发表机构
University of Technology Sydney; Shenzhen University（悉尼科技大学; 深圳大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对医学影像联邦遗忘问题，构建了Lethe基准测试，评估十二种方法在多任务、多遗忘粒度下的表现，发现遗忘请求难度而非方法本身是区分性能的关键，且医疗任务中移除客户端几乎不影响性能，需消除残留成员信息。
AI 中文摘要
联邦学习可让医学影像模型在多家医院间联合训练，而隐私法（尤其是GDPR的“被遗忘权”）要求从这类模型中移除某家医院、某类或某患者的影响，这一问题成为联邦遗忘问题。这一需求在医疗领域最为迫切，因为患者会撤回同意，医院也会退出合作。然而，几乎所有遗忘相关的研究证据都来自自然图像，其自然图像的异质性和任务结构与临床数据差异极大，因此现有方法能否迁移尚不明确，且目前尚无针对临床数据的统一协议。我们提出Lethe，这是一个面向医学影像联邦遗忘的基准测试。它在八个任务系列（涵盖分类、分割、去噪、跨模态合成、视觉-语言问答等）上，针对三种遗忘粒度，并结合重新训练的黄金标准，从效用、隐私和成本三个维度评估了十二种方法。核心结果表明，区分方法的关键在于遗忘请求的难度，而非方法本身。文献中占主导的简单移除任务会让保留效用的方法难以区分，只有困难的移除任务才能凸显它们的差异。更引人注目的是，在许多跨站点泛化的医疗任务中，移除一个客户端几乎不会改变任务性能，残留的成员信息才是必须被消除的信号。
英文摘要
Federated learning enables medical-imaging models to be trained across hospitals, and privacy law, most explicitly the GDPR ``right to be forgotten'', turns removing a hospital's, a class's, or a patient's influence from such a model into a federated unlearning problem. This need is most acute in medicine, where patients withdraw consent and hospitals leave collaborations. Yet nearly all unlearning evidence comes from natural images, whose heterogeneity and task structure differ sharply from clinical data, so it is unclear whether existing methods transfer, and no shared protocol covers clinical data. We present Lethe, a benchmark for federated unlearning in medical imaging. It evaluates twelve methods across eight task families, from classification and segmentation to denoising, cross-modality synthesis, and vision-language question answering, at three forgetting granularities and against a retrained gold standard on utility, privacy, and cost. The central result is that what separates methods is the difficulty of the forgetting request, not the method itself. The easy removals that dominate the literature leave the methods that preserve utility indistinguishable, while only hard ones separate them. More striking, on the many medical tasks that generalize across sites, forgetting a client barely changes task performance, leaving residual membership as the signal that must be erased.
Comments31 pages, 15 figures, benchmark paper
