Secure and Efficient Data Sharing Using Blockchain and Deep Learning for Industrial Healthcare Systems

Secure, reliable, scalable communication of private data is a necessity for IoMT devices used in industrial healthcare. Classical centralized architectures are not able to cope with such demands due to their inadequacy on privacy, data integrity, scalability, and cyber security. In this context, a decentralized industrial healthcare data sharing scheme built on permissioned blockchains is offered to alleviate the above challenges. Using smart contracts in permissioned blockchains, the framework guarantees controlled access, tamper resistance of data storage and trusted information kindles transference. Moreover, Support Vector Machines (SVM) was adopted to use with the LSTM network for data analytics, behavior modeling and enhanced attack detection. In order to guarantees patient privacy, homomorphic encryption is embedded in order to process encrypted healthcare data in the cloud. Experiments demonstrate that the proposed approach can be more accurate, robust, and scalable than other deep learning and machine learning methods, which could provide an intelligent method to learn useful representation of industrial healthcare data for future work.