A Blockchain-based Alignment System For Cross-institutional Patient Data

In medical practice,patient consultation information is often fragmented and scattered among multiple medical institutions,and the data systems among medical institutions are isolated from each other,making it difficult to align and complete information,while the acquisition of complete and long-period patient consultation data is important for medical research.The traditional regional medical data sharing system for data integration has a centralized architecture that is prone to security risks,and the system is complex and difficult to build.The existing blockchain-based distributed data sharing system often has too much synchronized data,low operational efficiency,low integration with hospital systems,and risks of privacy and data leakage,making it difficult to be well applied.How to achieve efficient integration of cross-institutional patient information while safeguarding patient data security and privacy is of great significance for data-driven clinical research.To address the need of cross-institutional patient information association,this paper designs and builds a blockchain-based cross-institutional patient data alignment system.The alignment concept in this paper is to align different characteristic quantities of patient data scattered in each hospital with the benchmark of patient identity,and to achieve the integration of information between different hospital databases without exposing the original patient data.The system uses a ciphertext matching method based on hash value groups for patient information alignment,which does not require the establishment of a full data uniqueness index,while reducing the amount of blockchain synchronization data.In addition,the combination of blockchain,attribute-based encryption and asymmetric encryption achieves permission control and security assurance.The system enables secure and efficient association,matching,alignment and aggregation of patient information across medical institutions,and enables researchers to visualize the distribution of patient data in multiple centers.The information association system built in this paper adopts decentralized architecture,which is more secure,lightweight and easy to implement than the traditional regional central node architecture.Compared with existing blockchain clinical data sharing systems,this system has been effectively improved in operational efficiency by optimizing the patient matching mechanism and reducing the amount of synchronized data.Under clinical data testing,the amount of synchronized data between nodes is reduced to about 10% of the original,with an improvement rate of86%-91%,and the system runtime is reduced to 10%-30% of the original,with a 60%-90% increase in runtime efficiency.The method in this paper provides new ideas for cross-hospital patient data association,complete visit record acquisition,and medical data sharing research.