[论文] Unsupervised Domain Adaptation for Calcification Classification in Mam...

论文概要

研究领域: CV 作者: Xuan Liu, Derek L. Nguyen, Emily C. Barre 发布时间: 2025-07-09 arXiv: 2507.06826

中文摘要

基于深度学习的计算机辅助诊断（CAD）系统在乳腺癌诊断中表现出强大性能，特别是在乳腺X线摄影分类任务中。然而，跨多站点数据集的域偏移仍然是一个挑战，尤其是当模型应用于未见域时。在本工作中，我们提出了一个钙化分类框架，以改进跨多站点乳腺X线摄影数据集的恶性与良性乳腺疾病分类。该框架包含两个组件：（1）基于风格迁移模型（AdaIN和CycleGAN）的无监督域适配模块，无需额外标注即可生成特定厂商和特定技术的训练样本；（2）使用Swin Transformer V2作为骨干的监督分类模块。我们在三个数据集上评估了所提出的方法：在OPTIMAM（英国国家医疗服务体系；n=2994）上进行交叉验证，随后在EMBED（埃默里大学；n=125）和Duke钙化数据集v1（n=788）上进行外部验证。这些数据集涵盖多个厂商，包括全视野数字乳腺X线摄影和源自数字乳腺断层合成的合成2D图像。所提出的框架提高了EMBED（AUC 0.68至0.72）和Duke钙化数据集（AUC 0.68至0.73）的跨站点性能。这些发现表明，域适配可以减少域偏移并提高跨多站点数据集的钙化分类泛化能力。

原文摘要

Deep learning-based computer-aided diagnosis (CAD) systems have shown strong performance in breast cancer diagnosis, particularly for classification tasks in mammography. However, domain shifts across multi-site datasets remain a challenge, especially when models are applied to unseen domains. In this work, we proposed a calcification classification framework to improve malignant versus benign breast disease classification across multi-site mammography datasets. The framework consisted of two components: (1) an unsupervised domain adaptation module based on style transfer models (AdaIN and CycleGAN) to generate vendor-specific and technique-specific training samples without additional annotations, and (2) a supervised classification module using Swin Transformer V2 as the backbone. We eval...

--- *自动采集于 2026-07-09*

#论文 #arXiv #CV #小凯