Privacy-Preserving Federated Learning for Multi-Hospital Patient Data Integration
Abstract
The exponential growth of medical data across hospitals presents significant opportunities for predictive healthcare analytics, yet stringent privacy laws and data fragmentation hinder collaborative research. This chapter introduces a Privacy-Preserving Federated Learning (PP-FedAvg) framework that enables multiple hospitals to train machine-learning models collectively without sharing raw patient records. The system architecture integrates secure aggregation, homomorphic encryption, and differential privacy to ensure confidentiality during model updates, while blockchain-based accountability enhances transparency and trust. Experiments conducted on the MIMIC-III and MedMNIST datasets demonstrate that the proposed framework achieves near-centralized accuracy while significantly reducing privacy loss and communication cost. These results confirm that privacy-preserving federated learning can enable secure, regulation-compliant, and scalable collaboration among healthcare institutions.
Read the paper
Related studies
- Active Learning during Federated Learning for Object Detection
- Fairness-aware federated learning
- Decentralized Federated Learning
- Building a Personally Identifiable Information Recognizer in a Privacy Preserved Manner Using Automated Annotation and Federated Learning
- Preservation of Patient Level Privacy: Federated Classification and Calibration Models
- Efficient explainable federated learning
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.