostate cancer. Accurate detection and localisation of the prostate tissue boundary on various MRI scans can be helpful for obtaining a region of interest for Computer Aided Diagnosis systems. In this paper, we present a fully automated detection and segmentation pipeline using a conditional Generative Adversarial Network (cGAN). We investigated the robustness of the cGAN model against adding Gaussian noise or removing noise from the training data. Based on the detection and segmentation metrics, de-noising did not show a significant improvement. However, by including noisy images in the training data, the detection and segmentation performance was improved in each 3D modality, which resulted in comparable to state-of-the-art results.

Original language	English
Title of host publication	Medical Image Understanding and Analysis
Subtitle of host publication	23rd Conference, MIUA 2019, Liverpool, UK, July 24–26, 2019, Proceedings
Editors	Yalin Zheng, Bryan M. Williams, Ke Chen
Publisher	Springer Nature
Pages	15-25
Number of pages	11
ISBN (Electronic)	978-3-030-39343-4
ISBN (Print)	978-3-030-39342-7
DOIs	https://doi.org/10.1007/978-3-030-39343-4_2
Publication status	Published - 24 Jan 2020
Event	Proceedings 23rd Conference on Medical Image Understanding and Analysis - University of Liverpool, Liverpool, United Kingdom of Great Britain and Northern Ireland Duration: 24 Jul 2019 → 26 Jul 2019

Publication series

Name	Communications in Computer and Information Science
Publisher	Springer Nature
Volume	1065
ISSN (Print)	1865-0929
ISSN (Electronic)	1865-0937

Conference

Conference	Proceedings 23rd Conference on Medical Image Understanding and Analysis
Abbreviated title	MIUA 2019
Country/Territory	United Kingdom of Great Britain and Northern Ireland
City	Liverpool
Period	24 Jul 2019 → 26 Jul 2019

Keywords

Prostate MRI
Computer Aided Diagnosis
Segmentation
Detection
Generative Adversarial Network

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1007/978-3-030-39343-4_2

Permanent link

Persistent link

Cite this

Grall, A., Hamidinekoo, A., Malcolm, P., & Zwiggelaar, R. (2020). Using a Conditional Generative Adversarial Network (cGAN) for Prostate Segmentation. In Y. Zheng, B. M. Williams, & K. Chen (Eds.), Medical Image Understanding and Analysis: 23rd Conference, MIUA 2019, Liverpool, UK, July 24–26, 2019, Proceedings (pp. 15-25). (Communications in Computer and Information Science; Vol. 1065). Springer Nature. https://doi.org/10.1007/978-3-030-39343-4_2

Grall, Amélie ; Hamidinekoo, Azam ; Malcolm, Paul et al. / Using a Conditional Generative Adversarial Network (cGAN) for Prostate Segmentation. Medical Image Understanding and Analysis: 23rd Conference, MIUA 2019, Liverpool, UK, July 24–26, 2019, Proceedings. editor / Yalin Zheng ; Bryan M. Williams ; Ke Chen. Springer Nature, 2020. pp. 15-25 (Communications in Computer and Information Science).

@inproceedings{7bdba6ed3ff34eb49aceccdaae870057,

title = "Using a Conditional Generative Adversarial Network (cGAN) for Prostate Segmentation",

abstract = "Prostate cancer is the second most commonly diagnosed cancer among men and currently multi-parametric MRI is a promising imaging technique used for clinical workup of prostate cancer. Accurate detection and localisation of the prostate tissue boundary on various MRI scans can be helpful for obtaining a region of interest for Computer Aided Diagnosis systems. In this paper, we present a fully automated detection and segmentation pipeline using a conditional Generative Adversarial Network (cGAN). We investigated the robustness of the cGAN model against adding Gaussian noise or removing noise from the training data. Based on the detection and segmentation metrics, de-noising did not show a significant improvement. However, by including noisy images in the training data, the detection and segmentation performance was improved in each 3D modality, which resulted in comparable to state-of-the-art results.",

keywords = "Prostate MRI, Computer Aided Diagnosis, Segmentation, Detection, Generative Adversarial Network",

author = "Am{\'e}lie Grall and Azam Hamidinekoo and Paul Malcolm and Reyer Zwiggelaar",

note = "Publisher Copyright: {\textcopyright} 2020, Springer Nature Switzerland AG.; Proceedings 23rd Conference on Medical Image Understanding and Analysis, MIUA 2019 ; Conference date: 24-07-2019 Through 26-07-2019",

year = "2020",

month = jan,

day = "24",

doi = "10.1007/978-3-030-39343-4_2",

language = "English",

isbn = "978-3-030-39342-7",

series = "Communications in Computer and Information Science",

publisher = "Springer Nature",

pages = "15--25",

editor = "Yalin Zheng and Williams, {Bryan M.} and Ke Chen",

booktitle = "Medical Image Understanding and Analysis",

address = "Switzerland",

}

Grall, A, Hamidinekoo, A, Malcolm, P & Zwiggelaar, R 2020, Using a Conditional Generative Adversarial Network (cGAN) for Prostate Segmentation. in Y Zheng, BM Williams & K Chen (eds), Medical Image Understanding and Analysis: 23rd Conference, MIUA 2019, Liverpool, UK, July 24–26, 2019, Proceedings. Communications in Computer and Information Science, vol. 1065, Springer Nature, pp. 15-25, Proceedings 23rd Conference on Medical Image Understanding and Analysis, Liverpool, United Kingdom of Great Britain and Northern Ireland, 24 Jul 2019. https://doi.org/10.1007/978-3-030-39343-4_2

Using a Conditional Generative Adversarial Network (cGAN) for Prostate Segmentation. / Grall, Amélie; Hamidinekoo, Azam; Malcolm, Paul et al.
Medical Image Understanding and Analysis: 23rd Conference, MIUA 2019, Liverpool, UK, July 24–26, 2019, Proceedings. ed. / Yalin Zheng; Bryan M. Williams; Ke Chen. Springer Nature, 2020. p. 15-25 (Communications in Computer and Information Science; Vol. 1065).

Research output: Chapter in Book/Report/Conference proceeding › Conference Proceeding (Non-Journal item)

TY - GEN

T1 - Using a Conditional Generative Adversarial Network (cGAN) for Prostate Segmentation

AU - Grall, Amélie

AU - Hamidinekoo, Azam

AU - Malcolm, Paul

AU - Zwiggelaar, Reyer

PY - 2020/1/24

Y1 - 2020/1/24

N2 - Prostate cancer is the second most commonly diagnosed cancer among men and currently multi-parametric MRI is a promising imaging technique used for clinical workup of prostate cancer. Accurate detection and localisation of the prostate tissue boundary on various MRI scans can be helpful for obtaining a region of interest for Computer Aided Diagnosis systems. In this paper, we present a fully automated detection and segmentation pipeline using a conditional Generative Adversarial Network (cGAN). We investigated the robustness of the cGAN model against adding Gaussian noise or removing noise from the training data. Based on the detection and segmentation metrics, de-noising did not show a significant improvement. However, by including noisy images in the training data, the detection and segmentation performance was improved in each 3D modality, which resulted in comparable to state-of-the-art results.

AB - Prostate cancer is the second most commonly diagnosed cancer among men and currently multi-parametric MRI is a promising imaging technique used for clinical workup of prostate cancer. Accurate detection and localisation of the prostate tissue boundary on various MRI scans can be helpful for obtaining a region of interest for Computer Aided Diagnosis systems. In this paper, we present a fully automated detection and segmentation pipeline using a conditional Generative Adversarial Network (cGAN). We investigated the robustness of the cGAN model against adding Gaussian noise or removing noise from the training data. Based on the detection and segmentation metrics, de-noising did not show a significant improvement. However, by including noisy images in the training data, the detection and segmentation performance was improved in each 3D modality, which resulted in comparable to state-of-the-art results.

KW - Prostate MRI

KW - Computer Aided Diagnosis

KW - Segmentation

KW - Detection

KW - Generative Adversarial Network

UR - http://www.scopus.com/inward/record.url?scp=85079091149&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-39343-4_2

DO - 10.1007/978-3-030-39343-4_2

M3 - Conference Proceeding (Non-Journal item)

SN - 978-3-030-39342-7

T3 - Communications in Computer and Information Science

SP - 15

EP - 25

BT - Medical Image Understanding and Analysis

A2 - Zheng, Yalin

A2 - Williams, Bryan M.

A2 - Chen, Ke

PB - Springer Nature

T2 - Proceedings 23rd Conference on Medical Image Understanding and Analysis

Y2 - 24 July 2019 through 26 July 2019

ER -

Grall A, Hamidinekoo A, Malcolm P, Zwiggelaar R. Using a Conditional Generative Adversarial Network (cGAN) for Prostate Segmentation. In Zheng Y, Williams BM, Chen K, editors, Medical Image Understanding and Analysis: 23rd Conference, MIUA 2019, Liverpool, UK, July 24–26, 2019, Proceedings. Springer Nature. 2020. p. 15-25. (Communications in Computer and Information Science). Epub 2020 Jan 23. doi: 10.1007/978-3-030-39343-4_2

Using a Conditional Generative Adversarial Network (cGAN) for Prostate Segmentation

Abstract

Publication series

Conference

Keywords

UN SDGs

Access to Document

Permanent link

Other files and links

Fingerprint

Cite this