Department of Infectious Disease, The First Affiliated Hospital of China Medial University, Shenyang 110000, China.
Department of Curriculum and Instruction, Faculty of Education, The Chinese University of Hong Kong, Hong Kong SAR, China.
Int J Environ Res Public Health. 2022 Jul 18;19(14):8733. doi: 10.3390/ijerph19148733.
Medical students learning to use artificial intelligence for medical practices is likely to enhance medical services. However, studies in this area have been lacking. The present study investigated medical students' perceptions of and behavioral intentions toward learning artificial intelligence (AI) in clinical practice based on the theory of planned behavior (TPB). A sum of 274 Year-5 undergraduates and master's and doctoral postgraduates participated in the online survey. Six constructs were measured, including (1) personal relevance (PR) of medical AI, (2) subjective norm (SN) related to learning medical AI, (3) perceived self-efficacy (PSE) of learning medical AI, (4) basic knowledge (BKn) of medical AI, (5) behavioral intention (BI) toward learning medical AI and (6) actual learning (AL) of medical AI. Confirmatory factor analysis and structural equation modelling were employed to analyze the data. The results showed that the proposed model had a good model fit and the theoretical hypotheses in relation to the TPB were mostly confirmed. Specifically, (a) BI had a significantly strong and positive impact on AL; (b) BI was significantly predicted by PR, SN and PSE, whilst BKn did not have a direct effect on BI; (c) PR was significantly and positively predicted by SN and PSE, but BKn failed to predict PR; (d) both SN and BKn had significant and positive impact on PSE, and BKn had a significantly positive effect on SN. Discussion was conducted regarding the proposed model, and new insights were provided for researchers and practitioners in medical education.
医学生学习使用人工智能进行医疗实践可能会提高医疗服务水平。然而，该领域的研究一直缺乏。本研究基于计划行为理论（TPB）调查了医学生对学习临床实践人工智能（AI）的看法和行为意向。共有 274 名五年级本科生、硕士和博士研究生参与了在线调查。共测量了六个构念，包括（1）医疗 AI 的个人相关性（PR），（2）学习医疗 AI 的主观规范（SN），（3）学习医疗 AI 的感知自我效能（PSE），（4）医疗 AI 的基础知识（BKn），（5）学习医疗 AI 的行为意向（BI）和（6）医疗 AI 的实际学习（AL）。采用验证性因子分析和结构方程模型对数据进行分析。结果表明，所提出的模型具有良好的模型拟合度，与 TPB 相关的理论假设大多得到了验证。具体来说，（a）BI 对 AL 有显著的正向影响；（b）BI 由 PR、SN 和 PSE 显著预测，而 BKn 对 BI 没有直接影响；（c）PR 由 SN 和 PSE 显著正向预测，但 BKn 无法预测 PR；（d）SN 和 BKn 对 PSE 均有显著的正向影响，而 BKn 对 SN 有显著的正向影响。对所提出的模型进行了讨论，为医学教育的研究人员和实践者提供了新的见解。
