Blockchain Enabled Federated Learning Framework for Privacy-Preserved Electronic Health Record

The healthcare sector has experienced significant transformations due to the expansion of industrial edge computing and the Internet of Things (IoT), resulting in an increased volume of distributed healthcare data. Nonetheless, preserving the privacy and security of healthcare data remains a paramount concern. This paper proposes a blockchain-based approach BlockDP-FL that integrates deep learning (DL) with federated learning (FL) enhanced by differential privacy (DP) to protect electronic health records (EHRs). This study leverages a Bi-directional Long Short-Term Memory (BiLSTM) network to analyze data and categorize users as normal or abnormal. By integrating DP within the FL framework, it ensures secure data management while accurately identifying and eliminating abnormal users, safeguarding critical healthcare information. The proposed framework, developed in Python, delivers improved classification accuracy and performance over existing methods while strengthening data privacy.