Federated Learning for Multi-Center Sepsis Early Prediction with Privacy-Preserving

2026-06-02

AI Summary

This study addresses the challenge of early sepsis prediction in multicenter healthcare settings, where data privacy concerns and decentralized data distributions hinder conventional modeling approaches. To overcome this, the authors propose a horizontal federated learning framework that enables multiple hospitals to collaboratively train a predictive model without sharing raw patient data. Leveraging real-world clinical data from three Chinese tertiary hospitals—strictly curated according to rigorous inclusion and exclusion criteria—the work systematically demonstrates the practicality, robustness, and privacy-preserving capabilities of federated learning for sepsis prediction. Notably, it provides the first empirical evidence that such a framework can effectively resist data reconstruction attacks. Experimental results show that the federated model achieves performance comparable to centrally trained models while fundamentally eliminating the risk of privacy leakage, thereby establishing a secure and viable paradigm for multicenter medical collaboration.

Abstract

Privacy-sensitive and distributed characteristics of multi-center medical data bring severe obstacles to centralized modeling for accurate early prediction of sepsis. Federated learning (FL) has attracted growing attention as a promising framework for collaborative model development, as it allows multiple institutions to jointly train predictive models without directly sharing or centralizing raw data. Nevertheless, its practical performance, robustness, and privacy-preserving benefits remain insufficiently evaluated using real-world clinical datasets. To bridge this gap, this study systematically examines the application of federated learning to multi-center sepsis prediction. The experimental dataset consists of 648 clinically screened samples collected from three tertiary hospitals in China, with rigorous inclusion and exclusion criteria. We establish a centralized training paradigm as the performance baseline, and then implement a horizontal federated learning framework for distributed collaborative modeling. Extensive experimental results demonstrate that the federated learning-based model achieves highly comparable prediction accuracy to the centralized counterpart, while fundamentally avoiding privacy leakage. Further privacy security analysis verifies that malicious attackers cannot reconstruct the original patient data from the transmitted model parameters, indicating strong resistance against data reconstruction attacks. This work not only validates the practicality and security of federated learning in clinical sepsis prediction, but also provides a reliable and feasible solution for privacy-preserving multi-center medical collaboration.

Problem

• Federated Learning
• Sepsis Prediction
• Privacy Preservation
• Multi-Center Medical Data
• Distributed Clinical Data

Innovation

• Federated Learning
• Sepsis Prediction
• Privacy Preservation
• Multi-Center Collaboration
• Data Reconstruction Attack