{"title":"利用半监督 CycleGAN 进行领域转换,提高甲状腺组织图像的分类性能。","authors":"Yoshihito Ichiuji, Shingo Mabu, Satomi Hatta, Kunihiro Inai, Shohei Higuchi, Shoji Kido","doi":"10.1007/s11548-024-03061-x","DOIUrl":null,"url":null,"abstract":"<p><strong>Purpose: </strong>A large number of research has been conducted on the classification of medical images using deep learning. The thyroid tissue images can be also classified by cancer types. Deep learning requires a large amount of data, but every medical institution cannot collect sufficient number of data for deep learning. In that case, we can consider a case where a classifier trained at a certain medical institution that has a sufficient number of data is reused at other institutions. However, when using data from multiple institutions, it is necessary to unify the feature distribution because the feature of the data differs due to differences in data acquisition conditions.</p><p><strong>Methods: </strong>To unify the feature distribution, the data from Institution T are transformed to have the closer distribution to that from Institution S by applying a domain transformation using semi-supervised CycleGAN. The proposed method enhances CycleGAN considering the feature distribution of classes for making appropriate domain transformation for classification. In addition, to address the problem of imbalanced data with different numbers of data for each cancer type, several methods dealing with imbalanced data are applied to semi-supervised CycleGAN.</p><p><strong>Results: </strong>The experimental results showed that the classification performance was enhanced when the dataset from Institution S was used as training data and the testing dataset from Institution T was classified after applying domain transformation. In addition, focal loss contributed to improving the mean F1 score the best as a method that addresses the class imbalance.</p><p><strong>Conclusion: </strong>The proposed method achieved the domain transformation of thyroid tissue images between two domains, where it retained the important features related to the classes across domains and showed the best F1 score with significant differences compared with other methods. In addition, the proposed method was further enhanced by addressing the class imbalance of the dataset.</p>","PeriodicalId":51251,"journal":{"name":"International Journal of Computer Assisted Radiology and Surgery","volume":null,"pages":null},"PeriodicalIF":2.3000,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Domain transformation using semi-supervised CycleGAN for improving performance of classifying thyroid tissue images.\",\"authors\":\"Yoshihito Ichiuji, Shingo Mabu, Satomi Hatta, Kunihiro Inai, Shohei Higuchi, Shoji Kido\",\"doi\":\"10.1007/s11548-024-03061-x\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Purpose: </strong>A large number of research has been conducted on the classification of medical images using deep learning. The thyroid tissue images can be also classified by cancer types. Deep learning requires a large amount of data, but every medical institution cannot collect sufficient number of data for deep learning. In that case, we can consider a case where a classifier trained at a certain medical institution that has a sufficient number of data is reused at other institutions. However, when using data from multiple institutions, it is necessary to unify the feature distribution because the feature of the data differs due to differences in data acquisition conditions.</p><p><strong>Methods: </strong>To unify the feature distribution, the data from Institution T are transformed to have the closer distribution to that from Institution S by applying a domain transformation using semi-supervised CycleGAN. The proposed method enhances CycleGAN considering the feature distribution of classes for making appropriate domain transformation for classification. In addition, to address the problem of imbalanced data with different numbers of data for each cancer type, several methods dealing with imbalanced data are applied to semi-supervised CycleGAN.</p><p><strong>Results: </strong>The experimental results showed that the classification performance was enhanced when the dataset from Institution S was used as training data and the testing dataset from Institution T was classified after applying domain transformation. In addition, focal loss contributed to improving the mean F1 score the best as a method that addresses the class imbalance.</p><p><strong>Conclusion: </strong>The proposed method achieved the domain transformation of thyroid tissue images between two domains, where it retained the important features related to the classes across domains and showed the best F1 score with significant differences compared with other methods. In addition, the proposed method was further enhanced by addressing the class imbalance of the dataset.</p>\",\"PeriodicalId\":51251,\"journal\":{\"name\":\"International Journal of Computer Assisted Radiology and Surgery\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":2.3000,\"publicationDate\":\"2024-11-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"International Journal of Computer Assisted Radiology and Surgery\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://doi.org/10.1007/s11548-024-03061-x\",\"RegionNum\":3,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"2024/1/18 0:00:00\",\"PubModel\":\"Epub\",\"JCR\":\"Q3\",\"JCRName\":\"ENGINEERING, BIOMEDICAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Journal of Computer Assisted Radiology and Surgery","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.1007/s11548-024-03061-x","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/1/18 0:00:00","PubModel":"Epub","JCR":"Q3","JCRName":"ENGINEERING, BIOMEDICAL","Score":null,"Total":0}
引用次数: 0
摘要
目的:人们利用深度学习对医学图像进行了大量分类研究。甲状腺组织图像也可按癌症类型进行分类。深度学习需要大量数据,但每个医疗机构都无法收集到足够数量的数据用于深度学习。在这种情况下,我们可以考虑将某个医疗机构训练的分类器在其他医疗机构重复使用,因为该医疗机构拥有足够数量的数据。但是,在使用多个机构的数据时,由于数据获取条件的不同,数据的特征也不尽相同,因此有必要统一特征分布:为了统一特征分布,使用半监督 CycleGAN 进行域转换,将来自 T 机构的数据转换为与来自 S 机构的数据分布更接近的数据。所提出的方法增强了 CycleGAN 的功能,考虑到了类的特征分布,从而为分类进行适当的域转换。此外,为了解决每种癌症类型的数据数量不同的不平衡数据问题,在半监督 CycleGAN 中应用了几种处理不平衡数据的方法:实验结果表明,当使用 S 机构的数据集作为训练数据,并对 T 机构的测试数据集进行域转换后进行分类时,分类性能得到了提高。此外,作为一种解决类不平衡的方法,焦点丢失对提高平均 F1 分数的贡献最大:结论:所提出的方法实现了甲状腺组织图像在两个域之间的域转换,保留了与跨域类别相关的重要特征,与其他方法相比,F1得分最高,差异显著。此外,通过解决数据集的类不平衡问题,所提出的方法得到了进一步增强。
Domain transformation using semi-supervised CycleGAN for improving performance of classifying thyroid tissue images.
Purpose: A large number of research has been conducted on the classification of medical images using deep learning. The thyroid tissue images can be also classified by cancer types. Deep learning requires a large amount of data, but every medical institution cannot collect sufficient number of data for deep learning. In that case, we can consider a case where a classifier trained at a certain medical institution that has a sufficient number of data is reused at other institutions. However, when using data from multiple institutions, it is necessary to unify the feature distribution because the feature of the data differs due to differences in data acquisition conditions.
Methods: To unify the feature distribution, the data from Institution T are transformed to have the closer distribution to that from Institution S by applying a domain transformation using semi-supervised CycleGAN. The proposed method enhances CycleGAN considering the feature distribution of classes for making appropriate domain transformation for classification. In addition, to address the problem of imbalanced data with different numbers of data for each cancer type, several methods dealing with imbalanced data are applied to semi-supervised CycleGAN.
Results: The experimental results showed that the classification performance was enhanced when the dataset from Institution S was used as training data and the testing dataset from Institution T was classified after applying domain transformation. In addition, focal loss contributed to improving the mean F1 score the best as a method that addresses the class imbalance.
Conclusion: The proposed method achieved the domain transformation of thyroid tissue images between two domains, where it retained the important features related to the classes across domains and showed the best F1 score with significant differences compared with other methods. In addition, the proposed method was further enhanced by addressing the class imbalance of the dataset.
期刊介绍:
The International Journal for Computer Assisted Radiology and Surgery (IJCARS) is a peer-reviewed journal that provides a platform for closing the gap between medical and technical disciplines, and encourages interdisciplinary research and development activities in an international environment.