JMIR Publications

ABSTRACT

With the deepening modernization of Traditional Chinese Medicine (TCM) and the widespread adoption of artificial intelligence technologies in healthcare, the structuring, standardization, and computability of TCM knowledge have become central imperatives for driving high-quality disciplinary development. This paper focuses on “TCM symptoms”—the foundational core of the syndrome differentiation and treatment system—and systematically proposes a theoretical framework and implementation pathway for constructing the Traditional Chinese Medicine Symptom Ontology (TCM-SO). Grounded in classical texts such as the Huangdi Neijing and Shanghan Lun, this ontology integrates modern TCM clinical practice data. Employing knowledge engineering ontology modeling methods, it constructs a multi-level, inferable knowledge system encompassing symptom naming, attribute definitions, semantic relationships, syndrome differentiation classifications, and modern medical correlations. It innovatively introduces HPO phenotype mapping and gene target association mechanisms, achieving precise mapping between TCM symptoms and the Human Phenotype Ontology (HPO). Standardized gene target information (NCBI Gene specification) is added to symptom instances, forming a cross-scale association network linking “symptom-HPO phenotype-gene.” Developed using OWL 2 language, the ontology features a clear hierarchical structure and cross-platform interoperability. It effectively supports intelligent diagnostic and therapeutic systems, clinical data mining, and integrated Chinese-Western medicine research, providing a unified semantic foundation for TCM internationalization. By integrating traditional medical knowledge with modern life sciences, this study offers technical support and a theoretical paradigm for the digital transformation of TCM knowledge, demonstrating significant academic value and practical significance.