Research Progress on Artificial Intelligence Empowered Digitization of Tongue Diagnosis in Traditional Chinese Medicine

Abstract

Tongue diagnosis, as the core content of inspection in Traditional Chinese Medicine (TCM), is a specific embodiment of the TCM diagnostic thinking of “observing the exterior to infer the interior” and has an irreplaceable value in TCM syndrome differentiation and treatment. However, traditional tongue diagnosis has inherent bottlenecks such as strong subjectivity, difficulty in quantification, and low inheritance efficiency, which restrict the modernization of TCM. With the breakthrough progress of artificial intelligence (AI) technology in fields such as computer vision, deep learning, and multimodal fusion, it has provided core technical support for solving the problems of “subjectivity and fuzziness” in tongue diagnosis and realizing the digitalization, objectivity, and intelligence of tongue image diagnosis. This paper systematically reviews the research progress of AI empowering the digitalization of tongue diagnosis, focuses on elaborating the evolution of core technologies such as tongue image preprocessing, intelligent feature recognition, and multimodal fusion, summarizes its clinical application practices in core scenarios including disease auxiliary diagnosis, health status identification, and dynamic efficacy monitoring, and deeply analyzes the current bottlenecks such as insufficient data standardization, lack of model interpretability, superficial integration of TCM theory and AI technology, and limited clinical promotion. In-depth analysis of these current situations and challenges is of great value for promoting the continuous deepening and application of AI in the digital diagnosis of TCM tongue image, and helps to facilitate the inheritance, innovation and modernization of TCM.