从数量到价值:利用人工智能与刻意练习促进放射学精准学习
本文综述人工智能(AI)与刻意练习理论在放射学培训中的相互作用,指出AI可能减少初级判读机会、助长自动化偏差和对黑箱算法的过度依赖,但也可作为精准学习的教育辅助工具,实时识别诊断错误并增强反馈。作者提出将AI融入住院医师培训的循证建议,强调AI素养、保留正常影像暴露、维持初级判读技能和结构化反馈。
要点
- AI整合可能减少初级判读机会、助长自动化偏差和过度依赖黑箱算法,削弱关键学习过程。
- AI可通过精准学习识别诊断错误并增强反馈,与刻意练习原则一致。
- 作者建议住院医师培训应纳入AI素养、保留正常影像暴露、维持初级判读技能和结构化反馈。
核心结论
- AI必须被有意纳入放射学培训以保护和增强诊断专长的培养,审慎应用AI技术有望优化刻意练习、加速技能获取,并使未来放射科医生不仅与AI协作,还能以更优的临床判断和专长超越其局限。
研究设计与样本
- 综述,批判性审视AI、刻意练习理论与放射学培训之间的相互作用,未报告检索范围与纳入文献数量。
临床应用与阅片要点
- 作者提出将AI整合入住院医师培训的循证建议,强调需要AI素养、保留对正常影像表现的暴露、维持初级判读技能和结构化反馈。
局限性
- 原文未报告局限性
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
The increasing integration of artificial intelligence (AI) into radiology practice presents both opportunities and challenges for the education of future radiologists. This review critically examines the interplay between AI, the theory of deliberate practice, and radiology training. Deliberate practice, defined by focused, goal-directed activities with immediate feedback and opportunities for refinement, has been shown to be superior to traditional volume- and experience-based learning models in developing clinical expertise. AI integration risks attenuating essential learning processes by reducing primary interpretation opportunities, fostering automation bias, and promoting over-reliance on "black box" algorithms. However, AI also offers a powerful educational adjunct through precision learning, curating personalised learning experiences based on individual needs. AI can identify diagnostic errors in real time and enhance feedback mechanisms, aligning with deliberate practice principles. We argue that AI must be deliberately incorporated into radiology training to safeguard and enhance the development of diagnostic expertise. We also consider the impact AI will have on the role of the future radiologist and the importance and challenges of acquiring the necessary non-interpretative skills. We propose evidence-based recommendations for the integration of AI into residency programmes, emphasising the need for AI literacy, preservation of exposure to normal imaging findings, maintenance of primary interpretation skills, and structured feedback. We contend that thoughtful application of AI technologies offers the potential to optimise deliberate practice, accelerate skill acquisition, and ensure that future radiologists are equipped not only to work alongside AI but to surpass its limitations with superior clinical judgment and expertise.
引用
- Kelly BS, Duignan S, Booth CC, et al. From volume to value: leveraging artificial intelligence and deliberate practice to foster precision learning in radiology. Pediatr Radiol. 2025. PMID: 41258969. doi:10.1007/s00247-025-06470-5
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。