透過您的圖書館登入
IP:18.225.7.243
  • 期刊

機器學習與海量資料在醫學教育之應用

The Application of Machine Learning and Big Data in Medical Education

摘要


隨科技之快速進展,醫療行為變得更困難而具風險,加上台灣人口老化、照護人力不足等困境,使醫學教育、醫療品質與病人安全面對大挑戰,以資訊科技裝備醫療教育成為一個極重要的解決策略。人工智慧與海量資料的分析與應用是近年來國際發展的重要趨勢,更需應用在醫療照護領域。智慧科技與海量資料在醫療教育與學研界之應用尚有待加強。以下就幾個熱點議題提出資訊工程面結合醫療教育應用面的說明:(I)臨床診斷輔助系統: 應用機器學習或深度學習的工具,加上人類疾病臨床資料,使用醫療專家診斷疾病的邏輯和規則,而找出最適當的診斷範圍,幫助初學者聚焦合理範圍診斷,提升學習成效,甚至藉早期診斷而改善疾病預後。(II)危重症即時警示系統:以深度遞迴神經網路智慧(RNN, Recurrent Neural Network)協助危重症即時警示:對病情變化快速的急重症患者,顯示警訊並自動給予最可能的診斷,甚至建議下一步處置。(III)影像辨識之的疾病診斷工具:建立病人影像資料庫,收集先天性異常疾病患者各種臨床資料,進行臉部(特別是頭臉眼耳鼻)及身體軀幹特徵定義與標準測量,專家標記,實現深度卷積神經網路(又稱卷基深度學習架構,CNN, Convolutional Neural Networks)使用深度網路模型,以協助先天疾病診斷。

並列摘要


With the rapid advances of digital technology, healthcare is becoming difficult and risky. When added by the factors of aging and healthcare manpower shortage, the quality of medical education and healthcare become challenging. The application of Artificial intelligence (AI) and big data is considered very important in many fields of the future world, including the HealthCare domain. However, there is needs for more efforts to facilitate the application in medicine and medical education. The paper highlights several hot topics in this domain: (I) Clinical diagnostic support system: With the uses of machine learning or deep learning, plus various forms of clinical data and expert system, the diagnostic support system is to find out a quality diagnoses list in order to improve the diagnostic accuracy and learning effectiveness. (II) Using RNN (recurrent neural network) to provide warming signals on critical emergent events in medical care, followed by suggesting possible diagnoses and appropriate treatment. (III) Diagnosis system by visual recognition: To help the diagnostic accuracy on the patients with congenital malformation disorders, the system uses CNN (Convolutional Neural Networks), coupling with patients' image data of face and body features, to generate disease diagnoses.

延伸閱讀