文献速递眼底病与葡萄膜炎

使用联邦学习与区块链的隐私保护技术防御视网膜影像对抗攻击

Privacy-Preserving Technology Using Federated Learning and Blockchain in Protecting against Adversarial Attacks for Retinal Imaging.
PubMed 文献 · Ophthalmology · 2024-10-16

该回顾性多队列研究利用来自新加坡、中国和台湾的27145张图像,设计了一种新的联邦学习聚合方法,用于在特征分布偏斜和标签分布不平衡场景下检测近视性黄斑变性和OCT黄斑疾病。其联邦学习模型对近视性黄斑变性检测的AUC为0.868±0.009,对OCT黄斑疾病分类的AUC为0.970±0.012,并在标签翻转和干净标签对抗攻击下保持稳健,区块链集成对模型开发时间影响最小。

要点

  • 联邦学习模型检测近视性黄斑变性的AUC为0.868±0.009,OCT黄斑疾病分类的AUC为0.970±0.012。
  • 在标签翻转攻击下,该联邦学习模型AUC为0.861±0.019,与集中式模型(AUC 0.856±0.015)相似,高于其他联邦学习模型(AUC 0.578-0.819)。
  • 在干净标签攻击下,该联邦学习模型AUC为0.878±0.006,与集中式模型(AUC 0.878±0.001)相当,优于其他先进联邦学习模型(AUC 0.529至0.838)。

核心结论

  • 所提出的联邦学习算法克服了传统联邦学习在非独立同分布情况下的缺点,并对对抗攻击保持稳健。
  • 区块链的集成在模型更新传输过程中增加了进一步的安全性。
  • 区块链赋能的联邦学习可以成为可信的协作健康人工智能研究平台。

研究设计与样本

  • 回顾性、多队列研究,使用来自新加坡、中国和台湾的共27145张图像。
  • 设计了一种新的联邦学习聚合方法,用于在特征分布偏斜和标签分布不平衡场景下,从眼底照片检测近视性黄斑变性以及从OCT扫描分类黄斑疾病。
  • 进行了对抗攻击(标签翻转和干净标签),并将区块链纳入联邦学习以演示模型更新在协作站点间的安全传输。

主要发现

  • 联邦学习模型对近视性黄斑变性检测的AUC为0.868±0.009,对OCT黄斑疾病分类的AUC为0.970±0.012。
  • 在标签翻转攻击中,联邦学习模型AUC为0.861±0.019,与集中式模型(AUC 0.856±0.015)相似,高于其他联邦学习模型(AUC 0.578-0.819)。
  • 在干净标签攻击中,联邦学习模型AUC为0.878±0.006,与集中式模型(AUC 0.878±0.001)相当,优于其他先进联邦学习模型(AUC 0.529至0.838)。
  • 模拟显示,在1个全局轮次中区块链增加的额外时间约为5秒。
  • 将区块链加入联邦学习框架是可行的,对模型开发时间影响最小。

临床怎么用

  • 原文未据此给出推荐。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

PURPOSE: Collaboration provides valuable data for robust artificial intelligence (AI) model development. Federated learning (FL) is a privacy-enhancing technology that allows collaboration while respecting privacy via the development of models without raw data transfer. However state-of-the-art FL models still face challenges in non-independent and identically distributed (non-IID) health care settings and remain susceptible to privacy breaches. We propose an FL framework coupled with blockchain technology to address these challenges. DESIGN: Retrospective, multicohort study. MAIN OUTCOME MEASURES: We evaluated our FL model performance in myopic macular degeneration (MMD) and OCT classification and compared our model against state-of the-art FL and centralized models. METHODS: A total of 27 145 images from Singapore, China, and Taiwan were used to design a novel FL aggregation method for the detection of MMD from fundus photographs and macular disease from OCT scans in feature distribution skew and label distribution imbalance scenarios. We further performed adversarial attacks (label flipping and clean label). As proof of concept, blockchain was incorporated into FL to demonstrate secure transfer of model updates across collaborating sites. RESULTS: Our FL model showed robust performance with an area under the curve (AUC) of 0.868 ± 0.009 for MMD detection and 0.970 ± 0.012 for OCT macular disease classification. In label flipping attack, our FL model had an AUC of 0.861 ± 0.019, similar to the centralized model (AUC 0.856 ± 0.015) and higher than other FL models (AUC 0.578-0.819). In clean label attack, our FL model had an AUC of 0.878 ± 0.006, which was comparable to the centralized model (AUC 0.878 ± 0.001) and superior to other state-of-the-art FL models with an AUC of 0.529 to 0.838. Simulation showed that the additional time with blockchain in 1 global epoch was approximately 5 seconds. The addition of blockchain to the FL framework was feasible with a minimal impact on model development time. CONCLUSIONS: Our proposed FL algorithm overcomes the shortcoming of the traditional FL in non-IID situations and remains robust against adversarial attacks. The integration of blockchain adds further security during the transfer of model updates. Blockchain-enabled FL can be a trusted platform for collaborative health AI research. FINANCIAL DISCLOSURE(S): The author(s) have no proprietary or commercial interest in any materials discussed in this article.

来源:Ophthalmology,查看原文。版权归原作者所有。

引用

  1. Teo ZL, Zhang X, Yang Y, et al. Privacy-Preserving Technology Using Federated Learning and Blockchain in Protecting against Adversarial Attacks for Retinal Imaging. Ophthalmology. 2024. PMID: 39424148. doi:10.1016/j.ophtha.2024.10.017

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。