Research On Medical Data Sharing And Access Control Mechanism Based On Blockchain

Medical data,including patient records,insurance contracts,drug circulation records and doctor information,are mostly stored independently by medical institutions to ensure safety due to their professionalism,complexity,privacy and high knowledge density.Isolation is not conducive to timely diagnosis of patients’ condition.If the data are shared and integrated in a scientific and standardized way,such as electronic medical records and physical examination reports,patients can enjoy more accurate diagnosis and convenient insurance services,and researchers will overcome difficult diseases,so as to promote the continuous and high-quality development of the whole industry.Therefore,data sharing has always been the research focus in medical field.Traditional medical data sharing schemes have some disadvantages: First,centralized data storage will increase the cost of supervision and bring the problem of data leakage;Second,it’s hard to apply data due to the difficulty of establishing trust relationship among different institutions;Third,it is hard to supervise and control the data sharing across institutions,because it is difficult to trace the source,which leads to the problem of doctor-patient disputes and unable to determine the responsibility;Fourth,complex cryptography algorithms produce a large amount of computing and communication consumption,which is not suitable for resource-constrained medical scenarios;Fifth,it can’t meet the needs of fine-grained access control.Patient medical records and physical examination data are private,which is inconvenient to share between unrelated departments.In view of the above problems,the main research contents and innovations of this thesis are summarized as follows:（1）A master and slave multi-chain structure composed of one master and many slaves is proposed,which can ensure the safe sharing of data across institutions.Decentralized trust environment will be established through blockchain,the existence of hash algorithm,Merkle tree can ensure that the data can’t be tampered with,and the timestamp makes the source of data traceable.Besides,all nodes don’t need to share risk,and the system will not be paralyzed as a whole due to attacks on an individual chain.The resource consumption will be greatly reduced due to the decrease of the number of nodes on a single chain and storing original data outsides.（2）A medical data access control mechanism based on blockchain is proposed,which can realize fine-grained access control.The access control policies formulated by the data provider is recorded on blockchain,which realizes the whole process monitoring of policy creation,update and revocation.In addition,four smart contracts,including policy enforcement contract,policy administration contract,policy decision contract and attribute authority contract,are designed and implemented,which can directly determine the access rights.The decision-making process is open,and the results are credible.（3）A group-based optimized practical byzantine fault tolerance consensus algorithm is proposed.The algorithm establishes a comprehensive reputation evaluation model to evaluate the credibility of nodes.The nodes with high reputation value are selected dynamically to form consensus group,and members in the group are constantly updated to avoid the problem of centralization.The number of nodes participating in consensus is reduced by the emergence of consensus group.The experimental results show that the performance is better than Practical Byzantine Fault Tolerance in terms of transaction delay and throughput,and the scalability is stronger.