目的:因應COVID-19疫情爆發,實習醫學生及實習生無法至醫療機構實習,故本部將LINE與醫學教育結合,建構核醫醫學教育機器人,達到停課不停學目的。方法:運用自然語言處理與知識本體技術,收集所有核醫檢查相關教學教材紙本及電子資料,進行資訊萃取分析後建置醫學教育課程指引準則知識本體,透過機器學習進行自然語言處理。結果:2020年10月至2021年4月進行問卷調查,結果學生對於核醫課程了解平均值落在1.8,透過使用醫學教育機器人滿意度平均值提升至4.21,「對該課程知識有增加」及「課程內容有符合我的需求」平均達到4.31的高分,證明機器人確實解決學生實習課室課程問題。結論:醫學教育機器人以幾近人類口吻與學生們進行互動式的教學,增加學生對實習流程前中後之課程規劃、評核制度、核醫檢查治療與雙向回饋機制的瞭解。
Objective: Due to the outbreak of the Covid-19 pandemic, medical students and interns have been unable to undergo internship at medical institutions. The department has therefore integrated LINE with medical education to build a nuclear medicine medical education chatbot for the purpose of ensuring uninterrupted studies despite suspended classes. Methods: Utilizing natural language processing and ontology technology, we collected all paper-based teaching materials and electronic information on nuclear medicine scans and proceeded with information extraction and analysis to establish an ontology of medical education curriculum guidelines, then used machine learning to engage in natural language processing. Results: In a survey conducted between October 2020 and April 2021, the score for students' understanding of nuclear medicine courses averaged at 1.8. Average satisfaction rate increased to 4.21 following use of the medical education chatbot with "Increased knowledge of the course" and "Course contents fulfill my needs" reaching high scores of 4.31 on average, thus proving the chatbot indeed resolves course issues encountered by students undergoing internship. Conclusions: The medical education chatbot engages in interactive teaching with students using near-human tones, increasing students' understanding of course plans before, during and after the internship procedure, the rules of assessment, nuclear medicine scans and treatments, as well as the bilateral feedback mechanism.