Securing IoMT healthcare systems with federated learning and BigchainDB

Abstract

The Internet of Medical Things (IoMT) is transforming healthcare by allowing the storage of patient data for diagnostics and treatment. However, this technology faces significant challenges, including ensuring data reliability, security, quality, and privacy. This study proposes a new architecture that uses Federated Learning (FL) and BigchainDB to address these issues. By using FL and BigchainDB, only authorized and trustworthy devices can store their data in the blockchain. This prevents unauthorized access to the blockchain and its stored data. We evaluated this architecture on a real-world model.

Our security mechanism successfully detects and blocks >89% of malicious attacks on the blockchain network. This filtering process ensures that only validated transactions are stored in the blockchain. As a result, fewer transactions are sent to the blockchain, and less data is placed in the memory pool. Our approach increases blockchain throughput while lowering latency. By using a multi-level blockchain, we enhance patient privacy by restricting access to personal data. This research contributes to the development of a secure, efficient, and privacy-preserving IoMT system. By leveraging the power of FL and BigchainDB, we can ensure that patient data is secure, reliable, and accessible only to authorized parties, ultimately improving the quality of care and patient outcomes.