用顶k梯度稀疏传输，降低医疗影像联邦学习通信开销
CFL-SparseMed: Communication-Efficient Federated Learning for Medical Imaging with Top-k Sparse Updates
- 只传梯度中最重要的前k个值，大幅减少通信量
- 在非独立同分布数据下保持模型准确率，提升诊断效果
- 适合医疗数据隐私要求高、网络带宽受限的场景
安全可靠的医学影像分类对有效治疗患者至关重要，但集中式模型面临数据和隐私问题。联邦学习（FL）可实现隐私保护协作，但在异构、非独立同分布（non-IID）数据下存在通信成本高的挑战，尤其在大规模网络中。我们提出CFL-SparseMed，一种基于顶k稀疏化的联邦学习方法，仅传输梯度中绝对值最大的前k个值，显著降低通信开销。该统一方案有效缓解数据异构性，同时保持模型精度。实验表明，该方法提升了联邦学习效率，保障了隐私，并改善了非IID医疗影像环境下的诊断准确率与患者护理质量。可复现代码已开源于Github。
原文摘要 · Abstract (English)
Secure and reliable medical image classification is crucial for effective patient treatment, but centralized models face challenges due to data and privacy concerns. Federated Learning (FL) enables privacy-preserving collaborations but struggles with heterogeneous, non-IID data and high communication costs, especially in large networks. We propose \textbf{CFL-SparseMed}, an FL approach that uses Top-k Sparsification to reduce communication overhead by transmitting only the top k gradients. This unified solution effectively addresses data heterogeneity while maintaining model accuracy. It enhances FL efficiency, preserves privacy, and improves diagnostic accuracy and patient care in non-IID medical imaging settings. The reproducibility source code is available on \href{https://github.com/Aniket2241/APK_contruct}{Github}.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。