Boosting multi-demographic federated learning for chest radiograph analysis using general-purpose self-supervised representations
This addresses data scarcity and variability in pediatric healthcare using FL, but it is incremental as it builds on existing self-supervised and FL methods.
The study tackled the problem of federated learning (FL) struggling with non-IID data in medical image analysis, particularly for pediatric cases, by using general-purpose self-supervised representations. It found that FL alone degraded performance for larger adult datasets and pediatric cases, but adding self-supervised weights significantly improved outcomes, with P-values as low as 0.031 for pediatric cases.
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.