Staingan: Stain style transfer for digital histological images

M. Tarek Shaban; Christoph Baur; Nassir Navab; Shadi Albarqouni

doi:10.1109/ISBI.2019.8759152

Staingan: Stain style transfer for digital histological images

M. Tarek Shaban, Christoph Baur, Nassir Navab, Shadi Albarqouni

Informatics 16 - Chair of Computer Aided Medical Procedures

Technical University of Munich

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

220 Scopus citations

Abstract

Digitized Histological diagnosis is in increasing demand. However, color variations due to various factors are imposing obstacles to the diagnosis process. The problem of stain color variations is a well-defined problem with many proposed solutions. Most of these solutions are highly dependent on a reference template slide. We propose a deep-learning solution inspired by CycleGANs that is trained end-to-end, eliminating the need for an expert to pick a representative reference slide. Our approach showed superior results quantitatively and qualitatively against the state of the art methods (10% improvement visually using SSIM). We further validated our method on a clinical use-case, namely Breast Cancer tumor classification, showing a 12% increase in AUC. The code is made publicly available ¹.¹https://github.com/xtarx/StainGAN.

Original language	English
Title of host publication	ISBI 2019 - 2019 IEEE International Symposium on Biomedical Imaging
Publisher	IEEE Computer Society
Pages	953-956
Number of pages	4
ISBN (Electronic)	9781538636411
DOIs	https://doi.org/10.1109/ISBI.2019.8759152
State	Published - Apr 2019
Event	16th IEEE International Symposium on Biomedical Imaging, ISBI 2019 - Venice, Italy Duration: 8 Apr 2019 → 11 Apr 2019

Publication series

Name	Proceedings - International Symposium on Biomedical Imaging
Volume	2019-April
ISSN (Print)	1945-7928
ISSN (Electronic)	1945-8452

Conference

Conference	16th IEEE International Symposium on Biomedical Imaging, ISBI 2019
Country/Territory	Italy
City	Venice
Period	8/04/19 → 11/04/19

Keywords

Deep learning
Generative adversarial networks
Histology images
Stain normalization

UN SDGs

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

Access to Document

10.1109/ISBI.2019.8759152

Cite this

Shaban, M. T., Baur, C., Navab, N., & Albarqouni, S. (2019). Staingan: Stain style transfer for digital histological images. In ISBI 2019 - 2019 IEEE International Symposium on Biomedical Imaging (pp. 953-956). Article 8759152 (Proceedings - International Symposium on Biomedical Imaging; Vol. 2019-April). IEEE Computer Society. https://doi.org/10.1109/ISBI.2019.8759152

@inproceedings{d54034bc0e39476b87ded6f045a73202,

title = "Staingan: Stain style transfer for digital histological images",

abstract = "Digitized Histological diagnosis is in increasing demand. However, color variations due to various factors are imposing obstacles to the diagnosis process. The problem of stain color variations is a well-defined problem with many proposed solutions. Most of these solutions are highly dependent on a reference template slide. We propose a deep-learning solution inspired by CycleGANs that is trained end-to-end, eliminating the need for an expert to pick a representative reference slide. Our approach showed superior results quantitatively and qualitatively against the state of the art methods (10% improvement visually using SSIM). We further validated our method on a clinical use-case, namely Breast Cancer tumor classification, showing a 12% increase in AUC. The code is made publicly available 1.1https://github.com/xtarx/StainGAN.",

keywords = "Deep learning, Generative adversarial networks, Histology images, Stain normalization",

author = "Shaban, {M. Tarek} and Christoph Baur and Nassir Navab and Shadi Albarqouni",

note = "Publisher Copyright: {\textcopyright} 2019 IEEE.; 16th IEEE International Symposium on Biomedical Imaging, ISBI 2019 ; Conference date: 08-04-2019 Through 11-04-2019",

year = "2019",

month = apr,

doi = "10.1109/ISBI.2019.8759152",

language = "English",

series = "Proceedings - International Symposium on Biomedical Imaging",

publisher = "IEEE Computer Society",

pages = "953--956",

booktitle = "ISBI 2019 - 2019 IEEE International Symposium on Biomedical Imaging",

}

Shaban, MT, Baur, C, Navab, N & Albarqouni, S 2019, Staingan: Stain style transfer for digital histological images. in ISBI 2019 - 2019 IEEE International Symposium on Biomedical Imaging., 8759152, Proceedings - International Symposium on Biomedical Imaging, vol. 2019-April, IEEE Computer Society, pp. 953-956, 16th IEEE International Symposium on Biomedical Imaging, ISBI 2019, Venice, Italy, 8/04/19. https://doi.org/10.1109/ISBI.2019.8759152

Staingan: Stain style transfer for digital histological images. / Shaban, M. Tarek; Baur, Christoph; Navab, Nassir et al.
ISBI 2019 - 2019 IEEE International Symposium on Biomedical Imaging. IEEE Computer Society, 2019. p. 953-956 8759152 (Proceedings - International Symposium on Biomedical Imaging; Vol. 2019-April).

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

TY - GEN

T1 - Staingan

T2 - 16th IEEE International Symposium on Biomedical Imaging, ISBI 2019

AU - Shaban, M. Tarek

AU - Baur, Christoph

AU - Navab, Nassir

AU - Albarqouni, Shadi

PY - 2019/4

Y1 - 2019/4

N2 - Digitized Histological diagnosis is in increasing demand. However, color variations due to various factors are imposing obstacles to the diagnosis process. The problem of stain color variations is a well-defined problem with many proposed solutions. Most of these solutions are highly dependent on a reference template slide. We propose a deep-learning solution inspired by CycleGANs that is trained end-to-end, eliminating the need for an expert to pick a representative reference slide. Our approach showed superior results quantitatively and qualitatively against the state of the art methods (10% improvement visually using SSIM). We further validated our method on a clinical use-case, namely Breast Cancer tumor classification, showing a 12% increase in AUC. The code is made publicly available 1.1https://github.com/xtarx/StainGAN.

AB - Digitized Histological diagnosis is in increasing demand. However, color variations due to various factors are imposing obstacles to the diagnosis process. The problem of stain color variations is a well-defined problem with many proposed solutions. Most of these solutions are highly dependent on a reference template slide. We propose a deep-learning solution inspired by CycleGANs that is trained end-to-end, eliminating the need for an expert to pick a representative reference slide. Our approach showed superior results quantitatively and qualitatively against the state of the art methods (10% improvement visually using SSIM). We further validated our method on a clinical use-case, namely Breast Cancer tumor classification, showing a 12% increase in AUC. The code is made publicly available 1.1https://github.com/xtarx/StainGAN.

KW - Deep learning

KW - Generative adversarial networks

KW - Histology images

KW - Stain normalization

UR - http://www.scopus.com/inward/record.url?scp=85072178672&partnerID=8YFLogxK

U2 - 10.1109/ISBI.2019.8759152

DO - 10.1109/ISBI.2019.8759152

M3 - Conference contribution

AN - SCOPUS:85072178672

T3 - Proceedings - International Symposium on Biomedical Imaging

SP - 953

EP - 956

BT - ISBI 2019 - 2019 IEEE International Symposium on Biomedical Imaging

PB - IEEE Computer Society

Y2 - 8 April 2019 through 11 April 2019

ER -

Staingan: Stain style transfer for digital histological images

Abstract

Publication series

Conference

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this