Huang Di and Huang Di’s Canon of Medicine

Liang, Huang

Artificial intelligence-driven intelligentization of traditional Chinese medicine diagnosis: Applications, challenges, and prospects of multimodal fusion and large language model

In the context of the rapid development of artificial intelligence (AI) technology, the intelligentization of traditional Chinese medicine (TCM) diagnosis has become a key research direction for promoting the modernization and internationalization of TCM. TCM diagnosis is based on information from four diagnostic methods, featuring multiple modalities, high dimensions, strong overallness, and significant reliance on experience. It has long faced challenges, such as insufficient objectification, standardization, and repeatability. AI technology has made significant progress in TCM diagnosis, gradually achieving intelligent perception and fusion modeling of multiple sources of diagnostic information and demonstrating good potential in pattern differentiation, reasoning for pattern differentiation, and decision-making for auxiliary diagnosis. Language models provide a novel technical paradigm for the expression, reasoning, and interaction of elements of TCM diagnosis. This article systematically discusses the theoretical basis, key technologies, and application progress of AI in the intelligentization of TCM diagnosis, focusing on the current development status and challenges of multimodal diagnostic modeling, intelligent expression of patterns, and the development of large model-driven diagnostic systems, to provide references for the application and development of intelligent TCM diagnosis.

Zhu, Zhang, She, Chen

Application of an artificial intelligence-based four-diagnostic-instrument in oncological symptom management

Oncological cytotoxic therapies, radiotherapy, and targeted or immunotherapy inevitably induce debilitating symptoms, such as fatigue, pain, and nausea or vomiting, severely impacting patients’ quality of life and treatment tolerance. Although traditional Chinese medicine (TCM) emphasizes personalized, holistic management through pattern differentiation, traditional practice is subjective and lacks standardization. This article proposes an artificial intelligence (AI)-secured four-diagnostic TCM tool for the management of oncological symptoms. The tool objectively quantifies TCM patterns in real time using digital tongue or face imaging, photoplethysmographic pulse waveforms, and pattern questionnaires, while concurrently assessing symptom severity using the MD Anderson Symptom Inventory (MDASI)-TCM. A pattern-symptom-technique smart matching algorithm then standardizes TCM intervention selection (e.g., acupoint patching, acupuncture), enabling a dynamic assessment-intervention-optimization closed-loop protocol that modernizes the TCM principle of “treating according to changing patterns.” This AI-driven approach shifts TCM from experience-based empiricism to objective data-driven practice, thereby enhancing the precision and standardization of integrative oncology by combining quantified patterns with MDASI-TCM symptom factors. The platform paves the way for the future integration of multi-omics data (imaging, genomics, proteomics, and metabolomics) to build predictive efficacy models and explore TCM patterns as prognostic biomarkers, ultimately providing a practical framework for improving the quality of life and delivering individualized integrative cancer care.

Huang, Wang, He, Mei, Zheng, Li, Tou, Shen, Zheng