可验证的云存储医疗加密数据统计分析方案

Verifiable Statistical Analysis Scheme for Encrypted Medical Data in Cloud Storage

为满足当前云存储医疗数据对敏感性、完整性以及统计分析可用性的需求,提出一种可验证的医疗加密数据统计分析方案。采用同态加密技术实现密文数据聚合并提高医疗数据的机密性,通过同态签名算法确保外包医疗加密数据的完整性。用户上传经过同态加密和签名的医疗数据到云服务器,云服务器在收到医疗数据分析中心的外包数据聚合请求后对密文医疗数据以及签名值进行聚合运算,并将相应结果返回给医疗数据分析中心,医疗数据分析中心验证云服务器外包同态加密数据聚合的完整性。在此基础上,医疗数据分析中心仅需使用私钥解密就能获得所有用户正确的原始医疗数据聚合结果,并据此进行统计分析。实验结果表明,该方案在医疗隐私大数据分析领域相对SPPDA等方案具有效率优势,医疗数据分析中心在验证数据完整性和分析聚合数据时计算开销保持恒定,与用户数量无关。

In order to meet the requirements of medical data in cloud for sensitivity,integrity and statistical analysis applicability,this paper proposes a verifiable statistical analysis scheme for encrypted medical data.The scheme employs the homomorphic encryption technique to achieve medical data confidentiality and encrypted data aggregation.In addition,the homomorphic signature algorithm is used to ensure the integrity of outsourced medical data.The scheme enables users to upload the encrypted medical data and the corresponding signatures to the cloud server for storage.Once receiving a request for outsourced data aggregation from a medical Data Analysis Center(DAC),the cloud server aggregates those encrypted data and the corresponding signatures,and returns the results to DAC.The DAC could verify the integrity of the encrypted data aggregated by the cloud server.By the private key for decryption,DAC could directly obtain correct results of the aggregated original medical data of all users,and further perform statistical analysis.The experimental results show that the calculation cost of DAC in this scheme is constant and independent of the number of users in data integrity verification and aggregated data analysis,and the proposed scheme is more efficient than SPPDA and other schemes in massive private medical data analysis.