A semi-supervised learning approach with consistency regularization for tumor histopathological images analysis
IntroductionManual inspection of histopathological images is important in clinical cancer diagnosis. Pathologists implement pathological diagnosis and prognostic evaluation through the microscopic examination of histopathological slices. This entire process is time-consuming, laborious, and challeng...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Frontiers Media S.A.
2023-01-01
|
Series: | Frontiers in Oncology |
Subjects: | |
Online Access: | https://www.frontiersin.org/articles/10.3389/fonc.2022.1044026/full |
_version_ | 1797956879287058432 |
---|---|
author | Yanyun Jiang Xiaodan Sui Yanhui Ding Wei Xiao Yuanjie Zheng Yongxin Zhang |
author_facet | Yanyun Jiang Xiaodan Sui Yanhui Ding Wei Xiao Yuanjie Zheng Yongxin Zhang |
author_sort | Yanyun Jiang |
collection | DOAJ |
description | IntroductionManual inspection of histopathological images is important in clinical cancer diagnosis. Pathologists implement pathological diagnosis and prognostic evaluation through the microscopic examination of histopathological slices. This entire process is time-consuming, laborious, and challenging for pathologists. The modern use of whole-slide imaging, which scans histopathology slides to digital slices, and analysis using computer-aided diagnosis is an essential problem.MethodsTo solve the problem of difficult labeling of histopathological data, and improve the flexibility of histopathological analysis in clinical applications, we herein propose a semi-supervised learning algorithm coupled with consistency regularization strategy, called“Semi- supervised Histopathology Analysis Network”(Semi-His-Net), for automated normal-versus-tumor and subtype classifications. Specifically, when inputted disturbing versions of the same image, the model should predict similar outputs. Based on this, the model itself can assign artificial labels to unlabeled data for subsequent model training, thereby effectively reducing the labeled data required for training.ResultsOur Semi-His-Net is able to classify patches from breast cancer histopathological images into normal tissue and three other different tumor subtypes, achieving an accuracy was 90%. The average AUC of cross-classification between tumors reached 0.893.DiscussionTo overcome the limitations of visual inspection by pathologists for histopathology images, such as long time and low repeatability, we have developed a deep learning-based framework (Semi-His-Net) for automatic classification subdivision of the subtypes contained in the whole pathological images. This learning-based framework has great potential to improve the efficiency and repeatability of histopathological image diagnosis. |
first_indexed | 2024-04-10T23:56:45Z |
format | Article |
id | doaj.art-668ecf134f1d4e868b7d6c321d52e31c |
institution | Directory Open Access Journal |
issn | 2234-943X |
language | English |
last_indexed | 2024-04-10T23:56:45Z |
publishDate | 2023-01-01 |
publisher | Frontiers Media S.A. |
record_format | Article |
series | Frontiers in Oncology |
spelling | doaj.art-668ecf134f1d4e868b7d6c321d52e31c2023-01-10T12:46:25ZengFrontiers Media S.A.Frontiers in Oncology2234-943X2023-01-011210.3389/fonc.2022.10440261044026A semi-supervised learning approach with consistency regularization for tumor histopathological images analysisYanyun Jiang0Xiaodan Sui1Yanhui Ding2Wei Xiao3Yuanjie Zheng4Yongxin Zhang5School of Mathematics and Statistics, Shandong Normal University, Jinan, ChinaSchool of Mathematics and Statistics, Shandong Normal University, Jinan, ChinaSchool of Mathematics and Statistics, Shandong Normal University, Jinan, ChinaShandong Provincial Hospital, Shandong University, Jinan, ChinaSchool of Mathematics and Statistics, Shandong Normal University, Jinan, ChinaSchool of Mathematics and Statistics, Shandong Normal University, Jinan, ChinaIntroductionManual inspection of histopathological images is important in clinical cancer diagnosis. Pathologists implement pathological diagnosis and prognostic evaluation through the microscopic examination of histopathological slices. This entire process is time-consuming, laborious, and challenging for pathologists. The modern use of whole-slide imaging, which scans histopathology slides to digital slices, and analysis using computer-aided diagnosis is an essential problem.MethodsTo solve the problem of difficult labeling of histopathological data, and improve the flexibility of histopathological analysis in clinical applications, we herein propose a semi-supervised learning algorithm coupled with consistency regularization strategy, called“Semi- supervised Histopathology Analysis Network”(Semi-His-Net), for automated normal-versus-tumor and subtype classifications. Specifically, when inputted disturbing versions of the same image, the model should predict similar outputs. Based on this, the model itself can assign artificial labels to unlabeled data for subsequent model training, thereby effectively reducing the labeled data required for training.ResultsOur Semi-His-Net is able to classify patches from breast cancer histopathological images into normal tissue and three other different tumor subtypes, achieving an accuracy was 90%. The average AUC of cross-classification between tumors reached 0.893.DiscussionTo overcome the limitations of visual inspection by pathologists for histopathology images, such as long time and low repeatability, we have developed a deep learning-based framework (Semi-His-Net) for automatic classification subdivision of the subtypes contained in the whole pathological images. This learning-based framework has great potential to improve the efficiency and repeatability of histopathological image diagnosis.https://www.frontiersin.org/articles/10.3389/fonc.2022.1044026/fulldeep learningsemi-supervised learningdata augmentationconsistency regularizatonwhole-slide images |
spellingShingle | Yanyun Jiang Xiaodan Sui Yanhui Ding Wei Xiao Yuanjie Zheng Yongxin Zhang A semi-supervised learning approach with consistency regularization for tumor histopathological images analysis Frontiers in Oncology deep learning semi-supervised learning data augmentation consistency regularizaton whole-slide images |
title | A semi-supervised learning approach with consistency regularization for tumor histopathological images analysis |
title_full | A semi-supervised learning approach with consistency regularization for tumor histopathological images analysis |
title_fullStr | A semi-supervised learning approach with consistency regularization for tumor histopathological images analysis |
title_full_unstemmed | A semi-supervised learning approach with consistency regularization for tumor histopathological images analysis |
title_short | A semi-supervised learning approach with consistency regularization for tumor histopathological images analysis |
title_sort | semi supervised learning approach with consistency regularization for tumor histopathological images analysis |
topic | deep learning semi-supervised learning data augmentation consistency regularizaton whole-slide images |
url | https://www.frontiersin.org/articles/10.3389/fonc.2022.1044026/full |
work_keys_str_mv | AT yanyunjiang asemisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis AT xiaodansui asemisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis AT yanhuiding asemisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis AT weixiao asemisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis AT yuanjiezheng asemisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis AT yongxinzhang asemisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis AT yanyunjiang semisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis AT xiaodansui semisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis AT yanhuiding semisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis AT weixiao semisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis AT yuanjiezheng semisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis AT yongxinzhang semisupervisedlearningapproachwithconsistencyregularizationfortumorhistopathologicalimagesanalysis |