Research And Implement Of PIX Based On IHE

With the advancement of the construction of Chinese medical industry information,a considerable number of hospitals have set up their own medical information system.These information systems may come from different manufacturers.However,the standard of each manufacturer’s production is different,which results in the heterogeneous medical systems.Nowadays,how to integrate the medical information system in further to solve the problem of "information island" and to achieve cross domain identity matching and medical information sharing have become one of the key issues in the development of China’s medical information industry.Therefore,how to solve the problem about the patient identification and the medical information sharing is the key issue to build the Patient Identifier Cross-referencing(PIX)system.The main research in this paper focuses on the following aspects:1.In this paper,the PIX integration specification and the Cross Enterprise Document Sharing(XDS)sharing model are studied,and they are introduced in detail.2.This paper introduces the patient master index scheme,according that it proposes a cross-domain patient identification cross-index system based on the Enterprise Master Patient Index(EMPI).According to the patient’s identity characteristics,the appropriate patient’s identity matching field can be selected,including exact match algorithm and incremental matching algorithm.Then we adopt the ID + PID scheme to determine the unique identification number of the patient.3.Based on the research above,this paper proposes the PIX system based on PIX integration specification and XDS sharing model,and then each module of the system was designed and implemented.The module is consisted of automatic matching module,artificial matching module,PIX query module and user management module.4.When the PIX system design is complete,the system needs to be tested.First of all,we do some functional testing for the PIX query module,and the test results show that the data is consistent.Then the performance of the matching module is tested.After that the throughput of the matching process is tested,and the result of the matching is verified by recall rate and correct rate.The experimental results show that the ability of the system to perform the matching algorithm is stronger,and the recall and the correct rate have been greatly improved after data preprocessing.