Boosting multi-demographic federated learning for chest x-ray analysis using general-purpose self-supervised representations

2025-04-11

AI Summary

Federated learning (FL) for chest X-ray analysis across multi-center, age-heterogeneous (adult/pediatric) populations suffers from severe non-IID data distributions and critical scarcity of pediatric annotations, leading to degraded model performance. Method: We propose the first FL framework integrating general-purpose self-supervised visual representations—specifically ViT-based DINO and MAE pretraining—into a multi-demographic FL pipeline, offering a plug-and-play robust initialization strategy. Contribution/Results: Our approach significantly improves classification accuracy on 9,125 pediatric images (P=0.031) and consistently enhances performance across 398,523 adult images from multiple centers (P<0.008). Results demonstrate that generic self-supervised representations universally strengthen generalization in cross-age, cross-institution FL settings. This work establishes a clinically viable, privacy-preserving pathway for collaborative modeling across heterogeneous medical imaging cohorts.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Computer Vision: Multi-modal Vision
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Data transparency and provenance

Abstract

Reliable artificial intelligence (AI) models for medical image analysis often depend on large and diverse labeled datasets. Federated learning (FL) offers a decentralized and privacy-preserving approach to training but struggles in highly non-independent and identically distributed (non-IID) settings, where institutions with more representative data may experience degraded performance. Moreover, existing large-scale FL studies have been limited to adult datasets, neglecting the unique challenges posed by pediatric data, which introduces additional non-IID variability. To address these limitations, we analyzed n=398,523 adult chest radiographs from diverse institutions across multiple countries and n=9,125 pediatric images, leveraging transfer learning from general-purpose self-supervised image representations to classify pneumonia and cases with no abnormality. Using state-of-the-art vision transformers, we found that FL improved performance only for smaller adult datasets (P<0.001) but degraded performance for larger datasets (P<0.064) and pediatric cases (P=0.242). However, equipping FL with self-supervised weights significantly enhanced outcomes across pediatric cases (P=0.031) and most adult datasets (P<0.008), except the largest dataset (P=0.052). These findings underscore the potential of easily deployable general-purpose self-supervised image representations to address non-IID challenges in clinical FL applications and highlight their promise for enhancing patient outcomes and advancing pediatric healthcare, where data scarcity and variability remain persistent obstacles.

Problem

• Addresses performance degradation in federated learning for non-IID medical data
• Improves pediatric chest X-ray analysis using self-supervised representations
• Enhances federated learning outcomes with transfer learning for diverse datasets

Innovation

• Federated learning for decentralized medical image analysis
• Self-supervised representations to handle non-IID data
• Transfer learning enhances pediatric and adult dataset performance