Research And Implementation Of The Privacy Protection Mechanisms On Personal Medical Data

With the development of mobile Internet technology and medical sensor technology,online disease diagnosis services combining big data technology have brought great convenience to people's lives by collecting users' physical conditions through sensors.Depending on the convenience and efficiency of medical sensors and mobile terminals,a growing number of online medical diagnosis services have been applied to many fields such as chronic disease monitoring,telemedicine and exercise guidance.Accessing an online disease diagnosis service through a smart terminal brings convenience to people's lives,however,it also has the potential dangers of leaking users' physical conditions data and the hospitals' diagnosis model parameters.Considering both users' physical conditions data and the hospitals' diagnostic model parameters are highly sensitive,the privacy of sensitive data related to participating entities in the online disease diagnosis system may seriously affect the application and development of online disease diagnosis technology.To protect the privacy of user's physical data and hospital diagnostic model parameters in the online disease diagnosis system mentioned above,the thesis designs a privacy protected hierarchical online medical diagnosis system,then proposes a secure communication protocol to meet corresponding security requirements,and implements the diagnostic module of the online diagnostic system.Finnally,to evaluate the effectiveness and security of the scheme,the test of overall function and performance of the system have been made.The main contribution of the thesis is as follows.Firstly,investigating the current online medical diagnosis workflow and analyzing the requirements of online medical diagnosis system for data privacy protection,the thesis designs a privacy-preserving online hierarchical diagnosis system,which is composed of authoritative agency,hospital and client,and protect the privacy of users' physical conditions data and hospital diagnosis model parameters.Secondly,considering system security requirements,the thesis designs a security communication protocols,including basic communication protocols,authentication protocols,key update protocols and secure transmission protocols,which provides mutual authentication between access entities,security of key updating,and the confidential transmission of interactive data.Thirdly,according to the privacy-preserving requirements of the online diagnosis module,a privacy-preserving classification model based on linear SVM is selected.And then we improved it by combining an online hierarchical diagnosis structure,proposing a privacy-preserving online disease diagnosis.To evaluate the performance,the accuracy,safety and computational complexity of the diagnostic module is analyzed,and the online disease diagnosis request capabilities in the system is realized.Finally,the overall function,time overhead of system access and concurrency performance of the system have been made,whose results proves that the system can provide secure and efficient online medical diagnostic services.