INTRODUCTION:
Child and Adolescent Obesity has become a global public health issue, and artificial intelligence (AI) technology offers innovative solutions for intervention and management in this field. This article systematically reviews the current applications of AI in Child and Adolescent Obesity, with a focus on the working principles of foundation models and the practical implementation of Retrieval-Augmented Generation (RAG) technology in obesity.This study aimed to develop and validate a locally-deployed large-language-model dialogue system for child-and-adolescent obesity management, and to quantify its accuracy and utility in completing—within a single automated consultation—diagnosis, risk assessment, and generation of personalised intervention plans, thereby providing evidence for future scaled deployment. management.

METHODS:
Through experiments comparing the performance of different AI models in obesity diagnosis, assessment, and intervention plan generation, the effectiveness of AI technology in the field of Child and Adolescent Obesity was validated. The results show that AI-based intelligent consultation systems significantly improve the accuracy of obesity diagnosis, while the incorporation of RAG technology further enhances the model's knowledge-updating capability and personalized service capacity. This article also explores future directions for AI in Child and Adolescent Obesity management, including dynamic knowledge updates, multimodal data integration, and the development of multi-scenario collaborative intervention strategies.

RESULTS:
This study aims to provide theoretical support and practical guidance for the deeper application of AI technology in Child and Adolescent Obesity.

CONCLUSIONS:
The results demonstrate that the developed system, integrating a large language model with retrieval-augmented generation, can achieve accurate diagnosis and generate personalised intervention plans for child and adolescent obesity within a single dialogue, significantly improving diagnostic accuracy and service efficiency, and showing feasibility for further clinical deployment and multi-scenario expansion.