Opportunities and challenges of traditional Chinese medicine in the era of digitalization and intelligence

Since its inception in the mid-20th century, artificial intelligence (AI) has evolved from early expert systems to the current intelligent phase, characterized by deep learning and large language models. This technological wave is reshaping the global healthcare industry across the board, with profound transformations occurring in drug development, clinical diagnosis, and medical service models. The digital and intelligent transformation of traditional Chinese medicine (TCM) is not merely a simple integration of technology, but rather marks a fundamental shift in the development model—from a traditional experience-driven approach to a “data-knowledge-intelligence” three-dimensional fusion-driven paradigm. AI empowers the field of TCM with unique advantages, particularly in the dual alignment between TCM’s theoretical system and research methodology. TCM emphasizes holistic, systematic syndrome differentiation, and its diagnostic and therapeutic processes involve high-dimensional, multivariable complex data analysis. This aligns well with AI’s strengths in handling nonlinear, multivariable complex systems. With the widespread adoption of electronic medical record systems, advancements in sensor technology, and breakthroughs in machine learning algorithms, the digital and intelligent transformation of TCM now has a solid technological foundation. Bibliometric analysis reveals that Chinese scholars contribute to 54.35% of the global publications in this field, while research institutions from 24 countries and regions, including the United States, Europe, and South Korea, are actively engaging in multinational collaborations to further advance the intelligent research in TCM.[1] 1. Development status of AI empowerment in the field of TCM 1.1. Intelligent breakthroughs in clinical medicine, education, and research AI technology has brought about a systematic transformation in TCM clinical research. Research case collection systems developed using AI, through automatic data capture, intelligent retrieval, information desensitization, and digital twin technologies, efficiently complete the entire process from data collection, governance, to analysis. These systems not only enable complex endpoint evaluation but also play a critical role in optimizing clinical research design, project registration, and management. They provide comprehensive support for researchers, significantly accelerating clinical trial progress and enhancing automation levels.[2,3] The rapid development of AI technologies such as machine vision and multimodal data fusion is driving the establishment of a new paradigm in the integration of TCM and Western diagnostic practices. This is evident in two key aspects. First, AI technology has addressed the challenge of quantifying the “four examinations” (observation, auscultation and olfaction, inquiry, and pulse diagnosis): sensor technologies capture pulse characteristics; natural language processing analyzes inquiry information; multimodal TCM diagnostic models have begun to realize the intelligent integration of the “comprehensive analysis of four examinations,” laying a technological foundation for the modernization of TCM diagnosis.[4–7] Secondly, in the field of Western medicine, deep learning-based imaging recognition systems have already achieved high proficiency in diagnosing common chronic diseases.[8] Multimodal fusion technology, which integrates imaging, laboratory, and genomic data, significantly enhances precision diagnostics and creates new technological opportunities for the integration of TCM and Western medicine. TCM’s holistic syndrome differentiation approach excels in capturing systemic and dynamic changes, while Western medicine’s focus on localized pathology is adept at pinpointing microscopic features. The complementary strengths of both systems can be leveraged. Constructing a multi-modal diagnostic model that integrates both macro-level syndromes and micro-level biomarkers is a key technological path for harnessing the unique advantages of TCM and promoting its modernization. At the level of clinical treatment decision-making, TCM intelligent decision support systems, trained on vast amounts of data, cover hundreds of disease and syndrome types.[9] These systems have demonstrated their value in predicting critical conditions such as sepsis, contributing to an 18.7% relative reduction in mortality rates.[10] In the realm of TCM clinical education, AI integrates technologies such as natural language processing, knowledge graphs, and virtual simulation to build intelligent teaching platforms that combine literature databases, TCM knowledge graphs, case maps, remote cloud consultation rooms, and 3D virtual diagnostic and therapeutic scenarios. These platforms support diverse teaching methods, including intelligent Q&A, case analysis, and augmented-reality-integrated courses, while incorporating intelligent assessment mechanisms to systematize the cultivation of clinical thinking. For example, the Tianyi large language model, jointly developed by institutions such as Beijing University of Chinese Medicine, adopts a progressive learning strategy of “reading-clinical practice-following mentors,” and can be applied to clinical education, decision support, and drug development.[11] 1.2. Full-chain innovation in the digital and intelligent Chinese medicine industry AI technology is reshaping the research paradigm of TCM through 3 key layers. First, at the target-molecular level, AI significantly enhances the efficiency of drug discovery, molecular screening, and structural optimization, while reducing costs.[12,13] Secondly, in the systematization of efficacy and safety evaluation, AI technologies, based on organoid and digital cell models, improve the safety and accuracy of clinical drug use through “virtual analysis.”[14] Finally, in supporting the modernization of TCM traditional theories, platforms such as the Traditional Chinese Medicine Active Ingredient Prediction Platform (TCM-AIPP) integrate data chains linking Chinese herbal medicines, components, targets, functions, pathways, and pharmacological properties, providing modern scientific explanations for the medicinal properties of Chinese medicines.[15] In the production sector, AI has covered the entire industry chain, including planting, harvesting, quality identification, drug extraction, pharmaceutical preparation, and warehousing logistics, driving a profound intelligent transformation in the Chinese medicine pharmaceutical industry. The mobile intelligent testing platform developed by the China Academy of Chinese Medical Sciences integrates rapid detection of pesticide residues, toxins, pollutants, and quality identification functions, meeting the demand for fast testing in complex scenarios. 2. Triple dilemmas of digital and intelligent transformation in TCM 2.1. Standardization dilemma: data heterogeneity and lack of interoperability The digital and intelligent transformation of TCM faces multiple standardization challenges. Although the group standard for evaluating large TCM models, released in 2024, established a quantitative assessment framework,[16] significant issues remain. At the clinical level, TCM diagnosis heavily relies on practitioners' subjective experience, and lacking unified quantification standards. At the equipment level, intelligent diagnostic devices for the “four examinations” have not seen widespread adoption, and interoperability between different devices is absent. At the data level, the data collection process has not yet realized open sharing, and leading to notable issues in terms of the stability and repeatabilitymulti-source data. The absence of a comprehensive data governance system remains the primary bottleneck restricting the digital and intelligent development of TCM. 2.2. AI ethics and safety risks: liability attribution and algorithm transparency The application of AI technology presents multidimensional risks and challenges, including ambiguous intellectual property rights, incomplete mechanisms for patient privacy protection, algorithmic bias that could lead to discriminatory risks, and difficulties in determining liability. When AI decision-making errors occur, there is currently no clear legal framework for attributing responsibility among developers, platforms, and users. The “black-box” nature of deep learning models makes the decision-making process difficult to explain, and such interpretability is crucial for building trust in the high-risk field of healthcare. 2.3. Talent dilemma: insufficient integration of medicine and engineering Currently, the integration between TCM professionals and AI technology experts remains superficial. What's more, the lack of digital and intelligent research led by TCM academic thought makes it difficult to achieve deep integration. There is a severe shortage of interdisciplinary talent with expertise in both TCM and AI, resulting in a lack of TCM-specific features in AI research at comprehensive universities. On the other hand, AI research in TCM universities is constrained, leading to insufficient equipment and service capabilities and low clinical trust in AI-driven TCM practices. 3. Development path: constructing a responsible digital and intelligent ecosystem for TCM 3.1. Upholding the core of TCM theory and avoiding the blindness of “technology-driven” approaches Interpreting the TCM principles through modern science requires adherence to the core of TCM theory, preservation of its distinctive characteristics, and vigilance against overreliance on technology. Methodological innovation must be firmly rooted in the ontological foundations and cognitive logic of TCM, avoiding the blindness of “technology-driven” approaches. The theoretical system of TCM encompasses original concepts such as visceral manifestation, pathogenesis, syndrome differentiation and treatment, meridians, collaterals, and acupoints, holism, medicinal properties, therapeutic principles and methods, and the treatment of both root and manifestations. When applying AI to enhance the interpretability in TCM, particular attention should be paid to the scientific connotations of fundamental TCM pathogenesis, cross-organ regulatory mechanisms related to the 5 Zang organs, modern scientific characterization of the 4 properties and 5 flavors of Chinese medicine, the scientific basis of the 7 emotions and synergistic compatibility of Chinese medicine combinations, and the scientific interpretation and regulation of Qi-blood dynamics. On this basis, there is an urgent need to establish a validation system with distinctive TCM characteristics, to move beyond traditional verification frameworks, and to construct a 3-dimensional evaluation system encompassing clinical effectiveness, technological advancement, and cultural inheritance. 3.2. Building high-quality professional datasets and advancing data governance system construction Adhering to principles of findability, accessibility, interoperability, and reusability, a data management system for TCM that meets international standards should be established. A unified TCM data standard and governance framework will better facilitate the close connection of different research outcomes. For example, establishing standardized pulse manifestation and waveform identification criteria for pulse diagnosis, and developing uniform standards for tongue diagnosis interpretation, will help create high-quality TCM diagnostic and treatment data. These standardized datasets can not only be used to train intelligent TCM models but also assist researchers in validating model effectiveness, thus reducing redundancy in research efforts. Under the premise of ensuring data sovereignty and patient privacy, advanced technological methods should be used to enable multicenter collaborative training of models without the need to share raw data. This provides an effective technical solution for breaking data silos. In an open, collaborative data ecosystem, promoting data sharing and innovation in the TCM field becomes a reality. 3.3. Constructing a “dual-helix” talent development system The digital and intelligent transformation of TCM spans multiple interdisciplinary fields, including chemistry, computer science, biology, science of Chinese materia medica, TCM science, pharmacy, automation, mathematics, clinical medicine, pharmacology, and software engineering. This requires the creation of a “dual-helix” training model that features “theoretical integration–technological convergence–practical application.” For instance, reforms to the curriculum system should integrate foundational TCM theories, machine learning algorithms, medical data mining, natural language processing, and other related content, designing AI courses with TCM-specific characteristics. In terms of continuing education, AI and data science training should be provided to practicing TCM professionals. Various formats, such as short-term training courses, online classes, and academic seminars, should be offered to help TCM practitioners acquire basic AI knowledge and tools, thereby enhancing their digital diagnosis and treatment capabilities. 3.4. Improving policies and regulations to build a responsible AI ecosystem It is essential to accelerate the development of specialized regulations for AI in healthcare, clearly defining responsibility boundaries and improving ethical review and regulatory mechanisms. By drawing on the regulatory frameworks from other countries and industries in the fields of healthcare and AI, a system should be established for the approval and clinical validation of intelligent diagnostic and treatment devices that are tailored to the unique characteristics of TCM. This will ensure the safe, effective, and equitable application of AI technologies in TCM practice. 4. Conclusion The digital and intelligent transformation of TCM is a profound paradigm revolution. It is not merely an update of technological tools, but a systemic transformation in thinking, research paradigms, and service models. This transformation will become a key engine driving the modernization of TCM and enhancing its global influence. In light of this, we extend a sincere invitation to colleagues in the global TCM community: let us unite in this grand endeavor, live up to the times, fulfill our mission, and contribute with wholehearted dedication to the modernization of TCM. Statement of ethics None. Conflict of interest statement Luqi Huang is Editor-in-Chief of Science of Traditional Chinese Medicine. The article was subject to the journal’s standard procedures. Funding source This work was supported by the National Key Research and Development Program of China – Young Scientist Project (Grant No. 2025YFC3510900). Data availability statement Data sharing not applicable to this article as no datasets were generated or analyzed during the current study. Author contributions Luqi Huang was responsible for the conception and writing of this paper. The author read and approved the final version of the manuscript.