论文标题
使用潜在扩散模型来生成匿名胸部X光片，用于训练胸部异常分类系统
Generation of Anonymous Chest Radiographs Using Latent Diffusion Models for Training Thoracic Abnormality Classification Systems
论文作者
论文摘要
大规模胸部X射线数据集的可用性是在胸部异常检测和分类中开发出良好表现的深度学习算法的要求。但是，由于患者重新识别的风险，胸部X光片中的生物识别标识符阻碍了出于研究目的的公众共享。为了解决此问题，合成数据生成为匿名医学图像提供了解决方案。这项工作采用潜在扩散模型来合成高质量类条件图像的匿名胸部X射线数据集。我们提出了一种增强隐私的抽样策略，以确保在图像生成过程中生物识别信息的不转移。在胸部异常分类任务上评估了生成图像的质量以及作为独家培训数据的可行性。与真实分类器相比，我们在接收器操作特征曲线下仅3.5％的性能差距获得了竞争成果。
The availability of large-scale chest X-ray datasets is a requirement for developing well-performing deep learning-based algorithms in thoracic abnormality detection and classification. However, biometric identifiers in chest radiographs hinder the public sharing of such data for research purposes due to the risk of patient re-identification. To counteract this issue, synthetic data generation offers a solution for anonymizing medical images. This work employs a latent diffusion model to synthesize an anonymous chest X-ray dataset of high-quality class-conditional images. We propose a privacy-enhancing sampling strategy to ensure the non-transference of biometric information during the image generation process. The quality of the generated images and the feasibility of serving as exclusive training data are evaluated on a thoracic abnormality classification task. Compared to a real classifier, we achieve competitive results with a performance gap of only 3.5% in the area under the receiver operating characteristic curve.
