• 国家药监局综合司 国家卫生健康委办公厅
  • 国家药监局综合司 国家卫生健康委办公厅

Randomized Controlled Study on AI Intelligent Feedback System Combined with Simulated Endoscopic Training in Improving Gastroscopy Operation Skills

Corresponding author: Wang Xiaobing, 13936135056@163.com
DOI: 10.12201/bmr.202605.00104
Statement: This article is a preprint and has not been peer-reviewed. It reports new research that has yet to be evaluated and so should not be used to guide clinical practice.
  •  

    Abstract: Objective: To investigate the effect difference between AI intelligent feedback system combined with simulated endoscopic training (group A) and feedback-free simulated training (group B) in improving gastroscopy operation skills. Methods: A single-center parallel-group randomized controlled trial design was adopted. Using block randomization (10 cases per block), 40 gastroenterology resident physicians receiving initial gastroscopy training were randomly divided into group A (n=20) and group B (n=20). Group A received real-time operation feedback from AI intelligent feedback system based on deep learning algorithm combined with traditional simulated endoscopic training, while group B received traditional feedback-free simulated endoscopic training. The training period was 4 weeks, 10 hours per week. The primary outcome indicator was the operation skill score based on the Global Endoscopic Activity Rating Scale (GEARS), and the secondary indicators included operation completion time, number of anatomical structure recognition errors, score of Cognitive Load Questionnaire (CLQ), and clinical practice assessment results after 3 months. Results: Prior to training, there was no significant difference in GEARS scores between the two groups (P>0.05). After training, Group A demonstrated superior skills in insertion techniques, observation ability, tissue exposure control, hemostasis management, emergency response, examination completeness, and overall evaluation scores compared to Group B, with statistically significant differences (P<0.05). Group A also showed shorter operation completion time, fewer anatomical structure recognition errors, and lower Cognitive Load Scale (CLQ) scores than Group B (P<0.05). Notably, Group A achieved higher pass rates in clinical practical assessments and exhibited significantly lower rates of virtual mucosal damage compared to Group B (P<0.05).Conclusion: The AI intelligent feedback system combined with simulated endoscopic training can significantly improve gastroscopy operation skills through real-time precise feedback mechanism, reduce cognitive load during learning, enhance assessment performance, and improve clinical operation safety.Key words: AI intelligent feedback system; simulated endoscopic training; gastroscopy operation; randomized controlled study; skill improvement

    Key words: AI intelligent feedback system; simulated endoscopic training; gastroscopy operation; randomized controlled study; skill improvement

    Submit time: 29 May 2026

    Copyright: The copyright holder for this preprint is the author/funder, who has granted biomedRxiv a license to display the preprint in perpetuity.
  • 图表

  • Effect of the Design of the Mastery Learning Course Based on Simulation in the Teaching of Clinical Nursing Skills. 2026. doi: 10.12201/bmr.202602.00016

    wudongfang, wulin, yuanjing, wangyanan. Exploration of the Application of Microteaching Combined with Quantitative Scoring Method in Skill Training of Infant Hip Joint Ultrasound. 2026. doi: 10.12201/bmr.202603.00083

    LiWeiwei, WangBaogong. A Randomized Controlled Study of Azilsartan Amlodipine Tablets vs. Alisartan Isoproxil Amlodipine Tablets in the Treatment of Essential Hypertension for 8 Weeks:A 4-Week Efficacy Analysis. 2026. doi: 10.12201/bmr.202606.00006

    TANG Wenting, LOU Wen, LIU Yanxi. Assessment of consistency in the risk of bias in medical randomized controlled trials in China. 2025. doi: 10.12201/bmr.202505.00027

    ZHANG Ying-mei, ZHENG Li-xia. Application of ERAS nursing combined with isokinetic muscle strength visual feedback training in knee joint injury. 2025. doi: 10.12201/bmr.202503.00050

    Miao Li, Airun Zhang, Jiahong Li, Qichuang Wang, Xiaoyi Fang. The Application of Scenario Simulation Training of Neonatal Resuscitation in Medical Interns. 2025. doi: 10.12201/bmr.202504.00052

    NIE Weilin, LIN Miaoling, YOU Yanzhu, WANG Yarong, HUANG Hongyan, CHEN Ting. Effects of multicomponent exercise on cognitive and physical functioning in older adults with cognitive frail: a systematic review and Meta-analysis of randomized controlled trials. 2025. doi: 10.12201/bmr.202511.00047

    YE Pei-pei, CAO Yong, OUYANG Li-zhi, LIN Qing-qing, ZHANG Jun-tao, DAI Ying, SUN Xiao-jun, LU Ying. . 2025. doi: 10.12201/bmr.202508.00010

    wangyunjun, guodongna. The effect of probiotics on blood pressure: a systematic review and meta-analysis of randomised controlled trials. 2026. doi: 10.12201/bmr.202606.00014

    Ye Wei. Effect Evaluation of mobile combined lead shields on Reducing Gonadal Radiation Dose to Operators During TIPS Procedures:A Randomized Controlled Trial. 2025. doi: 10.12201/bmr.202509.00036

  • ID Submit time Number Download
    1 2025-12-09

    10.12201/bmr.202605.00104V1

    Download
  • Public  Anonymous  To author only

Get Citation

作者:杜昊澎, Wu Zixiang, Wang Xiaobing. Randomized Controlled Study on AI Intelligent Feedback System Combined with Simulated Endoscopic Training in Improving Gastroscopy Operation Skills. 2026. biomedRxiv.202605.00104

Article Metrics

  • Read: 53
  • Download: 1
  • Comment: 0

Email This Article

User name:
Email:*请输入正确邮箱
Code:*验证码错误