大语言模型在营养管理的应用研究及展望  

Application research and prospects of large language models in nutritional management

作　　者：晋溶辰[1] 赵嶶 周子涵 刘英 杨卉[2] 黄金[2] JIN Rongchen;ZHAO Wei;ZHOU Zihan;LIU Ying;YANG Hui;HUANG Jin(Hunan University of Chinese Medicine,Changsha 410208,China;The Second Xiangya Hospital of Central South University,Changsha 410011,China)

机构地区：[1]湖南中医药大学护理学院,湖南省长沙市410208 [2]中南大学湘雅二医院,湖南省长沙市410011

出　　处：《护理实践与研究》2025年第3期398-405,共8页Nursing Practice and Research

基　　金：国家自然科学基金项目(72074227),湖南省社会科学成果评审委员会课题(XSP21YBC365);湖南省科学技术厅自然科学基金项目(2024JJ5474);湖南省教育厅科学研究项目(22B0390);湖南中医药大学本科生科研创新基金项目(2023BKS130);湖南中医药大学2024年大学生创新创业训练计划项目(S202410541121)。

摘　　要：科技进步和人们生活水平的提高,健康饮食越来越受到人们的关注。WHO表示,确保健康饮食和良好营养可降低患非传染性疾病的风险。营养管理是关注个体饮食和营养状况,以维持健康和预防疾病的重要领域。然而,目前的营养管理受到个体差异、文化和地域、科学研究和临床实践,以及食品供应和环境等多方面制约,最新证据表明,2021年全世界超过31亿人存在不健康饮食的问题,比2019年增加了1.34亿人。为充分发挥营养管理在疾病和生活中的效益,应用新的技术变得尤为关键。在此背景下,大语言模型作为人工智能领域最前沿的研究方向之一,尤其在医疗保健的应用方面备受关注,它具有强大的语言生成和理解能力,可以生成高质量的自然语言文本,还可以帮助人们更好地理解和解释自然语言文本的含义和语境,从而实现人类与机器之间的交互。当前,随着新一轮科技革命和产业变革的深入演进,基于大语言模型的新一代前沿技术和装备正在不断渗透食品营养领域,并带动全球食品营养行业向个性化方向快速发展。大语言模型不仅能够提供科学、个性化的食谱,还能提供交互式学习与情感支持,推动营养教育的普及。自提出以来,大语言模型已在国外的多个人群中进行验证,并展现出对不同类型人群的饮食的良好干预效果,本文将探讨大语言模型的定义、发展趋势、优势,以及其在不同人群下营养管理的应用。具体而言,将聚焦于肥胖人群、糖尿病人群、过敏人群,以及其他个性化需求的人群,并总结各类人群的应用效果、优缺点,旨在为今后开展更深入的研究和改进提供参考。With advancements in science and technology and improvements in people's living standards,healthy eating is receiving increasing attention.The WHO states that ensuring a healthy diet and good nutrition can reduce the risk of non-communicable diseases.Nutritional management is an important field that focuses on an individual's diet and nutritional status to maintain health and prevent disease.However,current nutritional management is constrained by various factors,including individual differences,culture and geography,scientific research and clinical practice,as well as food supply and the environment.The latest evidence shows that,in 2021,more than 3.1 billion people worldwide had unhealthy eating habits,an increase of 134 million compared to 2019.To fully realize the benefits of nutritional management in both disease and daily life,the application of new technologies becomes particularly critical.In this context,large language models,as one of the most cutting-edge research areas in artificial intelligence,have attracted considerable attention in healthcare applications particularly.They possess powerful language generation and comprehension capabilities,can produce high-quality natural language text,and help people better understand and interpret the meaning and context of such texts,thereby enabling interaction between humans and machines.At present,with the ongoing evolution of a new round of scientific and technological revolution and industrial transformation,a new generation of cutting-edge technologies and equipment based on large language models is steadily penetrating the field of food nutrition,driving the global food and nutrition industry to rapidly develop in the direction of personalization.Large language models can not only provide scientific and personalized recipes but also offer interactive learning and emotional support,thus promoting the widespread adoption of nutrition education.Since its introduction,the large language model has been validated in various populations abroad,demonstrating eff

关 键 词：语言模型 人工智能 营养管理 饮食 肥胖 糖尿病 

分 类 号：R47[医药卫生—护理学]