Abstract: Under the background of "new medical sciences" education reform, the cultivation of interdisciplinary clinical thinking ability has become a key dimension of the cultivation of outstanding medical talents. The current research on integrated medical teaching focuses on the mechanical combination of course modules, but the core issues how to achieve in-depth interdisciplinary connections through knowledge reconstruction and how to use intelligent technology to promote clinical cognitive transformation are not explored enough. This study, guided by the theory of systems medicine, constructs an integrated teaching model of pathophysiology and biochemistry with the dual-axis linkage of "disease phenotype - molecular mechanism" from three dimensions: knowledge graph construction, case-driven learning, and experimental verification chain design. By establishing a three-level integration model of "molecule - function - system", a dynamic matching mechanism between clinical cases and virtual simulation experiments was developed. Combined with typical teaching cases, this study explains how to use artificial intelligence (AI) to achieve accurate matching between students' cognitive levels and case complexity, then designs a cognitive training path based on the experimental verification chain, and finally forms a multimodal interactive learning system. This study provides a new perspective to reveal the cognitive transformation mechanism of clinical thinking training, and provides a practical model for promoting the reform of medical integration curriculum and improving the systematic medical literacy of medical students.

Key words: interdisciplinary integration, molecular mechanism, disease phenotype, AI-assisted teaching, intelligent medical education