- 期刊
基於知識圖譜與光學字元辨識技術之醫療資源整合系統
MEDICAL RESOURCE INTEGRATION SYSTEM BASED ON KNOWLEDGE GRAPH AND OPTICAL CHARACTER RECOGNITION TECHNOLOGY
摘要
近年來，由於新冠疫情的影響以及人口老齡化持續加劇，人們對醫療環境的關注度日漸提高，醫療資源的合理規劃對人們來說至關重要。本研究開發一款結合就診科別推薦、電子健康記錄和緊急呼叫等功能的醫療整合資源系統。本系統以React Native為框架，透過知識圖譜之推理機制推薦使用者適合的就診科別；醫療收據可透過光學字元辨識技術（OCR）和ckip-bert技術來完成圖片上文字之辨識，再呈現於健康紀錄；緊急呼叫鍵提供使用者聯絡緊急聯絡人，同時發送使用者定位給緊急聯絡人；系統後端則利用Neo4j圖形資料庫管理就診及使用者相關資料。系統完成後，本研究透過發放問卷進行使用者滿意度調查，結果顯示使用者對本系統的資訊品質、系統品質和使用者滿意度三個構面皆有滿意的表現。
並列摘要
In recent years, due to the impact of the Covid-19 and the aging of the population, people are increasingly concerned about the health care environment. It is important for people to plan their health care resources. This study develops a medical integrated resource system that can easily integrate the functions of medical recommendation, electronic health record, and emergency call. The app is developed by using the React Native framework. The appropriate medical department will be recommended by knowledge graph. We complete the text recognition on the medical receipt image by OCR and ckip-bert technology, and then present in electronic health record. Users can press the emergency call button to call the emergency contact. At the same time, the app will send user's position to the emergency contact. Neo4j graphical database is used to manage the medical consultation and user-related data in the backend of the system. After the app is developed, we conduct an user satisfaction survey by issuing questionnaires. The survey results show that users are satisfied with the information quality, system quality and user satisfaction of the system.