Skip to main content Skip to main navigation

Publikation

Asynchronous federated learning for web-based OCT image analysis

Hasan Md Tusfiqur Alam; Tim Maurer; Abdulrahman Mohamed Selim; Matthias Eiletz; Michael Barz; Daniel Sonntag
In: Journal of Medical Imaging (JMI), Vol. 13, No. 1, Pages 014501-014501, Society of Photo-Optical Instrumentation Engineers, 1/2026.

Zusammenfassung

Purpose: Centralized machine learning often struggles with limited data access and expert involvement. This study investigates decentralized approaches that preserve data privacy while enabling collaborative model training for medical imaging tasks. Approach: We explore asynchronous Federated Learning (FL) using the FedBuff algorithm for classifying Op- tical Coherence Tomography (OCT) retina images. Unlike synchronous algorithms like FedAvg, which require all clients to participate simultaneously, FedBuff supports independent client updates. We compare its performance to both centralized models and FedAvg. Additionally, we develop a browser-based proof-of-concept system using modern web technologies to assess the feasibility and limitations of interactive, collaborative learning in real-world settings. Results: FedBuff performs well in binary OCT classification tasks but shows reduced accuracy in more complex, multi-class scenarios. FedAvg achieves results comparable to centralized training, consistent with previous findings. While FedBuff underperforms compared to FedAvg and centralized models, it still delivers acceptable accuracy in less complex settings. The browser-based prototype demonstrates the potential for accessible, user-driven FL sys- tems but also highlights technical limitations in current web standards, especially regarding local computation and communication efficiency. Conclusions: Asynchronous FL via FedBuff offers a promising, privacy-preserving approach for medical image classification, particularly when synchronous participation is impractical. However, its scalability to complex classi- fication tasks remains limited. Web-based implementations have the potential to broaden access to collaborative AI tools, but limitations of the current technologies need to be further investigated.

Projekte

Weitere Links