The introduction of chaos-fractal theory and its application across various fields signals a shift in modern scientific development, characterized by systematic approaches that are often complex and ambiguous. This paradigm encourages the integration of new concepts and methods that bridge Chinese and Western medical ideologies.18 Digital intelligence, viewed as a process of entropy reduction, plays a crucial role in transforming disordered systems into more organized and predictable ones, a key principle in complexity science. The holographic principle and chaos-fractal theory provide both the theoretical foundation and experimental basis for developing a digital detection platform for perioperative TCM, enabling more precise and individualized diagnostic and therapeutic approaches.
The core objective of ERAS is to regulate perioperative homeostasis and maintain environmental balance throughout the perioperative period. Through multi-platform and interdisciplinary collaboration, even minor fluctuations can be amplified, initiating a cascade effect throughout the system—often referred to as the “Chaos Phenomenon”. With the advent of the “Internet+” era and the widespread adoption of 5G technology, perioperative artificial intelligence (AI), mobile healthcare, and intelligent medical devices are becoming increasingly prominent. The integration of individualized precision medicine, dynamic monitoring, and TCM facilitates deeper convergence across the healthcare sector, advancing digital transformation, enhancing information connectivity, and enabling cross-disciplinary technological integration.
Theoretical construction of the perioperative digital TCM diagnosis and treatment platform
The dynamic evolution of perioperative patients’ physiological states exhibits typical characteristics of complexity. Surgical trauma, anesthetic intervention, and individual compensatory responses together form a nonlinear interactive system, whose internal mechanisms are difficult to fully explain using traditional reductionist methods. By integrating the biological holographic phenomenon and chaos-fractal theory as the core theoretical framework, a perioperative TCM diagnosis and treatment model based on complexity science is developed to establish “hole-part” associations. Following the principle that “the part reflects the whole”, localized signs such as tongue characteristics and pulse conditions are considered micro-projections of the body’s overall Qi, blood, Yin, and Yang status. Through digital technologies such as tongue image analysis and pulse wave signal acquisition, a mapping relationship between localized signs and the overall functional state of perioperative patients can be established. The nonlinear dynamic characteristics of the perioperative physiological system can be quantified using indicators such as fractal dimension and sample entropy, revealing critical transition patterns between “chaos” and “order” during postoperative recovery. For example, a decrease in heart rate variability (HRV) fractal dimension during the early postoperative period indicates autonomic nervous system suppression (Yin-Yang imbalance), while an increase in fractal dimension during recovery corresponds to the restoration of Yin-Yang harmony. This theoretical framework transcends the limitations of traditional experience-based TCM diagnosis and treatment methods, providing a quantifiable and predictive scientific foundation for monitoring perioperative states.
Construction of methodology
Digitalization of the four diagnostic techniques in TCM
The four diagnostic methods form the practical foundation of TCM. While exploratory research on the quantification of these techniques has been underway, with some expert consensus emerging, progress has been slow and fragmented. Several factors contribute to this situation, including differing opinions on the value of such research, limitations in current technological capabilities, and the lack of a cohesive research framework and methodology.19,20
A bioinformatics prediction model, combined with visual analytics, knowledge graphs, and data processing software platforms, alongside large clinical diagnostic and treatment datasets, represents a cutting-edge approach to intelligent diagnostic research within the context of TCM’s four diagnostic methods. By integrating TCM’s physical characteristics and concepts of balance and harmony into these models, such platforms enable systematic management of perioperative stress responses, facilitating the monitoring of patient prognosis and accelerating postoperative recovery.21–23 Modern intelligent sensors, based on TCM principles, are capable of monitoring the dynamic, chaotic changes within the human body during the perioperative period. The integration of perioperative intelligent monitoring, assessment, and intervention supports key objectives such as “preventing disease before it occurs”, “preventing the progression of existing diseases”, and “preventing recurrence after recovery”. For example, the development of an intelligent pulse wave-recording wristwatch enables continuous, 24-h collection of pulse data, comparing normal and abnormal pulse patterns to build analytical models. This approach enhances the objectivity, visibility, and practicality of TCM diagnostic and treatment methods while supporting scientific research.24,25 In the era of big data, the application of AI in TCM diagnostics—such as AI-powered facial diagnosis, tongue diagnosis, pulse diagnosis, and medical history collection—has significantly improved diagnostic accuracy. Furthermore, the establishment of specialized disease libraries and the use of wearable devices present substantial potential. The use of AI-based diagnostic tools during the perioperative period enables comprehensive monitoring of the patient’s condition, helping to prevent complications and improve the quality of care.
Visualization of perioperative Yin and yang status
Although TCM practitioners can assess changes in Yin and Yang during the perioperative period using the four diagnostic methods, the complexity of diseases, surgical trauma, anesthesia, patient physiological variations, and limitations in clinical experience often result in uncertainty when interpreting the Yin and Yang status. This ambiguity frequently leads to a lack of evidence-based support for the use of specific Chinese herbal formulas. He Fudong suggests that the material basis of Yin and Yang manifestations in human physiopathology lies in the neuro-endocrine-immune network, which is regulated by the autonomic nervous system. In this framework, the sympathetic nervous system is associated with Yang manifestations, while the parasympathetic nervous system is linked to Yin manifestations.26
HRV provides an objective measure of autonomic function, enabling direct observation of the body’s Yin and Yang fluctuations throughout the day. The sympathetic nervous system, characterized by increased activity, outward movement, and upward orientation, is considered Yang. In contrast, the parasympathetic nervous system, associated with calming, inward, and downward tendencies, is regarded as Yin. This concept provides valuable insights into the Yin-Yang relationship within the postoperative autonomic nervous system. Studies have demonstrated that HRV indices, such as the standard deviation of normal RR intervals and the triangular index, reflect overall autonomic function, while metrics such as the percentage of adjacent normal-to-normal (NN) intervals differing by more than 50 milliseconds and high-frequency components indicate vagal activity.27 In perioperative patients with gastrointestinal cancers, significant fluctuations in HRV have been observed. Early postoperative periods typically show a marked reduction in autonomic function, evidenced by decreases in the high-frequency, standard deviation of normal RR intervals, triangular index, and the percentage of adjacent NN intervals differing by more than 50 milliseconds, reflecting an “imbalance of Yin and Yang”. Conversely, as patients gradually recover during later postoperative stages, improvements in these parameters reflect the “return of Yin and Yang to harmony”.28
In addition to linear methods used to analyze heartbeat rhythms, it is important to recognize that these rhythms also exhibit non-linear characteristics. The concept of “Sympathetic-Vagal Balance” reflects the dynamic regulation of the heart by the autonomic nervous system, which is inherently non-linear. The human body functions as a complex system, and its physiological rhythms display non-linear chaotic features. Given that HRV reflects the functional state of the autonomic nervous system, it too exhibits non-linear properties. The degree of non-linearity in these physiological rhythms may be positively correlated with Yin and negatively correlated with Yang.29 Thus, HRV could serve as a specific signal or “fingerprint” left by life processes. Our research team has previously identified that monitoring HRV during the perioperative period, particularly by analyzing circadian rhythms of the sympathetic and parasympathetic nervous systems, can indirectly reflect changes in the Yin and Yang balance within the body (Fig. 1).30 Utilizing objective indicators and visual representations to evaluate circadian rhythms and the Yin-Yang balance offers a practical approach for quantitative monitoring, individualized diagnosis, and treatment.
Application of AI biomarkers
Biomarkers have long been pivotal in medical research and clinical practice, playing a key role in disease diagnosis, predicting disease progression, and assessing the effectiveness of treatments. With the increasing integration of digital technologies in healthcare, mobile health has seen rapid advancements. Data collected from biomonitoring technologies, such as mobile health applications and wearable devices, can be transformed into indicators that reflect the state of Yin and Yang—referred to as AI biomarkers. The longitudinal data provided by these technologies holds significant potential for advancing our understanding of disease prevention, diagnosis, and treatment. AI biomarkers serve as objective indicators of disease progression or physiological responses to therapeutic interventions, collected through digitally intelligent devices.31,32 While traditional biomarkers have been valuable in clinical practice and research, they often require invasive procedures or are costly to measure. Furthermore, due to the dynamic, complex, and chaotic nature of the human body, traditional biomarkers capture only a limited range of physiological indicators, making it challenging to fully assess perioperative health. In contrast, AI biomarkers are generally non-invasive, modular, and cost-effective. The application of perioperative AI biomarkers allows for the detection of subtle changes in physiological states, which can be amplified through the multi-dimensional integration of organ functions. This approach facilitates continuous monitoring of the body’s Yin and Yang balance, representing a significant advancement in the field of perioperative TCM.
The human body functions as a microcosm of spatiotemporal chaotic phenomena, with its rhythms forming a complex, three-dimensional network of intersecting factors that interact in non-linear and non-accumulative ways. These rhythms manifest through various physiological expressions, such as pulses and behaviors. Rhythmicity is a fundamental characteristic of chaotic phenomena in the human body and often presents in a disordered yet specific manner. The use of smartwatches during the perioperative period allows for non-invasive monitoring of parameters such as blood oxygen levels, heart rate, and sleep patterns.33–35 Other metrics, including blood sugar, blood pressure, and body temperature, are tracked through sensors, providing valuable insights into fluctuations in heart rhythm, blood sugar levels, and blood pressure. These measurements enhance our understanding of the balance of Yin and Yang within the body, as well as the smooth flow of Qi and blood. Non-invasive monitoring of perioperative intestinal motility and circadian rhythms can also be achieved using an artificial intelligence bowel sound system (Fig. 2).36 The TCM ZiwuLiuzhu method can be applied to identify optimal times for specific activities and treatments. For example, early morning, when Qi flows most smoothly, is ideal for activities such as breathing exercises and defecation training. These practices aim to regulate the distribution of Qi, blood, Yin, and Yang, fostering positive Qi and accelerating recovery.37 Digital TCM diagnostic tools, along with smart wearable devices like pulse meters, blood glucose monitors, and health bracelets, can be used during perioperative care and post-discharge follow-up through the mobile Internet of Things. Establishing a specialized TCM database for these applications offers significant value for clinical and scientific research, disease treatment, risk prediction, and talent development (Fig. 3).
Despite the promising theoretical framework provided by the digital TCM perioperative diagnosis and treatment platform, several limitations remain. Firstly, the standardization of TCM’s four diagnostic methods (inspection, auscultation and olfaction, inquiry, and palpation) has yet to be unified, which may affect the model’s generalizability. Secondly, quantitative models based on chaos-fractal theory primarily focus on autonomic nervous system functions, such as HRV nonlinear entropy values, with insufficient integration of endocrine, immune, and other multi-system interactions. This limits the comprehensive reflection of TCM’s “holistic concept”. Additionally, the platform’s real-time dynamic regulation capabilities are constrained by algorithm latency and hardware computing power. Dynamic adjustments to TCM intervention strategies still rely heavily on manual expertise, and a fully automated decision-making loop has not yet been established. Finally, existing clinical validations are mostly centered on gastrointestinal surgeries, with limited studies on other surgical types and long-term outcomes, such as quality of life.
Future research should focus on three key areas to address these limitations: (1) establishing multi-center perioperative TCM specialty databases and developing standardized four-diagnosis collection devices; (2) advancing complexity science models by integrating multi-omics data to construct a cross-scale “gene-metabolism-Qi and blood” network; and (3) developing embedded fractal computing chips and autonomous decision-making engines to achieve real-time bedside diagnosis and dynamic intervention optimization. Ultimately, these efforts aim to transition digital TCM perioperative diagnosis and treatment from theoretical validation to clinical implementation.