| Neurosurgery Information Center is an electronic medical records system, developed for the neurosurgery departments of Shengli Hospital, Qilu Hospital and the Second Attached Hospital of Shandong University in Shandong province, which implements recording and management of various medical information including medical records and images, clinical reports, doctor's advice, etc., and provides functions of information retrieval and statistics as well. The Neurosurgery Information Center aims at, on the one hand, establishing a flexible and effective electronic medical records system for the neurosurgery departments of the hospitals to improve their working efficiency, and, on the other, paving the way for gathering full neurosurgery records information to establish the Shandong Neurosurgery Clinical Data Repository as well as for future neurosurgery research.The first Neurosurgery Information Center was finished in July 2007 and has been in use since then. This previous system, however, is based on relational database and has unchangeable and reduplicate data structures, so it is difficult for further development and hard to use, barely satisfying requirements from practitioners and the Shandong Neurosurgery Clinical Data Repository. On account of these problems, Jinan Diwei Corporation and the related research group at Shandong University decided to develop a new Neurosurgery Information Center using DB2 9 that is based on the pureXML technology.This thesis begins with problems of the previous Neurosurgery Information Center, and then describes the design and implement of the new Neurosurgery Information Center, according to the software development workflow. First, the system requirements are obtained using the United Modeling Language and the Object-oriented analyzing method, and use case specifications are set as well. Then the global, logical and storage architectures of the system are described. After that, how to design the medical records function and customized templates function with the Helper Design Pattern is presented in detail, and designs of database table structures, XML case templates and their metadata are emphasized. In the end, transforming of XML and HTML documents in the Case Records Module and the implement of information retrieval are carefully described.Using XML and metadata design, the new Neurosurgery Information Center highlights its flexibility, expansibility and accessibility. Firstly, in the previous Neurosurgery Information Center, different types of clinical documents are designed as different modules, which results in dispersive records, intricate interface and mixing of records and structure. The new system realizes division between functions and records, and uses only one module named Case Records Module instead of dozens of modules previously, whereas specific records of the clinical documents formatted as XML documents can be called freely, making a neat interface and improving expansibility of the system. Secondly, a new module named Data Admin, which enables doctors to create case templates at will, is designed to help solve problems related to unreasonable interface and faulty content in the previous system and ease designers' pressures of unfamiliarity of medicine terminology in the process of database design. Besides, the new system is capable of exact query and full text search in the XML documents. However, owing to the change of data storage type, the relational data in the previous system need to be transferred into the new hybrid database that is based on pureXML. This can be solved, nevertheless, and the key steps are changing relational data into XML data and then saving them in a DB2 9 server. |
PDF Full Text Request