发表机构
University of North Carolina at Charlotte; Wake Forest School of Medicine（北卡罗来纳大学夏洛特分校; 维克森林大学医学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
OCT-FedSIR提出可靠性感知谱框架，结合谱估计与重标记，在异构联邦眼科数据中有效识别并纠正注释噪声，提升分类准确率至86.73%。
AI 中文摘要
联邦学习能够在无需集中患者数据的情况下实现协作模型开发，但参与机构的注释可靠性并不总能得到保证。在眼科影像中，疾病患病率和类别构成的差异可能类似于由损坏监督引起的变化。我们提出了OCT-FedSIR，一种面向客户端依赖注释噪声和异构数据分布下的联邦OCT分类的可靠性感知谱框架。OCT-FedSIR结合了类别平衡谱估计、第一阶段logit调整、互补谱描述符、选择性谱重标记和噪声感知联邦优化。我们在Kermany、伊利诺伊大学芝加哥分校和Wake Forest数据集上，在对称和结构化非对称噪声以及三种非IID异构水平下评估了该框架。在117个实验条件下，OCT-FedSIR实现了86.73%的平均准确率，而RoFL为79.94%，FedCorr为78.75%。它在所有评估条件下正确区分了具有原始和损坏注释的客户端，而原始FedSIR识别过程在非对称噪声下鲁棒性较差。谱重标记恢复了77.2%的损坏注释，修正精度为91.3%，误修正率为3.5%。保留修正后的客户端平均比谱剪枝高出9.30个百分点。这些发现表明，注释噪声通常可以在不丢弃信息丰富的客户端数据的情况下被识别和纠正。
英文摘要
Federated learning enables collaborative model development without centralizing patient data, but annotation reliability at participating institutions cannot always be assumed. In ophthalmic imaging, differences in disease prevalence and class composition can resemble changes caused by corrupted supervision. We introduce OCT-FedSIR, a reliability-aware spectral framework for federated OCT classification under client-dependent annotation noise and heterogeneous data distributions. OCT-FedSIR combines class-balanced spectral estimation, Stage-I logit adjustment, complementary spectral descriptors, selective spectral relabeling, and noise-aware federated optimization. We evaluated the framework on the Kermany, University of Illinois Chicago, and Wake Forest datasets under symmetric and structured asymmetric noise and three levels of non-IID heterogeneity. Across 117 experimental conditions, OCT-FedSIR achieved a mean accuracy of 86.73%, compared with 79.94% for RoFL and 78.75% for FedCorr. It correctly separated clients with original and corrupted annotations across all evaluated conditions, while the original FedSIR identification procedure was less robust, particularly under asymmetric noise. Spectral relabeling recovered 77.2% of corrupted annotations with 91.3% correction precision and a 3.5% false-correction rate. Retaining corrected clients outperformed spectral pruning by 9.30 percentage points on average. These findings show that annotation noise can often be identified and corrected without discarding informative client data.
Comments38 pages, 7 figures, 4 tables, 6 tables in supplementary material
