FedMLP: Federated Multi-Label Medical Image Classification under Task Heterogeneity

TL;DR

FedMLP introduces a novel federated learning approach for multi-label medical image classification that effectively handles task heterogeneity caused by partial label availability across different clinical institutions.

Contribution

It formulates a realistic label missing setting in federated learning and proposes FedMLP, a two-stage method combining pseudo label tagging and global knowledge learning to address class missing.

Findings

Outperforms state-of-the-art methods on medical datasets

Effectively handles task heterogeneity in federated learning

Improves multi-label classification accuracy

Abstract

Cross-silo federated learning (FL) enables decentralized organizations to collaboratively train models while preserving data privacy and has made significant progress in medical image classification. One common assumption is task homogeneity where each client has access to all classes during training. However, in clinical practice, given a multi-label classification task, constrained by the level of medical knowledge and the prevalence of diseases, each institution may diagnose only partial categories, resulting in task heterogeneity. How to pursue effective multi-label medical image classification under task heterogeneity is under-explored. In this paper, we first formulate such a realistic label missing setting in the multi-label FL domain and propose a two-stage method FedMLP to combat class missing from two aspects: pseudo label tagging and global knowledge learning. The former utilizes a warmed-up model to generate class prototypes and select samples with high confidence to supplement missing labels, while the latter uses a global model as a teacher for consistency regularization to prevent forgetting missing class knowledge. Experiments on two publicly-available medical datasets validate the superiority of FedMLP against the state-of-the-art both federated semi-supervised and noisy label learning approaches under task heterogeneity. Code is available at https://github.com/szbonaldo/FedMLP.