以CommonKADS與語意網架構設計新興電子病歷推薦調閱系統

Using CommonKADS and Semantic Web Architecture to Building EMR Recommended System

林國中(Kuo-Chung Lin) ； 葉慶隆(Ching-Long Yeh)

《輔仁醫學期刊》 11卷4期 (2013/12) Pp. 267-284

摘要

近年來政府大力推行電子病歷以期降低醫院行政管理成本、縮小病歷儲存空間，朝向健全國民健康保險政策，避免醫療資源浪費等方向努力；且在2009年推動醫療影像報告、血液檢驗報告、門診用藥紀錄及出院病摘等4項之電子病歷交換標準。其方法為將原本分散儲存於各醫院之病歷資料，建置全國統一的交換中心（EEC）提供跨院區索引集中查詢。但因民眾大都選擇當下信賴的醫院就醫，導致每位病患之病歷項目繁多，醫師在透過健保卡尋找院外電子病歷時，往往出現很多的歷史病歷清單，醫師又得花費許多時間在調閱資料上，才能找到符合自己需求的病歷資料；也因此降低了門診醫師使用EEC平台的意願。本研究旨在設計推薦調閱系統來解決現行醫師在EEC平台上調閱所需病歷不易問題。首先，使用CommonKADS知識工程方法論來找到醫師在問診時的領域知識，並進而設計專屬之問診知識集，作為發展電子病歷交換推薦調閱介面的設計基礎。其次，設計推薦電子病歷交換平台之有效管理系統，來承載大量資料。再來，發展推薦調閱規則之SPARQL語言，針對運用現行存在EEC平台下載之電子病歷XML檔案，擴大應用各醫院間已簽章之電子病歷資料；最後，運用RDF相關技術來承載SPARQL查詢語言，進而提升EEC平台查詢調閱之效率與精確度。

關鍵字

推薦電子病歷 ； 知識工程 ； 語意網技術

並列摘要

The Department of Health (Executive Yuan) promoted four EMR (electronic medical record) exchange standards of the medical imaging report, blood examination, clinical meditation record, and discharge summary in 2009, the method of which was to centralize the medical records scattered and stored in various hospitals onto the EEC, so as to provide the centralized cross-hospital index inquiry. Nevertheless, the public often choose their trustful hospitals at the very moment to take medical treatment, resulting in each patient having manifold records. Seeking out-of-hospital EMR through the National Health Insurance Card, physicians have to spend a lot of time for selecting requirements information from a mass of historical case study before they could find out the exact one to meet their requirement. Thus, it has greatly reduced the clinical physicians' intention to use the EEC platform. The Purpose of this study is to design and recommend a checking system to relieve the predicament that current physicians are unable to access the necessarily medical record easily, and to improve the issues of inquiring efficiency in respect of the XML format file carried on the said platform. This study used CommonKADS knowledge-engineering methodology to locate the domain of the physician in diagnosed enquiry, and then, design an exclusive enquiring science anthology as a basis of accessing interface to develop, exchange, and recommend the EMR. Secondary, design and recommend an effective managing system of the EMR exchanging platform to carry bulk data, and then, develop and recommend the SPARQL language of the accessing rules, aiming at the existing EEC platform to download the XML file of EMR, so as to broaden the inter-hospital application of the signed EMR Finally, use RDF-related technology to carry SPARQL-enquiring language, and further to enhance the enquiring and accessing efficiency and precision of the EEC platform.

並列關鍵字

Recommended EMR System ； Knowledge Engineering ； Semantic Web Technology