A comprehensive survey on enhanced privacy techniques for federated learning in healthcare systems
Abstract
Federated learning (FL) is a decentralised approach to machine learning (ML) in which individuals build models utilising their own information without sharing it with a central server. This is particularly beneficial in healthcare systems, where devices like Medical Internet of Things (MIoT) and electronic health records (EHRs) can be educated remotely while maintaining patient privacy. FL enables collaboration and decision-making based on datasets from multiple participants, improving healthcare processes. However, FL faces privacy challenges as the algorithmic changes exchanged between the client and server might potentially reveal the personal information of the client, violating confidentiality laws, especially in sensitive fields such as medical care. To tackle these challenges, researchers have proposed various techniques to strengthen safety and privacy in FL in medical systems. This chapter presents a literature review that examines existing techniques in this area. The review analysed eight primary methods used to enhance privacy in healthcare FL systems. These techniques include Hierarchical Methods, Peer-to-Peer Sharing, Differential Privacy, Homomorphic Encryption, Edge Device Intelligence, Blockchain, Hybrid, and Diverse Methods, and Industry 5.0. The review analysed the advantages, disadvantages, and compromises of each method while also highlighting possible future paths for improving privacy in healthcare federated learning systems.
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.