医疗大模型驱动的精准筛查与虚拟助手减重管理研究

Large-scale model-driven medical precision screening and virtual assistant for weight management

摘要

目的 旨在利用人工智能(artificial intelligence,AI)技术构建的专属医疗大模型,分析其在实际减重管理临床工作中的应用价值。方法 首先采用武汉大学中南医院院内标记的特征数据及院外优质数据,进行专属减重大模型的训练和优化,最终形成 AI减重系统。然后对患者使用 AI减重系统3个月后健康效果与目标达成、治疗依从性等结果数据进行分析。结果 系统试运行3个月,累计识别管理用户1 492人,患者依从性提高了43%,减重效果提高了30%,患者满意度提高了35%,医生工作效率提高了50%。减重过程数据精准度提高了40%,减重治疗临床成本降低了39%。结论 AI医疗大模型能辅助临床医师提高减重治疗工作效率,提升患者的减重效果。

Abstract

Objective To leverage an exclusive medical large-model constructed with artificial intelligence (AI) technology to analyze its value in clinical weight management practices. Methods First, proprietary weight management data from Wuhan University Zhongnan Hospital and high-quality external datasets were utilized to train and optimize a dedicated weight management large model, ultimately forming an AI-powered weight loss system. Subsequently, an analysis was conducted on health outcomes, goal achievement, and treatment compliance among patients who had used the AI weight loss system over a three-month period. Results During the three-month trial operation, the system cumulatively identified and managed 1 492 patients. Patient compliance increased by 43%, weight loss effectiveness improved by 30%, patient satisfaction rose by 35%, and physician efficiency was enhanced by 50%. The accuracy of weight loss process data was improved by 40%, while the clinical costs for weight loss treatment were reduced by 39%. Conclusion The AI medical large model can assist clinicians in improving the efficiency of weight loss treatments and enhance patient weight management outcomes.

关键词

医疗大模型 / 精准筛查 / 人工智能(AI)虚拟助手 / 减重管理

Key words

medical large model / precision screening / artificial intelligence (AI) assistant / weight management

参考文献

[1] 国家卫生健康委疾病预防控制局 . 中国居民营养与慢性病状况报告(2020年)[M].北京: 人民卫生出版社, 2021.

[2] National Health Commission Disease Control and Prevention Bureau. Report on nutrition and chronic disease status of Chinese residents (2020)[M]. Beijing:The People's Health Press Co., Ltd, 2021.

[3] 《中国成人超重和肥胖预防控制指南》修订委员会 . 中国成人超重和肥胖症预防控制指南(2021)[M].北京: 人民卫生出版社, 2021.

[4] Revision Committee of the Guidelines for Prevention and Control of Overweight and Obesity in Chinese Adults. Guidelines for prevention and control of overweight and obesity in Chinese adults (2021 ed)[M]. Beijing:The People's Health Press Co., Ltd, 2021.

[5] 张璇, 张飞, 李铭麟, 等 . 智能机器人在基层慢性病管理中的应用与挑战[J]. 中国全科医学, 2025, 28(1): 7—12, 19.

[6] ZHANG X, ZHANG F, LI M L, et al. Application and challenges of intelligent robots in grassroots chronic disease management[J]. Chin Gen Pract, 2025, 28(1): 7—12, 19.

[7] 刘哲屹, 景慎艳 . 基于多源异构数据融合技术的 AI大模型在慢性病居家管理平台中的应用研究[J]. 计算机应用文摘, 2025, 41(8): 118—119, 122.

[8] LIU Z Y, JING S Y. Application research of AI big model based on multi—source heterogeneous data fusion technology in chronic disease home management platform[J]. J Comput Appl, 2025, 41(8): 118—119, 122.

[9] 杨华, 陈俭 . 突破大模型的天花板: 物理语言在下一代人工智能中的作用[J]. 中国科技纵横, 2024(15): 20-23.

[10] YANG H, CHEN J. Breaking the ceiling of large language models: the role of physical language in next—generation artificial intelligence[J]. China Sci & Technol Panor Mag, 2024(15): 20-23.

[11] 邢倩, 何达 . 医疗大语言模型的评价现状及思考[J]. 中国卫生资源, 2025, 28(1): 65-72.

[12] XING Q, HE D. Evaluation status and reflections on large language models in healthcare[J]. Chin Health Resour, 2025, 28(1): 65-72.

[13] 倪翠霞, 徐迪 . 大语言模型运行机制基础研究———以通义千问大语言模型为例[J]. 科学技术创新, 2025(9): 101-104.

[14] NI C X, XU D. Basic research on the operating mechanism of large language models — using the Tongyi Qianwen large language model as an example[J]. Sci Technol innov, 2025(9): 101-104.

[15] 徐伟恺 . 基于大语言模型的逻辑推理算法研究[D].成都: 电子科技大学, 2025.

[16] XU W K. Research on logical reasoning algorithm based on large language model[D]. Chengdu:University of Electronic Science and Technology of China, 2025.

[17] GBD 2015 Obesity Collaborators, Afshin A, Forouzanfar MH, et al. Health effects of overweight and obesity in 195 countries over 25 years[J]. N Engl J Med, 2017, 377(1): 13-27.

[18] 国家卫生健康委食品安全标准与监测评估司, 中国疾病预防控制中心营养与健康所, 成人肥胖食养指南编写专家组, 等 . 成人肥胖食养指南(2024年版)[J]. 卫生研究, 2024, 53(3): 347-351.

[19] Food Safety Standards and Monitoring Evaluation Department of the National Health Commission, Institute of Nutrition and Health, Chinese Center for Disease Control and Prevention, Expert Group for Writing Dietary Guidelines for Adult Obesity, et al. Dietary guidelines for adult obesity (2024 edition)[J]. J Hyg Res, 2024, 53(3): 347-351.

[20] 何静, 谢润锋, 戴田宇 . 基于 AI群体应用的社区健康咨询研究[J]. 中国卫生事业管理, 2025, 42(1): 26-31, 117.

[21] HE J, XIE R F, DAI T Y. Research on community health counseling based on AI group applications[J]. Chin Health Serv Manag, 2025, 42(1): 26-31, 117.

[22] 李振环, 王爱胜 . 基于 AI赋能的信息系统设计实验教学的探索实践———以《网页设计的 AI生成》一课为例[J]. 中国信息技术教育, 2025(1): 32-34.

[23] LI Z H, WANG A S. Exploration and practice of experimental teaching for AI—empowered information system design: taking the course “AI—generated Web Design” as an example[J]. China Infor Technol Ed, 2025(1): 32-34.

[24] 张蕾, 董丽霞 . 肥胖自我管理研究现状[J]. 智慧健康, 2025, 11(18): 15-18.

[25] ZHANG L, DONG L X. Research status of obesity self—management[J]. Smart Healthc, 2025, 11(18): 15-18.

[26] 王红秀, 赵海, 杨雪云, 等 . 社区卫生服务中心与三甲医院联合管理模式对社区中心性肥胖2型糖尿病患者干预效果研究[J]. 中国全科医学, 2017, 20(27): 3338-3342.

[27] WANG H X, ZHAO H, YANG X Y, et al. Effect of intervention on type 2 diabetic patients with central obesity in the community through joint management mode between the community health service center and the tertiary grade a hospital[J]. Chin Gen Pract, 2017, 20(27): 3338-3342.

[28] 张如意, 周运翱, 彭迎春 . “医生—医疗人工智能—患者”新型医患关系建构路径探析[J]. 中国医学伦理学, 2025, 38(1): 103-108.

[29] ZHANG R Y, ZHOU Y A, PENG Y C. Exploring the construction path of the new doctor—patient relationships of “doctor—medical artificial intelligence—patient”[J]. Chin Med Ethics, 2025, 38(1): 103-108.

基金资助

2024年国家卫生健康委医院管理研究所科研项目(YLXX24AIB006)

2024国家重点研发计划中医药现代化重点专项(2024YFC3505105)