FedMP: Tackling Medical Feature Heterogeneity in Federated Learning from a Manifold Perspective

2025-08-06

AI Summary

To address feature heterogeneity and poor model convergence caused by non-IID data in medical image federated learning, this paper proposes FedMP. The method tackles these challenges through three key innovations: (1) modeling inter-client medical feature distribution discrepancies from a manifold perspective—introducing the first random feature manifold completion mechanism to mitigate local manifold collapse; (2) enabling class-prototype-guided cross-client feature alignment to achieve discriminative manifold matching within a semantically consistent subspace; and (3) incorporating privacy-preserving feature aggregation. Extensive experiments on multi-center medical imaging datasets and multi-domain natural image benchmarks demonstrate that FedMP significantly improves model accuracy and convergence speed, while maintaining high communication efficiency and strong privacy guarantees. Overall, FedMP outperforms state-of-the-art federated learning methods across all evaluated metrics.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Security and Privacy: Data transparency and provenance

Abstract

Federated learning (FL) is a decentralized machine learning paradigm in which multiple clients collaboratively train a shared model without sharing their local private data. However, real-world applications of FL frequently encounter challenges arising from the non-identically and independently distributed (non-IID) local datasets across participating clients, which is particularly pronounced in the field of medical imaging, where shifts in image feature distributions significantly hinder the global model's convergence and performance. To address this challenge, we propose FedMP, a novel method designed to enhance FL under non-IID scenarios. FedMP employs stochastic feature manifold completion to enrich the training space of individual client classifiers, and leverages class-prototypes to guide the alignment of feature manifolds across clients within semantically consistent subspaces, facilitating the construction of more distinct decision boundaries. We validate the effectiveness of FedMP on multiple medical imaging datasets, including those with real-world multi-center distributions, as well as on a multi-domain natural image dataset. The experimental results demonstrate that FedMP outperforms existing FL algorithms. Additionally, we analyze the impact of manifold dimensionality, communication efficiency, and privacy implications of feature exposure in our method.

Problem

• Addresses feature distribution shifts in federated medical imaging
• Enhances FL model convergence under non-IID data heterogeneity
• Aligns client feature manifolds to improve decision boundaries

Innovation

• Stochastic feature manifold completion for enriched training
• Class-prototypes guide feature manifold alignment
• Enhances FL in non-IID medical imaging scenarios