Federated Learning in Infectious Diseases: A Scoping Review
Year: 2026
Type: preprint
Abstract: Federated learning (FL) is a decentralized machine learning approach that allows multiple institutions to collaboratively train models without exchanging raw data. This makes FL particularly relevant for infectious disease research and public health surveillance, where data privacy, sovereignty, and cross-border regulatory constraints often limit centralized data sharing. Early studies have applied FL to COVID-19 imaging, viral genomics, ICU risk prediction, and HIV/STI risk assessment, demonstrating its technical feasibility across different data modalities. However, evidence in this area remains dispersed across biomedical, computer science, and engineering domains, and varies substantially in methodological detail and reporting quality. Existing reviews either address FL in healthcare broadly or focus specifically on COVID-19, leaving no comprehensive synthesis of FL applications across the wider landscape of infectious diseases. Moreover, governance, regulatory, and equity considerations—central to real-world deployment of FL—are seldom examined in a systematic manner. Given the global burden of infectious diseases and the increasing need for privacy-preserving, multi-site analytic approaches, a consolidated mapping of current FL applications is needed. This scoping review will synthesize the available literature, identify gaps, and outline opportunities for research and policy development, particularly for settings where data sharing is constrained or resources are limited.
Source: OSF Preprints (OSF Preprints)
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7165433593