From Isolation to Collaboration: Federated Class-Heterogeneous Learning for Chest X-Ray Classification

TL;DR

This paper introduces surgical aggregation, a federated learning method that effectively trains a global chest X-ray classification model across distributed datasets with different class sets, enhancing privacy and generalizability.

Contribution

The paper proposes a novel federated learning approach that handles class heterogeneity without requiring shared class labels or fully annotated datasets, improving collaborative medical imaging analysis.

Findings

Outperforms existing federated learning methods in class-heterogeneous settings

Demonstrates robustness in both IID and non-IID data distributions

Achieves better generalization across diverse datasets

Abstract

Federated learning (FL) is a promising paradigm to collaboratively train a global chest x-ray (CXR) classification model using distributed datasets while preserving patient privacy. A significant, yet relatively underexplored, challenge in FL is class-heterogeneity, where clients have different sets of classes. We propose surgical aggregation, a FL method that uses selective aggregation to collaboratively train a global model using distributed, class-heterogeneous datasets. Unlike other methods, our method does not rely on the assumption that clients share the same classes as other clients, know the classes of other clients, or have access to a fully annotated dataset. We evaluate surgical aggregation using class-heterogeneous CXR datasets across IID and non-IID settings. Our results show that our method outperforms current methods and has better generalizability.