Construction of a Traditional Chinese Medicine Knowledge Base and Intelligent Inference for Precision Diagnosis and Treatment

This study presents a modular framework that integrates artificial intelligence (AI) with Traditional Chinese Medicine (TCM) to enable precision and personalized diagnosis. Addressing the semantic ambiguity and unstructured nature of TCM knowledge, we construct an ontology-driven knowledge graph capturing over 4,000 symptoms, 900 syndromes, and 1,800 herbs from classical texts, clinical guidelines, and electronic records. A four-stage inference engine, encompassing symptom clustering, Bayesian syndrome-disease mapping, rule-based prescription generation, and patient-specific adjustments, delivers explainable, adaptable recommendations. The system is deployed using a microservice architecture with Docker-based SaaS access and real-time API integration. Evaluated on 500 clinical cases and a validation set of 50 expert-reviewed cases, the framework achieved high diagnostic accuracy (Top-1: 79.4%), prescription precision (83.6%), and interpretability (mean expert rating: 4.52/5), outperforming rule-based and black-box baselines. This research advances the formalization of TCM through AI-enhanced reasoning, bridging symbolic knowledge with statistical learning to support scalable, trustworthy decision-making. Future work will extend to multimodal diagnostic integration and clinical deployment in diverse care settings.