Blockchain-based storage and retrieval framework for medical imaging data in OLTP systems

Medical imaging data ranging from X-rays, CT scans, to MRIs has continued to grow exponentially. Healthcare systems that have a model of centralized storage and retrieval have significant challenges because of their dependence on older technologies. For example, healthcare systems utilizing traditional PACS systems and EHR systems, experience challenges with latency, limited interoperability, and security, which weakly function in an OLTP model, where real-time access is necessary. Hence, a blockchain-based framework to enable secure, scalable, high-performance retrieval of medical imaging data has been proposed. The framework proposal is based on Hyperledger Fabric and features a novel empirical block size tuning model that optimizes OLTP throughput while maintaining immutability and traceability of the data. The framework takes a hybrid on-chain/off-chain storage approach where the blockchain ledger is dedicated to storing metadata, while the medical images are stored off the ledger itself which greatly reduced the storage overhead on the blockchain. Experimental evaluation suggests a mid-sized block of approximately 5000 records is optimal for balancing speed and stability in the network and additionally reducing average read latency by almost 50% compared to other configurations. We utilize CouchDB for querying and smart contract access control based on the immutability of data to satisfy HIPAA/GDPR requirements. Ultimately, the proposed architecture serves to demonstrate how blockchain can evolve from an immutable ledger, to an entire infrastructure of data, and therefore realize the benefits of real-time health applications, and the results have the opportunity to serve as the basis for future research and potentially larger or national studies, and also designated for artificial intelligence (AI) integration for performance and reliability in clinical contexts.