Federated learning in low-resource settings: A chest imaging study in Africa -- Challenges and lessons learned

2025-05-20

AI Summary

In resource-constrained African settings, TB diagnosis from chest X-rays faces critical challenges—including stringent privacy requirements, severe scarcity of annotated data, unstable connectivity, low digital literacy, and absent AI governance frameworks. Method: This study pioneers the first multi-center, real-world deployment of federated learning (FL) across eight sub-Saharan African countries, implementing localized image preprocessing, heterogeneous data co-optimization, and distributed model training—ensuring raw data never leaves participating institutions. Contribution/Results: The resulting FL-trained TB classifier achieves superior generalization over most site-specific models while preserving data privacy. Empirical evaluation confirms FL’s clinical deployability; however, network unreliability, variable institutional engagement, and lack of regulatory infrastructure emerge as key implementation bottlenecks. This work establishes a reproducible methodological paradigm and actionable insights for trustworthy, context-aware medical AI in low-resource environments.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Humans and AI: Human-in-the-loop Machine Learning
• Computer Vision: Bias, Fairness & Privacy

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

This study explores the use of Federated Learning (FL) for tuberculosis (TB) diagnosis using chest X-rays in low-resource settings across Africa. FL allows hospitals to collaboratively train AI models without sharing raw patient data, addressing privacy concerns and data scarcity that hinder traditional centralized models. The research involved hospitals and research centers in eight African countries. Most sites used local datasets, while Ghana and The Gambia used public ones. The study compared locally trained models with a federated model built across all institutions to evaluate FL's real-world feasibility. Despite its promise, implementing FL in sub-Saharan Africa faces challenges such as poor infrastructure, unreliable internet, limited digital literacy, and weak AI regulations. Some institutions were also reluctant to share model updates due to data control concerns. In conclusion, FL shows strong potential for enabling AI-driven healthcare in underserved regions, but broader adoption will require improvements in infrastructure, education, and regulatory support.

Problem

• Federated Learning for TB diagnosis in low-resource African settings
• Addressing data scarcity and privacy in centralized AI models
• Challenges: infrastructure, internet, literacy, regulations hinder FL adoption

Innovation

• Federated Learning enables collaborative AI without data sharing
• Study compares local and federated models for TB diagnosis
• FL faces infrastructure and regulatory challenges in Africa