人工智能(AI)在医疗健康领域的应用日益深入,逐渐拓展至营养管理,为老年及肿瘤患者营养不良这一全球性挑 战提供新的解决方案。 传统营养管理方法效率低、主观性强且难以个性化,AI 通过机器学习、自然语言处理及多模态数据分 析,实现了高效、精准的营养筛查、评估、干预与监测。 在营养筛查与评估方面,基于 AI 的自动化工具(如面部图像识别模型) 可快速识别高风险人群;多维度数据驱动的预测模型有助于实现更精准的营养状况判定及分级。 在干预环节,借助 AI 技术 探索个体数据(基因组、微生物、代谢组和行为)和影响营养之间的相互关系,从而设计个性化的膳食与营养支持方案。 在监 测与预后方面,AI 借助图像识别、可穿戴设备等技术实时追踪营养状况,动态调整干预策略;机器学习模型还能够基于营养指 标预测并发症、生存率及身体功能变化,辅助临床预后判断。 尽管 AI 应用于营养管理展现出巨大潜力,仍面临数据标准化不 足和伦理隐私等挑战。 未来应聚焦高质量多中心数据集构建、可解释算法开发及临床转化验证,推动 AI 技术在营养管理中 的规范化、规模化应用,最终改善患者生活质量和健康结局。

Artificial intelligence AI is being increasingly applied in the healthcare field gradually extending to nutritional management and offering new solutions to the global challenge of malnutrition among elderly and cancer patients. Traditional nutritional management methods are often inefficient subjective and difficult to personalize. Through machine learning natural language processing and multimodal data analysis AI enables efficient and precise nutritional screening assessment intervention and monitoring. In the area of nutritional screening and assessment AI-based automated tools such as a facial image recognition model can quickly identify high-risk patients. Multidimensional data-driven predictive models contribute to more accurate determination and grading of nutritional status. In the intervention phase AI technology is used to explore the relationships between individual datagenomic microbial metabolomic and behavioral-and nutritional influences thereby designing personalized dietary and nutritional support plans. For monitoring and prognosis AI utilizes technologies such as image recognition and wearable devices to track nutritional status in real time and dynamically adjust intervention strategies. Machine learning models can also predict complications survival rates and changes in physical function based on nutritional indicators assisting in clinical prognosis evaluation. Although AI shows great potential in nutritional management it still faces challenges such as insufficient data standardization and ethical privacy concerns. Future efforts should focus on constructing high-quality multi-center datasets developing interpretable algorithms and validating clinical applications to promote the standardized and scalable use of AI in nutritional management ultimately improving patients' quality of life and health outcomes.