| 摘要: | 人工智慧(Artificial intelligence, AI)於醫療影像應用透過機器學習工具、大數據及類神經網路在醫學影像分析中提高醫學診斷的效率和效果。儘管醫界強調AI輔助診斷為醫院提供精準醫療的機會，但整體採用率仍然很低。回顧過去相關文獻主要探討AI醫療應用對醫療過程的影響、臨床診斷與治療建議，而無法充分解釋醫師對AI輔助診斷的使用行為。基於個人採用之觀點，本研究開發兩階段採用研究模型解釋醫師對AI輔助診斷的先前行為(前採用階段)和持續使用(後採用階段)。本研究首先探討消費者價值、個人涉入和科技準備度對醫師採用AI輔助診斷行為意圖的影響；其次為探討成功期望、任務價值和社會資本對持續使用意圖及其績效之影響。本研究採實證研究方法，以衛生福利部所公告醫院評鑑及教學醫院評鑑合格名單為樣本?源，第一年以醫師為研究對象進行問卷調查，第二年以有AI輔助診斷使用經驗的醫師進行問卷調查；再以結構方程模式進?資?分析?驗證研究模式變?間之因果關係，藉以評估與驗證此整合模式及影響因素之關係。因此，本研究成果不僅提供醫院管理者選擇適合的導入策略，並可提供醫學影像製造商、軟件開發商和政府機構對AI輔助診斷行銷和管理策略之?考，並增?學術界對於AI科技使用行為衡?之相關研究。
Artificial intelligence (AI) medical applications enable increased efficiency in and effectiveness of medical diagnostics via machine learning tools, big data, and neural networks in medical image analysis. Although some medical practitioners encourage the use of AI-assisted diagnostic tools due to the precise diagnostic opportunities that they offer hospitals, their overall adoption rate remains low. Several prior studies have only focused on AI medical applications that impact the health care process, provide clinical diagnoses, and suggest treatment; therefore, they are insufficient for fully explaining physician AI-assisted diagnosis usage behaviors. Based on the perspective of individual adoption, this study has developed two-stage adoption research models to explain a physician’s prior behavior (pre-adoption stage) and AI-assisted diagnosis continued use (post-adoption stage), respectively. Thus, this study first aimed to explore the consumer value, personal involvement, and technology readiness factors that influence physicians’ intentions to adopt AI-assisted diagnostic tools. Next, this study aimed to explain the expectancy for success, task values, and social capital factors that influence the antecedents of continuous usage intention of AI-assisted diagnosis and performance impact. A sample source was achieved by using the roster of the Taiwan Ministry of Health and Welfare. A series of surveys will be conducted to empirically test the pre- adoption research model with practicing physicians in the first year. Next, a series of surveys will be conducted to empirically test the post-adoption stage research model with physicians with experience using AI-assisted diagnostic tools in the second year. Structural equation modeling was employed to test two research models. The results of this study provide useful insights that will not only help hospital managers choose an appropriate AI-assisted diagnosis implementation strategy but also enable medical imaging manufacturers, software developers, and government agencies to develop and appropriate their own marketing and administrative strategies for the future. Furthermore, this study provides grounds for a model of innovative AI technology adoption, which can serve as the starting point for future research in this relatively unexplored yet potentially fertile area of research. |
