Modelling the Ki67 Index in Synthetic HE-Stained Images Using Conditional StyleGAN Model

Hematoxylin and Eosin (HE) staining is the gold standard in histopathological examination of cancer tissue, representing the first step towards cancer diagnosis. The second step is a series of immunohistochemical stainings, including cell proliferation markers called the Ki67 index. Deep learning mo...

Full description

Saved in:
Bibliographic Details
Main Authors: Lucia Piatriková, Katarína Tobiášová, Andrej Štefák, Dominika Petríková, Lukáš Plank, Ivan Cimrák
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Bioengineering
Subjects:
Online Access:https://www.mdpi.com/2306-5354/12/5/476
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850127277772242944
author Lucia Piatriková
Katarína Tobiášová
Andrej Štefák
Dominika Petríková
Lukáš Plank
Ivan Cimrák
author_facet Lucia Piatriková
Katarína Tobiášová
Andrej Štefák
Dominika Petríková
Lukáš Plank
Ivan Cimrák
author_sort Lucia Piatriková
collection DOAJ
description Hematoxylin and Eosin (HE) staining is the gold standard in histopathological examination of cancer tissue, representing the first step towards cancer diagnosis. The second step is a series of immunohistochemical stainings, including cell proliferation markers called the Ki67 index. Deep learning models offer promising solutions for improving medical diagnostics, while generative models provide additional explainability of predictive models that is essential for their adoption in clinical practice. Our previous work introduced a novel approach that utilises a conditional StyleGAN model for generating HE-stained images conditioned on the Ki67 index. This study proposes to employ this model for generating sequences of HE-stained images reflecting varying Ki67 index values. Sequences enable exploration of hidden relationships between HE and Ki67 staining and can enhance the explainability of predictive models, e.g., by generating counterfactual examples. While our previous research focused on assessing the quality of generated HE images, this study extends that work by evaluating the model’s ability to capture Ki67-related variations in HE-stained images. Additionally, expert pathologists evaluated generated sequences and proposed criteria for assessing their relevance. Our findings demonstrate the potential of the conditional StyleGAN model as part of an explainable framework for analysing and predicting immunohistochemical information from HE-stained images. Results highlight the relevance of generative models in histopathology and their potential applications in cancer progression analysis.
format Article
id doaj-art-1a8a83917b1b4b8bbbb574ac63d66297
institution OA Journals
issn 2306-5354
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Bioengineering
spelling doaj-art-1a8a83917b1b4b8bbbb574ac63d662972025-08-20T02:33:43ZengMDPI AGBioengineering2306-53542025-04-0112547610.3390/bioengineering12050476Modelling the Ki67 Index in Synthetic HE-Stained Images Using Conditional StyleGAN ModelLucia Piatriková0Katarína Tobiášová1Andrej Štefák2Dominika Petríková3Lukáš Plank4Ivan Cimrák5Department of Software Technologies, Faculty of Management Science and Informatics, University of Žilina, 010 26 Žilina, SlovakiaDepartment of Pathological Anatomy, Jessenius Faculty of Medicine in Martin, Comenius University Bratislava and University Hospital, 036 01 Martin, SlovakiaDepartment of Pathological Anatomy, Jessenius Faculty of Medicine in Martin, Comenius University Bratislava and University Hospital, 036 01 Martin, SlovakiaDepartment of Software Technologies, Faculty of Management Science and Informatics, University of Žilina, 010 26 Žilina, SlovakiaDepartment of Pathological Anatomy, Jessenius Faculty of Medicine in Martin, Comenius University Bratislava and University Hospital, 036 01 Martin, SlovakiaDepartment of Software Technologies, Faculty of Management Science and Informatics, University of Žilina, 010 26 Žilina, SlovakiaHematoxylin and Eosin (HE) staining is the gold standard in histopathological examination of cancer tissue, representing the first step towards cancer diagnosis. The second step is a series of immunohistochemical stainings, including cell proliferation markers called the Ki67 index. Deep learning models offer promising solutions for improving medical diagnostics, while generative models provide additional explainability of predictive models that is essential for their adoption in clinical practice. Our previous work introduced a novel approach that utilises a conditional StyleGAN model for generating HE-stained images conditioned on the Ki67 index. This study proposes to employ this model for generating sequences of HE-stained images reflecting varying Ki67 index values. Sequences enable exploration of hidden relationships between HE and Ki67 staining and can enhance the explainability of predictive models, e.g., by generating counterfactual examples. While our previous research focused on assessing the quality of generated HE images, this study extends that work by evaluating the model’s ability to capture Ki67-related variations in HE-stained images. Additionally, expert pathologists evaluated generated sequences and proposed criteria for assessing their relevance. Our findings demonstrate the potential of the conditional StyleGAN model as part of an explainable framework for analysing and predicting immunohistochemical information from HE-stained images. Results highlight the relevance of generative models in histopathology and their potential applications in cancer progression analysis.https://www.mdpi.com/2306-5354/12/5/476hematoxylin and eosinKi67conditional GANStyleGANdigital pathology
spellingShingle Lucia Piatriková
Katarína Tobiášová
Andrej Štefák
Dominika Petríková
Lukáš Plank
Ivan Cimrák
Modelling the Ki67 Index in Synthetic HE-Stained Images Using Conditional StyleGAN Model
Bioengineering
hematoxylin and eosin
Ki67
conditional GAN
StyleGAN
digital pathology
title Modelling the Ki67 Index in Synthetic HE-Stained Images Using Conditional StyleGAN Model
title_full Modelling the Ki67 Index in Synthetic HE-Stained Images Using Conditional StyleGAN Model
title_fullStr Modelling the Ki67 Index in Synthetic HE-Stained Images Using Conditional StyleGAN Model
title_full_unstemmed Modelling the Ki67 Index in Synthetic HE-Stained Images Using Conditional StyleGAN Model
title_short Modelling the Ki67 Index in Synthetic HE-Stained Images Using Conditional StyleGAN Model
title_sort modelling the ki67 index in synthetic he stained images using conditional stylegan model
topic hematoxylin and eosin
Ki67
conditional GAN
StyleGAN
digital pathology
url https://www.mdpi.com/2306-5354/12/5/476
work_keys_str_mv AT luciapiatrikova modellingtheki67indexinsynthetichestainedimagesusingconditionalstyleganmodel
AT katarinatobiasova modellingtheki67indexinsynthetichestainedimagesusingconditionalstyleganmodel
AT andrejstefak modellingtheki67indexinsynthetichestainedimagesusingconditionalstyleganmodel
AT dominikapetrikova modellingtheki67indexinsynthetichestainedimagesusingconditionalstyleganmodel
AT lukasplank modellingtheki67indexinsynthetichestainedimagesusingconditionalstyleganmodel
AT ivancimrak modellingtheki67indexinsynthetichestainedimagesusingconditionalstyleganmodel