Abstract: The experimental course of traditional Chinese medicine (TCM) chemistry plays a crucial role in cultivating students' practical abilities and innovative thinking in TCM-related programs.However,the traditional blended teaching model faces challenges such as fragmented teaching resources and overly simplified content.This study proposes an artificial intelligence-driven teaching reform pathway characterized by a closed-loop interaction among "resource reconstruction—pattern innovation—evaluation upgrading." By integrating a full-process knowledge graph,and leveraging virtual laboratory platforms and intelligent devices,the model enables coordinated online-offline interaction.Multi-dimensional feedback is generated from both online learning behaviors and offline experimental practices,forming a continuous improvement cycle among "evaluation—resources—models." This approach effectively enhances students' practical innovation abilities and overall teaching quality,offering a reference for the digital transformation and quality improvement of experimental teaching in traditional Chinese medicine.