Privacy-Preserving Medical Data Sharing Scheme Based on Two-Party Cloud-Assisted PSI
Chengzhe Lai, Hanyue Zhang, Rongxing Lu, Dong Zheng
Abstract
The conflict between data privacy and sharing among healthcare institutions creates data silos, causing wasteful duplication, incomplete information, and potential hindrances to scientific research. In this article, we present a privacy-preserving medical data sharing scheme based on cloud-assisted private set intersection (PSI) and aggregate signature technique. First, we propose a novel authenticated cloud-assisted PSI, named AC-PSI, which can achieve client authentication and randomized processing of private data by using Diffie–Hellman-based oblivious pseudorandom function (DH-OPRF) and vector oblivious linear-function evaluation-based oblivious pseudorandom function (VOLE-OPRF), respectively. Second, based on the AC-PSI and locally verifiable signature (LVS), we design a privacy-preserving and secure medical data sharing scheme, which can provide enhanced security features by enabling access control of computing resources and resist precomputation attacks from external sources. Our approach has been proven through a rigorous analysis of security. Finally, through comparative analysis with the existing schemes, it is demonstrated that the proposed AC-PSI and medical data sharing scheme has low communication and computation overhead while achieving a higher level of privacy preservation and security.