Wang Wei, Jiang Ran, Cui Ning, Li Qian, Yuan Feng, Xiao Zhifeng
Department of Breast Surgery, Hubei Provincial Clinical Research Center for Breast Cancer, Hubei Cancer Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, China.
Department of Thyroid and Breast Surgery, Maternal and Child Health Hospital of Hubei Province, Wuhan, Hubei, China.
Front Pharmacol. 2022 Jul 22;13:929755. doi: 10.3389/fphar.2022.929755. eCollection 2022.
Various imaging techniques combined with machine learning (ML) models have been used to build computer-aided diagnosis (CAD) systems for breast cancer (BC) detection and classification. The rise of deep learning models in recent years, represented by convolutional neural network (CNN) models, has pushed the accuracy of ML-based CAD systems to a new level that is comparable to human experts. Existing studies have explored the usage of a wide spectrum of CNN models for BC detection, and supervised learning has been the mainstream. In this study, we propose a semi-supervised learning framework based on the Vision Transformer (ViT). The ViT is a model that has been validated to outperform CNN models on numerous classification benchmarks but its application in BC detection has been rare. The proposed method offers a custom semi-supervised learning procedure that unifies both supervised and consistency training to enhance the robustness of the model. In addition, the method uses an adaptive token sampling technique that can strategically sample the most significant tokens from the input image, leading to an effective performance gain. We validate our method on two datasets with ultrasound and histopathology images. Results demonstrate that our method can consistently outperform the CNN baselines for both learning tasks. The code repository of the project is available at https://github.com/FeiYee/Breast-area-TWO.
各种成像技术与机器学习（ML）模型相结合，已被用于构建用于乳腺癌（BC）检测和分类的计算机辅助诊断（CAD）系统。近年来，以卷积神经网络（CNN）模型为代表的深度学习模型的兴起，将基于ML的CAD系统的准确性提升到了一个与人类专家相当的新水平。现有研究探索了多种CNN模型在BC检测中的应用，监督学习一直是主流。在本研究中，我们提出了一种基于视觉Transformer（ViT）的半监督学习框架。ViT是一种在众多分类基准上已被验证优于CNN模型的模型，但其在BC检测中的应用却很少见。所提出的方法提供了一种定制的半监督学习过程，该过程统一了监督训练和一致性训练，以增强模型的鲁棒性。此外，该方法使用了一种自适应令牌采样技术，该技术可以从输入图像中有策略地采样最重要的令牌，从而有效提高性能。我们在两个包含超声和组织病理学图像的数据集上验证了我们的方法。结果表明，我们的方法在这两个学习任务上都能持续优于CNN基线。该项目的代码库可在https://github.com/FeiYee/Breast-area-TWO获取。
