Research On Key Technologies Of Medical Data Privacy Protection Based On Query Rewrite

With the continuous development of ‘Medical + Internet’,medical digitization has become the norm,and a large amount of medical data has been generated.These data contain important information about patients and hospitals.Therefore,the privacy and security of medical data should be paid attention to.The existing privacy protection mechanisms have good theoretical value,but they are difficult to apply in practice.The key to the problem is that most privacy protection mechanisms require in-depth transformation of the database,which greatly hinders the operation of existing systems.To ensure that the privacy of the database is protected,but also to minimize the overhead caused by system development has become a problem that needs to be solved at present.This system realizes the privacy protection of medical and health data and protects the privacy of the data returned by the query without changing the database.First,a privacy protection framework based on query rewriting RW-SQL is proposed.Through the analysis and processing of the structured query language,different privacy protection strategies can be dynamically selected,which can protect the data returned by different queries on the database,and has a strong compatibility.Secondly,in view of the privacy budget overflow risk in differential privacy interactive query,a privacy budget allocation method AQPB based on related attribute query is proposed.This method analyzes the correlation between current query and historical query to give a privacy budget allocation overflow risk assessment.The method finally defines the upper bound of privacy budget allocation,reducing the risk of data privacy leakage.The microservice system architecture designed in this paper is low in coupling,the privacy protection module and the database are independent of each other,and the RPC(Remote Procedure Call)technology is used to achieve high concurrent access to the system.After experimental testing,this system can realize the privacy protection of medical data through extremely low privacy protection overhead and a reasonable privacy budget allocation strategy.