A loss-based patch label denoising method for improving whole-slide image analysis using a convolutional neural network
Abstract This paper proposes a deep learning-based patch label denoising method (LossDiff) for improving the classification of whole-slide images of cancer using a convolutional neural network (CNN). Automated whole-slide image classification is often challenging, requiring a large amount of labeled...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2022-01-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-022-05001-8 |
| _version_ | 1819282310625755136 |
|---|---|
| author | Murtaza Ashraf Willmer Rafell Quiñones Robles Mujin Kim Young Sin Ko Mun Yong Yi |
| author_facet | Murtaza Ashraf Willmer Rafell Quiñones Robles Mujin Kim Young Sin Ko Mun Yong Yi |
| author_sort | Murtaza Ashraf |
| collection | DOAJ |
| description | Abstract This paper proposes a deep learning-based patch label denoising method (LossDiff) for improving the classification of whole-slide images of cancer using a convolutional neural network (CNN). Automated whole-slide image classification is often challenging, requiring a large amount of labeled data. Pathologists annotate the region of interest by marking malignant areas, which pose a high risk of introducing patch-based label noise by involving benign regions that are typically small in size within the malignant annotations, resulting in low classification accuracy with many Type-II errors. To overcome this critical problem, this paper presents a simple yet effective method for noisy patch classification. The proposed method, validated using stomach cancer images, provides a significant improvement compared to other existing methods in patch-based cancer classification, with accuracies of 98.81%, 97.30% and 89.47% for binary, ternary, and quaternary classes, respectively. Moreover, we conduct several experiments at different noise levels using a publicly available dataset to further demonstrate the robustness of the proposed method. Given the high cost of producing explicit annotations for whole-slide images and the unavoidable error-prone nature of the human annotation of medical images, the proposed method has practical implications for whole-slide image annotation and automated cancer diagnosis. |
| first_indexed | 2024-12-24T01:13:33Z |
| format | Article |
| id | doaj.art-0d62aa252d124e35a6bb66a84e357ac3 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-12-24T01:13:33Z |
| publishDate | 2022-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-0d62aa252d124e35a6bb66a84e357ac32022-12-21T17:22:50ZengNature PortfolioScientific Reports2045-23222022-01-0112111810.1038/s41598-022-05001-8A loss-based patch label denoising method for improving whole-slide image analysis using a convolutional neural networkMurtaza Ashraf0Willmer Rafell Quiñones Robles1Mujin Kim2Young Sin Ko3Mun Yong Yi4Department of Industrial and Systems Engineering, Graduate School of Knowledge Service Engineering, Korea Advanced Institute of Science and TechnologyDepartment of Industrial and Systems Engineering, Graduate School of Knowledge Service Engineering, Korea Advanced Institute of Science and TechnologyDepartment of Industrial and Systems Engineering, Graduate School of Knowledge Service Engineering, Korea Advanced Institute of Science and TechnologyPathology Center, Seegene Medical FoundationDepartment of Industrial and Systems Engineering, Graduate School of Knowledge Service Engineering, Korea Advanced Institute of Science and TechnologyAbstract This paper proposes a deep learning-based patch label denoising method (LossDiff) for improving the classification of whole-slide images of cancer using a convolutional neural network (CNN). Automated whole-slide image classification is often challenging, requiring a large amount of labeled data. Pathologists annotate the region of interest by marking malignant areas, which pose a high risk of introducing patch-based label noise by involving benign regions that are typically small in size within the malignant annotations, resulting in low classification accuracy with many Type-II errors. To overcome this critical problem, this paper presents a simple yet effective method for noisy patch classification. The proposed method, validated using stomach cancer images, provides a significant improvement compared to other existing methods in patch-based cancer classification, with accuracies of 98.81%, 97.30% and 89.47% for binary, ternary, and quaternary classes, respectively. Moreover, we conduct several experiments at different noise levels using a publicly available dataset to further demonstrate the robustness of the proposed method. Given the high cost of producing explicit annotations for whole-slide images and the unavoidable error-prone nature of the human annotation of medical images, the proposed method has practical implications for whole-slide image annotation and automated cancer diagnosis.https://doi.org/10.1038/s41598-022-05001-8 |
| spellingShingle | Murtaza Ashraf Willmer Rafell Quiñones Robles Mujin Kim Young Sin Ko Mun Yong Yi A loss-based patch label denoising method for improving whole-slide image analysis using a convolutional neural network Scientific Reports |
| title | A loss-based patch label denoising method for improving whole-slide image analysis using a convolutional neural network |
| title_full | A loss-based patch label denoising method for improving whole-slide image analysis using a convolutional neural network |
| title_fullStr | A loss-based patch label denoising method for improving whole-slide image analysis using a convolutional neural network |
| title_full_unstemmed | A loss-based patch label denoising method for improving whole-slide image analysis using a convolutional neural network |
| title_short | A loss-based patch label denoising method for improving whole-slide image analysis using a convolutional neural network |
| title_sort | loss based patch label denoising method for improving whole slide image analysis using a convolutional neural network |
| url | https://doi.org/10.1038/s41598-022-05001-8 |
| work_keys_str_mv | AT murtazaashraf alossbasedpatchlabeldenoisingmethodforimprovingwholeslideimageanalysisusingaconvolutionalneuralnetwork AT willmerrafellquinonesrobles alossbasedpatchlabeldenoisingmethodforimprovingwholeslideimageanalysisusingaconvolutionalneuralnetwork AT mujinkim alossbasedpatchlabeldenoisingmethodforimprovingwholeslideimageanalysisusingaconvolutionalneuralnetwork AT youngsinko alossbasedpatchlabeldenoisingmethodforimprovingwholeslideimageanalysisusingaconvolutionalneuralnetwork AT munyongyi alossbasedpatchlabeldenoisingmethodforimprovingwholeslideimageanalysisusingaconvolutionalneuralnetwork AT murtazaashraf lossbasedpatchlabeldenoisingmethodforimprovingwholeslideimageanalysisusingaconvolutionalneuralnetwork AT willmerrafellquinonesrobles lossbasedpatchlabeldenoisingmethodforimprovingwholeslideimageanalysisusingaconvolutionalneuralnetwork AT mujinkim lossbasedpatchlabeldenoisingmethodforimprovingwholeslideimageanalysisusingaconvolutionalneuralnetwork AT youngsinko lossbasedpatchlabeldenoisingmethodforimprovingwholeslideimageanalysisusingaconvolutionalneuralnetwork AT munyongyi lossbasedpatchlabeldenoisingmethodforimprovingwholeslideimageanalysisusingaconvolutionalneuralnetwork |