Research On Privacy Protection Of Medical Information

A large amount of medical data have been produced everyday,with the rapid development of medical and health services,information technology and the popular use of electronic medical record.Now,we are facing the problem of big medical data.As medical data are widely used to the area of research,collection and application,data leakage happens sometimes,which will lead to the disclosure of personal privacy information.Data leakage will endanger personal privacy,once disclosed it will bring patients tremendous troubles and impacts whether in life or physical.So,how to strengthen privacy protection under background of medical big data has been an urgent problem to be resolved.In the past few years,lots of scholars had studied on privacy protection,giving the methods of K-anonymity,L-diversity,(a,k)– anonymity and so on.In general,these researchers are mostly focusing on how to hid the real information with the validity and authenticity of the data,but they rarely pay attention on the protection of data availability,distortion and generalization number of constraints and other aspects of the original data.(1)After studying some algorithms which will easily lead to data availability reducing or even lost in the process of anonymous generalization anonymous,a new algorithm called(K,L)anonymous was proposed.The algorithm achieves the goal of eventual anonymous protection by setting constraints(or boundary)for identifier attribute.The experiment result shows that the new algorithm can not only protect privacy security,but also improve the availability of information,and greatly reduces the time overhead for generalization information.(2)A medical management system based on JavaWeb is designed,The system implements access control for users by RBAC model,handles personal information for patients by(K,L)-anonymous algorithm based on generalized constraint to get the sharing publish anonymous tables.The system can protect privacy information for patients by the two ways in most degree,so as to reduce the risk of privacy information to be disclosed.At the same time,it also proves the practicability andfeasibility of above-mentioned constraint-based generalization(K,L)-anonymous algorithm.