人工智能在放射科住院医师规范化培训中的调查与实践  

Investigation and Practice of Artificial Intelligence in Standardized Training for Radiology Residents

作　　者：于洋[1] 宋欣颖[1] 任莹[1] 侯阳[1] 孙洪赞[1] YU Yang;SONG Xinying;REN Ying;HOU Yang;SUN Hongzan(Department of Radiology,Shengjing Hospital of China Medical University,Shenyang Liaoning 110004,China)

机构地区：[1]中国医科大学附属盛京医院放射科,辽宁沈阳110004

出　　处：《中国继续医学教育》2025年第4期150-154,共5页China Continuing Medical Education

基　　金：辽宁省医学教育研究项目(No.2022-N004-28)。

摘　　要：目的调查参与放射科住院医师规范化培训(简称“住培”)的医生对人工智能(artificial intelligence,AI)的认知和态度情况,并探讨在放射科住培中加入AI课程的可行性。方法于2022年12月—2023年1月,对中国医科大学附属盛京医院放射科参加住培的医生79名(含专硕并轨研究生)进行AI相关知识授课,在课程前、后填写调查问卷,以了解住培医生对AI的认知和态度,并进行课后满意度调查。结果42名住培医生自认为精通技术。他们了解AI的主要渠道包括社交媒体(70名)、师生交流(62名)及相关讲座(53名)。72名住培医生听说过AI与深度学习,34名住培医生对AI的基本算法有了解。73名住培医生认为AI将改善放射医学。不同住培年级的人在问题11、14(人工智能将改善放射学,对于人工智能的了解不会影响我从事放射诊断工作)的选择上差异均有统计学意义(P<0.05);是否精通技术在问题11~15的选择上差异均有统计学意义(P<0.05)。课后调查结果显示,68名住培医生认为AI应成为医疗培训的一部分,AI课程的增设有助于他们的临床和科研工作。72名住培医生愿意帮助创造或训练一些人工智能算法。结论本调查结果显示,放射住培医生在目前社会背景下对AI具有一定程度的认知,但在工作、科研方面对AI的使用仍有待提高。经过AI相关知识授课后,大部分人能在自身心理状态、临床及科研能力方面得到综合性提升。未来在放射住培中加入AI课程可能是一项提高住培医生综合素质的有效方法。Objective To explore the perceptions and attitudes of physicians enrolled in standardized radiology residency training towards artificial intelligence(AI),and to consider the feasibility of incorporating AI coursework within the radiology residency curriculum.Methods From December 2022 to January 2023,a total of residents(including dual-track graduate students)participating in the residency at the department of radiology of Shengjing Hospital Affiliated to China Medical University,were provided with education on AI-related knowledge.Survey questionnaires were filled out before and after the course to understand the residents'knowledge and attitudes towards AI,and a post-course satisfaction survey was conducted.Results A total of 42 resident trainees considered themselves proficient in technology.Their primary channels for learning about AI included social media(70 residents),communication with faculty and peers(62 residents),and related lectures(53 residents).A total of 72 resident trainees had heard of AI and deep learning,while 34 demonstrated understanding of basic AI algorithms.A total of 73 resident trainees believed AI would improve radiological medicine.There were statistically significant differences in the choice of questions 11 and 14(artificial intelligence will improve radiology,and the understanding of artificial intelligence will not affect my work in radiology diagnosis)among residents of different residential training grades(P<0.05).There was statistical significance in the choice of questions 11 to 15(P<0.05).Post-course survey results indicated that 68 residents believed AI should be incorporated into medical training,and that AI curriculum development would benefit their clinical practice and research work.A total of 72 resident expressed willingness to assist in creating or training AI algorithms.Conclusion The survey results indicate that under the current societal context,radiology residents have a certain degree of awareness of AI,but the use of AI in their professional practice and researc

关 键 词：放射科 住院医师规范化培训 人工智能 认知度 医学影像学 教学改革 效果评价 问卷调查 

分 类 号：G642[文化科学—高等教育学]