Automated annotations of epithelial cells and stroma in
hematoxylin–eosin-stained whole-slide images using cytokeratin
re-staining

J 2022

Automated annotations of epithelial cells and stroma in hematoxylin–eosin-stained whole-slide images using cytokeratin re-staining

BRÁZDIL, Tomáš, Matej GALLO, Rudolf NENUTIL, Andrej KUBANDA, Martin TOUFAR et. al.

Basic information

Original name

Automated annotations of epithelial cells and stroma in hematoxylin–eosin-stained whole-slide images using cytokeratin re-staining

Authors

BRÁZDIL, Tomáš (203 Czech Republic, belonging to the institution), Matej GALLO (703 Slovakia, belonging to the institution), Rudolf NENUTIL (203 Czech Republic), Andrej KUBANDA (703 Slovakia, belonging to the institution), Martin TOUFAR (203 Czech Republic) and Petr HOLUB (203 Czech Republic, guarantor, belonging to the institution)

Edition

The Journal of Pathology: Clinical Research, 2022, 2056-4538

Other information

Language

English

Type of outcome

Článek v odborném periodiku

Field of Study

10201 Computer sciences, information science, bioinformatics

Country of publisher

Czech Republic

Confidentiality degree

není předmětem státního či obchodního tajemství

References:

Wiley Online

Impact factor

Impact factor: 4.100

RIV identification code

RIV/00216224:14330/22:00125032

Organization unit

Faculty of Informatics

DOI

http://dx.doi.org/10.1002/cjp2.249

UT WoS

000712864400001

Keywords in English

U-Net; artificial intelligence; digital pathology; H&E; immunohistochemistry; deep learning; tissue registration

Abstract

V originále

The diagnosis of solid tumors of epithelial origin (carcinomas) represents a major part of the workload in clinical histopathology. Distinguishing stroma from epithelium is a critical component of artificial intelligence (AI) methods developed to detect and analyze carcinomas. In this paper, we propose a novel automated workflow that enables large-scale guidance of AI methods to identify the epithelial component. The workflow is based on re-staining existing hematoxylin and eosin (H&E) formalin-fixed paraffin-embedded (FFPE) slides by immunohistochemistry for cytokeratins - cytoskeleton components specific to epithelial cells. We also demonstrate how the automatically generated masks can be used to train modern AI image segmentation based on U-Net, resulting in reliable detection of epithelial regions in previously unseen H&E slides.

Links

LM2018140, research and development project

Name: e-Infrastruktura CZ (Acronym: e-INFRA CZ)

Investor: Ministry of Education, Youth and Sports of the CR

MUNI/A/1195/2021, interní kód MU

Name: Aplikovaný výzkum v oblastech vyhledávání, analýz a vizualizací rozsáhlých dat, zpracování přirozeného jazyka a aplikované umělé inteligence

Investor: Masaryk University

90125, large research infrastructures

Name: BBMRI-CZ III

Citovat

BRÁZDIL, Tomáš, Matej GALLO, Rudolf NENUTIL, Andrej KUBANDA, Martin TOUFAR and Petr HOLUB. Automated annotations of epithelial cells and stroma in hematoxylin–eosin-stained whole-slide images using cytokeratin re-staining. The Journal of Pathology: Clinical Research. 2022, vol. 8, No 2, p. 129-142. ISSN 2056-4538. Available from: https://dx.doi.org/10.1002/cjp2.249.

@article{1801217,
   author = {Brázdil, Tomáš and Gallo, Matej and Nenutil, Rudolf and Kubanda, Andrej and Toufar, Martin and Holub, Petr},
   article_number = {2},
   doi = {http://dx.doi.org/10.1002/cjp2.249},
   keywords = {U-Net; artificial intelligence; digital pathology; H&E; immunohistochemistry; deep learning; tissue registration},
   language = {eng},
   issn = {2056-4538},
   journal = {The Journal of Pathology: Clinical Research},
   title = {Automated annotations of epithelial cells and stroma in hematoxylin–eosin-stained whole-slide images using cytokeratin re-staining},
   url = {https://onlinelibrary.wiley.com/doi/10.1002/cjp2.249},
   volume = {8},
   year = {2022}
}

TY  - JOUR
ID  - 1801217
AU  - Brázdil, Tomáš - Gallo, Matej - Nenutil, Rudolf - Kubanda, Andrej - Toufar, Martin - Holub, Petr
PY  - 2022
TI  - Automated annotations of epithelial cells and stroma in hematoxylin–eosin-stained whole-slide images using cytokeratin re-staining
JF  - The Journal of Pathology: Clinical Research
VL  - 8
IS  - 2
SP  - 129-142
EP  - 129-142
SN  - 20564538
KW  - U-Net
KW  - artificial intelligence
KW  - digital pathology
KW  - H&E
KW  - immunohistochemistry
KW  - deep learning
KW  - tissue registration
UR  - https://onlinelibrary.wiley.com/doi/10.1002/cjp2.249
N2  - The diagnosis of solid tumors of epithelial origin (carcinomas) represents a major part of the workload in clinical histopathology. Distinguishing stroma from epithelium is a critical component of artificial intelligence (AI) methods developed to detect and analyze carcinomas. In this paper, we propose a novel automated workflow that enables large-scale guidance of AI methods to identify the epithelial component. The workflow is based on re-staining existing hematoxylin and eosin (H&E) formalin-fixed paraffin-embedded (FFPE) slides by immunohistochemistry for cytokeratins - cytoskeleton components specific to epithelial cells. We also demonstrate how the automatically generated masks can be used to train modern AI image segmentation based on U-Net, resulting in reliable detection of epithelial regions in previously unseen H&E slides.
ER  -

BRÁZDIL, Tomáš, Matej GALLO, Rudolf NENUTIL, Andrej KUBANDA, Martin TOUFAR and Petr HOLUB. Automated annotations of epithelial cells and stroma in hematoxylin–eosin-stained whole-slide images using cytokeratin re-staining. \textit{The Journal of Pathology: Clinical Research}. 2022, vol.~8, No~2, p.~129-142. ISSN~2056-4538. Available from: https://dx.doi.org/10.1002/cjp2.249.

Detailed Information on Publication Record