深度学习辅助诊断技术在中医院校影像教学中的应用探索
蒋蕾 , 罗鹏 , 李冠武 , 鲁丽 , 王真真
中国医学教育技术 ›› 2026, Vol. 40 ›› Issue (3) : 346 -351.
深度学习辅助诊断技术在中医院校影像教学中的应用探索
Exploration of the application of deep learning-assisted diagnostic technology in imaging education of TCM colleges and universities
目的 探讨基于深度学习技术的人工智能辅助诊断(artificial intelligence supported diagnosis,AISD)软件在中医院校医学影像实践教学中的应用价值。针对传统教学模式因二维图像认知难度大、课时有限而导致的学生对复杂影像理解不足与实操能力短板的问题,本研究旨在通过实证研究,为构建智能且高效的医学影像教学模式提供依据。 方法 研究对象为2023年7月至2024年7月在上海中医药大学参加医学影像课程的100名2020级本科见习生,随机分为试验组(n=50)和对照组(n=50)。试验组使用头颈动脉AI辅助诊断系统,通过影像后处理(包括曲面重建、容积重建、最大/最小密度投影)到智能分析(涵盖斑块性质识别、狭窄程度分析、数据测量工具)等功能的全流程学习,实现对头颈动脉影像的多角度、定量化探索,对照组则采用传统多媒体幻灯片教学。两组均完成2课时学习。通过笔试、自我效能感和自我导向学习能力测试,对两组学生的学习效果进行评估,并对数据进行统计学比较。 结果 教学结束后,试验组学生的笔试总成绩(86.50±3.16)分高于对照组(77.38±3.47)分(P<0.001)。在具体项目评分上,试验组头颈动脉识别得分为(52.28±2.11)分,对照组为(46.30±1.88)分(P<0.001);病变动脉狭窄程度评判得分为(16.84±1.17)分,对照组为(15.58±1.80)分(P<0.001);斑块性质判断得分为(17.38±0.92)分,对照组为(15.50±0.95)分(P<0.001)。教学前后,试验组在自我效能感上的提升[(4.38±1.43)分]大于对照组[(1.02±0.65)分](P<0.001);在自我导向学习能力上的提升[(21.8±1.90)分]也高于对照组[(10.5±1.37)分](P<0.001)。 结论 基于深度学习的AISD软件可提升中医院校学生的影像理解能力、阅读技能和三维思维能力,增强自我效能感与自主学习能力,为医学影像教育智能化转型提供了可参考的实践模式。
Objective To exploring the application value of artificial intelligence supported diagnosis (AISD) software based on deep learning technology in the practical teaching of Medical Imaging in traditional Chinese medicine (TCM) colleges and universities. In view of the problems of students’ insufficient understanding of complex images and short practical ability caused by the difficulty of two-dimensional image cognition and limited class hours in the traditional teaching mode, this study aims to provide the basis for the construction of intelligent and efficient Medical Image teaching mode through empirical research. Methods The study involved 100 undergraduate interns of grade 2020 who participated in Medical Imaging course in Shanghai University of Traditional Chinese Medicine from July 2023 to July 2024. They were randomly divided into the experimental group (n=50) and the control group (n=50). The experimental group used the AI-aided head and neck artery diagnosis system to realize the multi-angle and quantitative exploration of the head and neck artery image through the whole process of image post-processing (including surface reconstruction, volume reconstruction, maximum/minimum density projection) to intelligent analysis (covering plaque property identification, stenosis analysis, data measurement tools), while the control group used traditional multimedia slide teaching. Both groups completed 2 hours of study. Through the written test, self-efficacy and self-directed learning ability test, the learning effects of the two groups of students were evaluated, and the data were statistically compared. Results When the teaching ended, the total written test score was significantly higher in the experimental group (86.50±3.16) than in the control group (77.38±3.47), (P<0.001). In terms of specific item scores, the experimental group also outperformed the control group acrossed domains: head and neck artery recognition [(52.28±2.11) vs. (46.30±1.88), P<0.001], evaluation of arterial stenosis severity in the diseased group [(16.84±1.17) vs. (15.58±1.80), P<0.001], and characterization of plaque nature [(17.38±0.92) vs. (15.50±0.95), P<0.001]. Before and after teaching, the experimental group showed a greater improvement in self-efficacy (4.38±1.43) compared with the control group (1.02±0.65) (P<0.001); The improvement in self-directed learning ability (21.8±1.90) was also higher than that of the control group (10.5±1.37) (P<0.001). Conclusion The AISD software based on deep learning can improve the image comprehension ability, reading skills, and three-dimensional thinking ability of students in TCM colleges and universities, enhance their self-efficacy and self-learning ability, and provide a practical reference mode for the intelligent transformation of medical imaging education.
| [1] |
|
| [2] |
|
| [3] |
胡方云, 孙振虎, 周建国. 人工智能技术在医学影像学教学中的应用与评估[J]. 中国卫生产业, 2024, 21(12): 152-155. |
| [4] |
张配配, 冯朝燕. 人工智能技术在医学影像课程实习教学中的应用思考[J]. 智慧健康, 2024, 10(8):1-4. |
| [5] |
李新春, 王鹏, 苏曦, |
| [6] |
韩莎莎, 程留慧, 张卉, |
| [7] |
|
| [8] |
|
| [9] |
|
| [10] |
杨阳, 刁楠, 黄增发, |
| [11] |
梁宇颂. 大学生学业自我效能感与心理健康的相关性研究[J]. 中国临床康复, 2004(24): 4962-4963. |
| [12] |
|
| [13] |
|
| [14] |
陈炜昊. 基于卷积神经网络的医学影像辅助诊断方法研究[D]. 咸阳: 西北农林科技大学, 2024. |
| [15] |
刘子仪. 基于可变形卷积的医学影像辅助诊断算法研究[D]. 成都: 西南交通大学, 2023. |
| [16] |
|
| [17] |
张添辉, 刘舒珊, 曾锦梁, |
| [18] |
岳梅, 张叶江. 人工智能时代医学教学改革方向研究[J]. 中国继续医学教育, 2020, 12(7): 6-9. |
上海中医药大学第二十三期课程建设项目(KECJ2024139)
/
| 〈 |
|
〉 |