Applications of Blockchain and Federated Learning for Secure Medical Imaging

Abstract

Artificial intelligence in medical imaging is becoming more common to improve clinical decision-making and accuracy of the diagnostic findings. Nevertheless, the high-level centralization of medical image storage and processing brings up serious challenges related to the privacy, security of data, regulatory advocacy, and trust. Federated learning and blockchain have become complementary technologies, which are able to resolve those issues. Blockchain is used to offer decentralized data control, immutability, and transparent access control, and federated learning for collaborative training of models without disclosing sensitive imaging data. The chapter gives an in-depth discussion on the use of blockchain and federated learning in secure medical imaging. It discusses their own functions, their combined design, and presents them in the application to clinical diagnostics, collaboration in research, and regulatory compliance. The problem and perspective of future research are also addressed, which can help to see the future of trustful, scaled and ethical medical imaging ecosystem.