An AI intelligent syndrome differentiation and acupoint positioning auxiliary system for regulating menstruation and promoting pregnancy thirteen needles

Abstract

The present application relates to the technical field of regulating menstruation and promoting pregnancy, and particularly relates to an AI intelligent syndrome differentiation and acupoint positioning auxiliary system for regulating menstruation and promoting pregnancy thirteen needles, which acquires initial syndrome differentiation parameters by collecting pulse conditions, tongue conditions and symptom distribution information, compares the initial syndrome differentiation parameters with standard syndrome type parameters, and generates a syndrome differentiation deviation signal when an abnormality occurs. Based on the signal, the system dynamically tracks a symptom change sequence, calculates a symptom deviation degree, evaluates an influence level, and generates a high-level syndrome differentiation signal. Then, the system fuses a syndrome type conversion coefficient and a constitution compensation value to obtain a comprehensive syndrome differentiation correction value. Finally, the system takes a weighted mean of the symptom deviation degree as a dynamic acupoint distribution coefficient, fuses standard acupoint parameters, and realizes individualized acupoint positioning optimization. The system realizes closed-loop optimization from intelligent syndrome differentiation to dynamic acupoint distribution, improves the accuracy, objectivity and intelligent level of the regulating menstruation and promoting pregnancy needle method, and has good clinical application value.

Description

Technical Field

[0001] This invention relates to the field of menstrual regulation and fertility promotion technology, and in particular to an AI intelligent diagnosis and acupoint positioning auxiliary system for thirteen acupuncture points for menstrual regulation and fertility promotion. Background Technology

[0002] With the deep integration of artificial intelligence technology and traditional Chinese medicine, the intelligent and standardized development of TCM diagnosis and treatment has ushered in new opportunities. The Thirteen Acupuncture Points for Regulating Menstruation and Promoting Fertility, a commonly used acupuncture treatment, is widely applied in the treatment of gynecological diseases such as menstrual disorders, ovulation disorders, and infertility in women. Its efficacy depends on accurate syndrome differentiation and precise acupoint location. Traditional TCM diagnosis and treatment mainly rely on the physician's subjective experience in the four diagnostic methods (inspection, auscultation and olfaction, inquiry, and palpation) for syndrome differentiation and treatment. This approach suffers from problems such as large individual differences, low repeatability, and long learning cycles, especially in the identification of complex syndromes and dynamic monitoring of disease progression, making quantification and standardization difficult. Furthermore, acupoint location often relies on surface landmarks and experience-based selection, which is easily affected by factors such as the patient's body type and posture, leading to fluctuations in treatment accuracy.

[0003] In recent years, although some studies have attempted to introduce technologies such as image recognition and sensor acquisition into TCM intelligent auxiliary diagnosis and treatment systems, shortcomings remain in areas such as dynamic modeling of syndrome differentiation, analysis of symptom evolution trends, and personalized acupoint optimization. Existing systems mostly remain at the static syndrome differentiation level, lacking continuous tracking and quantitative evaluation of dynamic changes in patient symptoms, making it difficult to achieve real-time linkage and optimization of syndrome differentiation and acupoint selection. Furthermore, dedicated intelligent auxiliary systems for acupuncture techniques aimed at regulating menstruation and promoting fertility are still lacking, and an integrated solution encompassing the four diagnostic methods (inspection, auscultation and olfaction, and palpation) and intelligent syndrome differentiation to precise acupoint location has not yet been formed.

[0004] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0005] The main objective of this invention is to provide an AI-powered intelligent syndrome differentiation and acupoint positioning assistance system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility. This system aims to address the technical problem that existing TCM diagnostic and treatment systems lack quantitative analysis of dynamic changes in syndrome differentiation and intelligent linkage of personalized acupoint positioning in acupuncture for regulating menstruation and promoting fertility. This results in strong subjectivity in syndrome differentiation, insufficient accuracy in acupoint selection, and difficulty in achieving precise treatment that combines standardization and individualization.

[0006] To achieve the above objectives, this invention provides an AI-powered intelligent diagnostic and acupoint location assistance system for thirteen acupuncture points used to regulate menstruation and promote fertility. The system includes: The four diagnostic methods data acquisition module is used to collect information from patients using the four diagnostic methods of traditional Chinese medicine, and to obtain diagnostic data including symptoms and signs. The diagnostic data includes pulse characteristics, tongue characteristics and symptom distribution information. Initial diagnostic parameters are obtained by identifying the feature distribution information. The dialectical comparison module is used to compare the initial dialectical parameters with the standard syndrome parameter range based on the initial dialectical parameters. If the initial dialectical parameters exceed the standard syndrome parameter range, a dialectical deviation signal is generated. The symptom tracking module is used to track the dynamic changes of symptoms in the diagnostic data based on the diagnostic deviation signal. Within the tracking period, the symptom change sequence is obtained, the symptom deviation degree is acquired, and it is compared with the deviation degree threshold to generate a diagnostic influence level signal. If the symptom deviation degree is greater than or equal to the deviation threshold, a high-level diagnostic signal is generated. The syndrome differentiation correction module is used to obtain the syndrome type conversion coefficient and the constitution compensation value based on the high-level syndrome differentiation signal, and to calculate the comprehensive syndrome differentiation correction value by superimposing the syndrome type conversion coefficient and the constitution compensation value. The acupoint optimization module is used to obtain dynamic acupoint coefficients based on comprehensive syndrome differentiation correction values. It integrates standard acupoint parameters with dynamic acupoint coefficients to calculate the required acupoint parameters for adjustment, thereby optimizing the positioning of the thirteen acupoints for regulating menstruation and promoting fertility. The dynamic acupoint coefficients are obtained by weighting and summing all symptom deviations and taking the average value.

[0007] Optionally, the initial dialectical parameters are obtained in the following way: The symptom regions in the syndrome differentiation data are divided into several syndrome differentiation sub-regions. The number of symptom features in each syndrome differentiation sub-region is obtained. The number of symptom features in each syndrome differentiation sub-region is weighted, summed, and averaged to obtain the symptom feature density value. The initial diagnostic parameters are obtained by calculating the ratio between the symptom feature density value and the standard syndrome density value.

[0008] Optionally, the method for obtaining the symptom deviation degree is as follows: Obtain the primary symptom offset component and the secondary symptom offset component, and then perform vector synthesis on the primary symptom offset component and the secondary symptom offset component to obtain the symptom deviation degree.

[0009] Optionally, the method for obtaining the main symptom offset component is as follows: The ratio of the symptom change interval value and the symptom characteristic quantification value within the same tracking period is calculated to obtain the symptom synchronization coefficient. Periods with symptom synchronization coefficients that meet preset conditions are selected and marked as valid diagnosis periods. The duration of the effective diagnostic period is obtained, and the ratio of the duration of the effective diagnostic period to the total duration of the tracking period is calculated to obtain the main symptom offset component.

[0010] Optionally, the symptom change interval value and symptom feature quantification value are obtained in the following way: Within the tracking period, the tracking period is divided into several time segments. A time-symptom coordinate system is established, with the horizontal dimension representing the time segment and the vertical dimension representing the symptom feature quantification value corresponding to each time segment. The obtained symptom feature quantification value is substituted into the time-symptom coordinate system to draw the symptom change curve. Based on the symptom change curve, the peak points and valley points of symptoms are obtained. The difference between the coordinates of adjacent peak points and valley points is calculated to obtain the quantitative value of symptom characteristics. Within the tracking period, the tracking period is divided into several time segments, and a time-symptom change coordinate system is established. The horizontal dimension represents the time segment, and the vertical dimension represents the frequency of symptom changes corresponding to each time segment. The obtained symptom change frequency is substituted into the time-symptom change coordinate system to draw the symptom change frequency curve. Based on the symptom change frequency curve, the points of increasing and decreasing frequency of change are obtained. The difference between the coordinates of adjacent points of increasing and decreasing frequency of change is calculated to obtain the symptom change interval value.

[0011] Optionally, the secondary symptom offset component is obtained in the following way: The symptom deviation value of a sub-region is compared with the allowable deviation value of the sub-region. If the symptom deviation value of a sub-region is greater than or equal to the allowable deviation value of the sub-region, the sub-region is marked as a symptom abnormal sub-region. If the symptom deviation value of a sub-region is less than the allowable deviation value of the sub-region, the sub-region is marked as a symptom normal sub-region. The correlation value of the sub-region's vital signs is compared with the correlation threshold of the sub-region. If the correlation value of the sub-region's vital signs is less than the correlation threshold of the sub-region, the sub-region is marked as a sub-region with abnormal vital signs correlation. If the correlation value of the sub-region's vital signs is greater than or equal to the correlation threshold of the sub-region, the sub-region is marked as a sub-region with normal vital signs correlation. Obtain the symptom abnormal sub-regions, perform spatial overlap analysis on the symptom abnormal sub-regions and the sign-related abnormal sub-regions to obtain overlapping abnormal sub-regions, obtain the number of overlapping abnormal sub-regions, sum the number of overlapping abnormal sub-regions to obtain the total number of abnormal regions, calculate the ratio of the total number of abnormal regions to the total number of syndrome differentiation sub-regions to obtain the secondary symptom offset component.

[0012] Optionally, the sub-region symptom deviation value is obtained in the following way: The diagnostic data is divided into regions to obtain diagnostic sub-region data. Based on the diagnostic sub-region data, the actual quantitative value of symptoms in each diagnostic sub-region is extracted and marked as the sub-region actual quantitative value. The difference between the sub-region actual quantitative value and the sub-region standard quantitative value is calculated to obtain the sub-region symptom deviation value. The method for obtaining the sub-regional vital sign correlation value is as follows: the vital sign correlation detection is performed on the syndrome differentiation data, the syndrome differentiation data is divided into several syndrome differentiation sub-regions, the vital sign correlation degree of each syndrome differentiation sub-region is obtained and marked as the sub-region measured correlation degree, and the difference between the sub-region measured correlation degree and the sub-region standard correlation degree is calculated to obtain the sub-regional vital sign correlation value.

[0013] Optionally, the method for obtaining the physical compensation value is as follows: Obtain the quantitative values ​​of symptom characteristics and the interval values ​​of symptom changes within the effective diagnostic period. Calculate the ratio between the quantitative values ​​of symptom characteristics and the interval values ​​of symptom changes to obtain the symptom change rate value. Sum the symptom change rate values ​​within each effective diagnostic period and take the average value to obtain the constitution compensation value.

[0014] Optionally, the method for obtaining the certificate type conversion coefficient is as follows: Based on the symptom synchronization coefficient corresponding to the effective differentiation period, it is marked as the effective synchronization coefficient. The difference between all effective synchronization coefficients in the tracking period and the standard value of the synchronization coefficient is calculated to obtain the synchronization coefficient deviation value. The variance of the synchronization coefficient deviation value is calculated to obtain the syndrome conversion coefficient.

[0015] Optionally, the verification method for the comprehensive dialectical correction value is as follows: Obtain the constitution compensation value and syndrome conversion coefficient within the effective diagnosis period, calculate the ratio between the constitution compensation value and the syndrome conversion coefficient to obtain the correction verification coefficient, and confirm the validity of the comprehensive diagnosis correction value if the correction verification coefficient is within the preset verification range.

[0016] In the AI-powered intelligent syndrome differentiation and acupoint positioning auxiliary system for thirteen acupuncture points for regulating menstruation and promoting pregnancy, the system achieves deep integration and dynamic linkage between TCM syndrome differentiation and acupuncture point selection by constructing a full-link intelligent process that includes four diagnostic methods data collection, syndrome differentiation comparison, dynamic tracking, syndrome differentiation correction, and acupoint optimization. First, the four diagnostic methods acquisition module performs structured recognition of multi-dimensional data such as pulse, tongue appearance, and symptom distribution to generate initial diagnostic parameters, improving the objectivity and completeness of information collection. The diagnostic comparison module, through intelligent comparison with standard syndrome parameter ranges, promptly detects anomalies and triggers diagnostic deviation signals, enhancing the sensitivity and early warning capabilities of diagnosis. The symptom tracking module performs time-series modeling of dynamic symptom changes, quantifies symptom deviation, and achieves precise capture and level assessment of disease fluctuations. The diagnostic correction module introduces syndrome conversion coefficients and constitution compensation values ​​to comprehensively reflect the evolution trend of syndromes and individual constitution differences, significantly improving the personalization and accuracy of diagnostic conclusions. Finally, the acupoint optimization module generates dynamic acupoint coefficients based on comprehensive diagnostic correction values, quantifying the degree of symptom deviation and integrating it into the adjustment of standard acupoint parameters, achieving personalized and dynamic optimization of acupoint positioning. Overall, this system effectively solves the problems of strong subjectivity in diagnosis, discontinuous disease tracking, and lack of quantitative basis for acupoint selection in traditional acupuncture treatment, improving the standardization, precision, and intelligence of the thirteen-needle treatment for regulating menstruation and promoting fertility. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the first embodiment of the AI ​​intelligent diagnosis and acupoint positioning auxiliary system for the thirteen acupuncture points for regulating menstruation and promoting fertility of the present invention; Figure 2 This is a flowchart illustrating the specific steps involved in obtaining the symptom change interval value and symptom characteristic quantification value in the AI ​​intelligent diagnosis and acupoint positioning auxiliary system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility of the present invention. Figure 3 This is a flowchart illustrating the specific steps involved in obtaining the secondary symptom offset component in the AI-powered intelligent diagnosis and acupoint positioning auxiliary system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility, as described in this invention.

[0018] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0020] In one embodiment, such as Figure 1As shown, an AI-powered intelligent diagnosis and acupoint location assistance system for thirteen acupuncture points for regulating menstruation and promoting fertility is provided. The system includes: The four diagnostic methods data acquisition module 10 is used to collect information on patients from the four diagnostic methods of traditional Chinese medicine and obtain diagnostic data containing symptoms and signs. The diagnostic data includes pulse characteristics, tongue characteristics and symptom distribution information. Initial diagnostic parameters are obtained by identifying the feature distribution information. The four diagnostic methods (inspection, auscultation, inquiry, and palpation) acquisition module is a system component used to collect information from patients through these four diagnostic methods. It integrates multi-dimensional data such as pulse, tongue appearance, and symptom distribution, and can be used to generate structured initial diagnostic parameters, improving the objectivity and completeness of information collection. In this embodiment, the module can simultaneously collect information from the four diagnostic methods (inspection, auscultation, inquiry, and palpation) through sensors, image recognition, and structured questionnaires. Diagnostic data can be a structured set of clinical information containing the patient's pulse characteristics, tongue appearance, and symptom distribution, which can be used as the basic input data for diagnostic analysis. Furthermore, diagnostic data can include, but is not limited to, pulse datasets, tongue image sets, and symptom description sets. Pulse characteristics can be quantitative indicators collected by pulse diagnosis equipment reflecting the patient's pulse rhythm, strength, frequency, and other characteristics, which can be used to determine the state of Qi and blood circulation and the function of the internal organs. In an exemplary embodiment, pulse characteristics can include, but are not limited to, floating and sinking pulse characteristics, slow and rapid pulse characteristics, and weak and strong pulse characteristics.

[0021] Tongue features can be visual parameters such as tongue color, coating texture, and shape extracted through image recognition. These can be used to assist in judging the body's internal imbalances, including cold / heat, deficiency / excess, and the balance of body fluids. For example, tongue features may include, but are not limited to, tongue color, coating color, and tongue shape. Symptom distribution information can be a structured description of the patient's complaints and physical signs in terms of time, location, and nature. This can be used to support preliminary syndrome classification and parameter generation. In a specific embodiment, symptom distribution information may include, but is not limited to, menstrual cycle symptom distribution, accompanying symptom distribution, and emotion-related symptom distribution. Initial diagnostic parameters can be numerical vectors generated by the four diagnostic methods acquisition module to characterize the patient's current syndrome state. These can be used as input benchmarks for syndrome comparison. Furthermore, initial diagnostic parameters may include, but are not limited to, parameters for qi deficiency, blood stasis, and liver stagnation.

[0022] Collecting TCM diagnostic information from patients can involve simultaneously acquiring information from the four diagnostic methods: observation, auscultation and olfaction, inquiry, and palpation. Furthermore, this information can be collected using a combination of intelligent pulse diagnosis instrument, tongue image camera, and structured electronic questionnaire, resulting in multi-dimensional diagnostic data input. Obtaining diagnostic data including symptoms and signs can be achieved by integrating the results of the four diagnostic methods to form a structured dataset. In one specific embodiment, this data can be obtained by converting the raw signals into feature vectors and storing them in a database, providing standardized input for subsequent diagnostics. Initial diagnostic parameters are obtained by identifying feature distribution information, which can involve pattern recognition of pulse, tongue, and symptom distributions and mapping them to numerical parameters. For example, obtaining initial diagnostic parameters by identifying feature distribution information can be achieved by extracting tongue features using a convolutional neural network and mapping them to a syndrome space, thus enabling the objective quantification of subjective experience.

[0023] The dialectical comparison module 20 is used to compare the initial dialectical parameters with the standard syndrome parameter range based on the initial dialectical parameters. If the initial dialectical parameters exceed the standard syndrome parameter range, a dialectical deviation signal is generated. The diagnostic comparison module can be a logical processing unit that intelligently compares the initial diagnostic parameters with the standard syndrome parameter range. It can be used to identify diagnostic deviations and trigger warning signals, thereby enhancing diagnostic sensitivity. In this embodiment, the diagnostic comparison module can achieve parameter range matching judgment based on a rule engine or machine learning model. The standard syndrome parameter range can be a set of parameter boundaries corresponding to various gynecological syndromes constructed based on traditional Chinese medicine classics and clinical guidelines, which can be used to provide a reference benchmark for diagnostic comparison. For example, the standard syndrome parameter range may include, but is not limited to, the parameter ranges for kidney deficiency syndrome, phlegm-dampness syndrome, and blood-heat syndrome. The diagnostic deviation signal can be a logical flag triggered when the initial diagnostic parameters exceed the standard syndrome parameter range, which can be used to initiate the symptom tracking process. In an exemplary embodiment, the diagnostic deviation signal may include, but is not limited to, a mild deviation signal, a moderate deviation signal, and a significant deviation signal.

[0024] Comparing the initial diagnostic parameters with the standard diagnostic parameter range can involve checking each parameter to see if it falls within a preset diagnostic range. Furthermore, this comparison can be achieved by calculating the matching degree using an interval membership function, thereby identifying potential diagnostic anomalies. If the initial diagnostic parameters exceed the standard diagnostic parameter range, a diagnostic deviation signal is generated, which can be achieved by triggering a logical flag when any parameter exceeds its limit. In one specific embodiment, generating a diagnostic deviation signal if the initial diagnostic parameters exceed the standard diagnostic parameter range can be achieved by setting a Boolean deviation flag, thereby activating a dynamic tracking mechanism.

[0025] The symptom tracking module 30 is used to track the dynamic changes of symptoms in the diagnostic data based on the diagnostic deviation signal. During the tracking period, the symptom change sequence is obtained, the symptom deviation degree is acquired and compared with the deviation degree threshold to generate a diagnostic influence level signal. If the symptom deviation degree is greater than or equal to the deviation threshold, a high-level diagnostic signal is generated. The symptom tracking module can be an analysis module that performs time-series modeling of dynamic symptom change data and quantifies the degree of deviation. It can be used to capture disease fluctuations and generate graded diagnostic influence signals. In this embodiment, the symptom tracking module can use time series analysis methods to model repeatedly collected diagnostic data. The dynamic symptom change data can be a longitudinal record of the symptoms and signs of the same patient collected at multiple time points, which can be used to support the analysis of disease evolution trends. For example, the dynamic symptom change data can include, but is not limited to, symptom fluctuation data within a cycle, pre- and post-treatment comparison data, and diurnal rhythm symptom data. The tracking period can be the time window in which the symptom tracking module performs data collection and analysis, which can be used to limit the time range of dynamic modeling. In a specific embodiment, the tracking period can include, but is not limited to, menstrual cycle tracking, treatment course stage tracking, and diurnal fluctuation tracking.

[0026] The symptom change sequence can be a time-series vector composed of dynamic symptom change data arranged chronologically within a tracking period, which can be used to calculate the symptom deviation. Further, the symptom change sequence can include, but is not limited to, primary symptom change sequences, secondary symptom change sequences, and tongue and pulse combined change sequences. The symptom deviation can be the quantified degree of deviation of the symptom change sequence from the baseline state, which can be used to measure the intensity of disease fluctuations. For example, the symptom deviation can include, but is not limited to, amplitude deviation, frequency deviation, and persistence deviation. The deviation threshold can be a preset critical value used to determine whether the symptom deviation reaches the clinical intervention level, which can be used to trigger the generation of a syndrome differentiation influence level signal. In an exemplary embodiment, the deviation threshold can include, but is not limited to, a mild intervention threshold, a moderate intervention threshold, and a severe intervention threshold. The syndrome differentiation influence level signal can be a graded response signal generated based on the comparison result of the symptom deviation and the threshold, which can be used to guide the intensity of subsequent syndrome differentiation correction. Further, the syndrome differentiation influence level signal can include, but is not limited to, a first-level influence signal, a second-level influence signal, and a third-level influence signal. A high-level syndrome differentiation signal can be a specific syndrome differentiation influence level signal generated when the symptom deviation is greater than or equal to the deviation threshold, which can be used to activate the syndrome differentiation correction module. For example, high-level diagnostic signals may include, but are not limited to, significant fluctuation signals, early warning signals of syndrome transformation, and signals of constitution imbalance.

[0027] Based on the diagnostic deviation signal, the dynamic changes in symptoms in the diagnostic data can be tracked by initiating periodic data collection after the deviation signal is activated. In one specific embodiment, tracking the dynamic changes in symptoms in the diagnostic data based on the diagnostic deviation signal can be achieved by automatically scheduling daily symptom reports from the patient's mobile app, thereby establishing a window for observing the evolution of the disease. Within the tracking period, a symptom change sequence is obtained, which can be the symptom data points collected in chronological order. Furthermore, within the tracking period, the symptom change sequence can be obtained by constructing a database table with timestamp indexes, thereby forming an analyzable time-series structure. Obtaining the symptom deviation degree and comparing it with a deviation threshold can be achieved by calculating the difference between the symptom change sequence and the baseline and comparing it with a preset threshold. For example, obtaining the symptom deviation degree and comparing it with the deviation threshold can be achieved by calculating the overall deviation degree using Euclidean distance, thereby quantifying the fluctuation of the disease and determining its clinical significance.

[0028] Generating a diagnostic impact level signal can be achieved by outputting a graded signal based on the relative relationship between the deviation degree and a threshold. In an exemplary embodiment, generating the diagnostic impact level signal can be achieved by outputting discrete level codes (such as 0 / 1 / 2), thereby providing a basis for differentiated response strategies. If the symptom deviation degree is greater than or equal to the deviation threshold, generating a high-level diagnostic signal can be achieved by activating a specific signal when the deviation degree meets the threshold condition. Furthermore, if the symptom deviation degree is greater than or equal to the deviation threshold, generating a high-level diagnostic signal can be achieved by setting a high-priority interrupt signal, thereby triggering the diagnostic correction process.

[0029] The syndrome differentiation correction module 40 is used to obtain the syndrome type conversion coefficient and the constitution compensation value based on the high-level syndrome differentiation signal, and to calculate the comprehensive syndrome differentiation correction value by superimposing the syndrome type conversion coefficient and the constitution compensation value. The syndrome differentiation correction module is a processing unit that generates a comprehensive syndrome differentiation correction value by fusing syndrome transformation coefficients and constitution compensation values ​​based on high-level syndrome differentiation signals. This can be used to improve the personalization and accuracy of syndrome differentiation conclusions. In this embodiment, the syndrome differentiation correction module can integrate multi-source correction factors through weighted superposition or neural network fusion mechanisms. Syndrome transformation coefficients can be quantitative factors reflecting the evolution trend of syndromes from one type to another, and can be used to dynamically adjust syndrome differentiation conclusions to match disease progression. For example, syndrome transformation coefficients may include, but are not limited to, the liver stagnation to fire coefficient, the qi deficiency to yang deficiency coefficient, and the blood stasis to phlegm-dampness coefficient. Constitution compensation values ​​can be correction amounts applied to syndrome differentiation parameters based on individual innate constitution differences, and can be used to improve the individual adaptability of syndrome differentiation conclusions. In a specific embodiment, constitution compensation values ​​may include, but are not limited to, balanced constitution compensation values, yang deficiency constitution compensation values, and yin deficiency constitution compensation values.

[0030] The comprehensive diagnostic correction value can be the final correction parameter obtained by superimposing the syndrome transformation coefficient and the constitution compensation value, and can be used as the basis for acupoint selection optimization. Furthermore, the comprehensive diagnostic correction value can include, but is not limited to, dynamic constitution correction value, syndrome migration correction value, and composite correction value. Based on high-level diagnostic signals, obtaining the syndrome transformation coefficient and constitution compensation value can be achieved by querying a pre-trained model or rule base to extract the corresponding correction factors. For example, based on high-level diagnostic signals, obtaining the syndrome transformation coefficient and constitution compensation value can be achieved by retrieving the coefficients corresponding to the transformation path from the syndrome evolution map, thereby introducing dynamic and individualized correction criteria. Superimposing the syndrome transformation coefficient and constitution compensation value to obtain the comprehensive diagnostic correction value can be achieved by performing a mathematical synthesis operation on the two correction factors. In an exemplary embodiment, superimposing the syndrome transformation coefficient and constitution compensation value to obtain the comprehensive diagnostic correction value can be achieved by performing weighted linear superposition, thereby generating a unified diagnostic correction parameter.

[0031] The acupoint optimization module 50 is used to obtain dynamic acupoint coefficients based on comprehensive syndrome differentiation correction values. It integrates standard acupoint parameters with dynamic acupoint coefficients to calculate the required acupoint parameters and completes the optimization of the positioning of the thirteen acupoints for regulating menstruation and promoting pregnancy. The dynamic acupoint coefficients are obtained by weighting and summing all the symptom deviations and taking the average value.

[0032] The acupoint optimization module can be an execution module that dynamically adjusts standard acupoint parameters based on comprehensive diagnostic correction values ​​to optimize the positioning of the thirteen acupuncture points for regulating menstruation and promoting fertility. It can be used to achieve personalized and dynamic optimization of acupoint positioning. In this embodiment, the acupoint optimization module can perform mathematical fusion calculations between the dynamic acupoint coefficient and the standard acupoint coordinates. The dynamic acupoint coefficient can be a coefficient used to adjust acupoint parameters, obtained by weighting and averaging the deviations of all symptoms, and can be used to achieve real-time personalized adjustment of acupoint positioning. For example, the dynamic acupoint coefficient may include, but is not limited to, the primary symptom weight coefficient, the secondary symptom weight coefficient, and the tongue and pulse synergy coefficient. The standard acupoint parameters can be the standard anatomical positioning coordinates and operational specifications of each acupoint in the thirteen acupuncture point scheme for regulating menstruation and promoting fertility, and can be used as a benchmark reference for acupoint optimization. In a specific embodiment, the standard acupoint parameters may include, but are not limited to, the standard parameters of Guanyuan (CV4), Sanyinjiao (SP6), and Zigung (GV15) acupoints.

[0033] The required acupoint parameters for adjustment can be personalized acupoint positioning parameters obtained by fusing standard acupoint parameters with dynamic acupoint matching coefficients, which can be used to guide actual acupuncture operations. Furthermore, the required acupoint parameters for adjustment can include, but are not limited to, offset coordinate parameters, depth adjustment parameters, and stimulation intensity parameters. The thirteen acupoint positioning for regulating menstruation and promoting fertility can be a spatial positioning set of thirteen specific acupoints used to treat menstrual disorders and infertility, which can be used as the anatomical basis for implementing acupuncture interventions. For example, the thirteen acupoint positioning for regulating menstruation and promoting fertility can include, but is not limited to, the positioning of acupoints on the Ren meridian, the Foot Taiyin Spleen meridian, and the Foot Shaoyin Kidney meridian.

[0034] Based on the comprehensive diagnostic correction value, the dynamic acupoint selection coefficient is obtained, which can be achieved by mapping the comprehensive diagnostic correction value to the acupoint selection adjustment weight. In an exemplary embodiment, the dynamic acupoint selection coefficient based on the comprehensive diagnostic correction value can be obtained by transforming the correction value into coefficients using a piecewise linear function, thereby establishing a quantitative correlation between diagnostics and acupoint selection. The standard acupoint parameters and the dynamic acupoint selection coefficient are fused together to obtain the required adjusted acupoint parameters, which can be achieved by applying coefficient-driven offset or scaling to the standard coordinates. Furthermore, the fusion of the standard acupoint parameters and the dynamic acupoint selection coefficient to obtain the required adjusted acupoint parameters can be achieved by performing vector offset on a three-dimensional anatomical model, thereby outputting personalized acupoint positioning results.

[0035] Optimizing the location of the thirteen acupuncture points for regulating menstruation and promoting fertility can result in the output of the final adjusted acupuncture point parameters for clinical use. For example, this optimization can be achieved by generating an AR visualization guide map, enabling precise acupuncture intervention. The dynamic acupuncture point selection coefficient is obtained by weighted summation and averaging of all symptom deviations. This can be done by assigning weights to the deviations of each symptom dimension and calculating the average. In a specific embodiment, the weights for weighted summation and averaging of all symptom deviations to obtain the dynamic acupuncture point selection coefficient can be determined using an expert weighting method, thus comprehensively reflecting the influence of multidimensional symptoms on acupuncture point selection.

[0036] Taking the intelligent diagnosis and treatment of patients with delayed menstruation and ovulation disorders as an example, the AI ​​intelligent syndrome differentiation and acupoint positioning assistance system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility in this embodiment can be as follows: During the patient's initial consultation, the four diagnostic methods acquisition module obtains a weak pulse through an intelligent pulse diagnosis instrument, the tongue image camera identifies a pale tongue and a thin white coating, and an electronic questionnaire records symptoms such as delayed menstruation and soreness in the lower back and knees; the system generates initial syndrome differentiation parameters pointing to kidney yang deficiency syndrome. The syndrome differentiation comparison module finds that some parameters are close to but do not completely fall within the standard range of kidney yang deficiency, triggering a slight syndrome differentiation deviation signal. The symptom tracking module collects basal body temperature, premenstrual breast tenderness, and mood fluctuation data daily during the subsequent 28-day menstrual cycle to form a symptom change sequence; if the LH peak does not appear on the 21st day, the symptom deviation exceeds the threshold, generating a high-level syndrome differentiation signal. The syndrome differentiation correction module calls the syndrome conversion coefficient of "transformation from kidney yang deficiency to yin and yang deficiency" and combines it with the patient's usual cold-sensitive yang deficiency constitution compensation value to calculate a comprehensive syndrome differentiation correction value. Based on this, the acupoint optimization module increases the needle insertion depth of the originally planned acupoints such as Guanyuan and Qihai by 15%, and slightly adjusts the position of Sanyinjiao towards the medial side of the tibia. The dynamic acupoint selection coefficient is calculated by weighting the primary symptom (delayed menstruation) by 0.6, the secondary symptom (aversion to cold) by 0.3, and the tongue and pulse by 0.1. Finally, personalized acupoint parameters are output to the physician's terminal to achieve precise treatment with the thirteen acupuncture points for regulating menstruation and promoting fertility.

[0037] In one embodiment, the initial dialectical parameters are obtained as follows: The symptom regions in the syndrome differentiation data are divided into several syndrome differentiation sub-regions. The number of symptom features in each syndrome differentiation sub-region is obtained. The number of symptom features in each syndrome differentiation sub-region is weighted, summed, and averaged to obtain the symptom feature density value. The initial diagnostic parameters are obtained by calculating the ratio between the symptom feature density value and the standard syndrome density value.

[0038] The symptom region can be the spatial or logical range corresponding to a structured set of symptoms in a patient's clinical manifestations. It can serve as the basic unit for dividing the syndrome differentiation sub-regions, carrying spatial distribution information of symptom features. In this embodiment, the symptom region may include, but is not limited to, one or more of menstrual-related symptom regions, emotion-related symptom regions, and physical sign-related symptom regions. The syndrome differentiation sub-region can be a logical unit that further subdivides the symptom region according to anatomical, functional, or syndrome dimensions. It can be used to preserve the spatial heterogeneity of symptom distribution and support local feature density calculation. Furthermore, the syndrome differentiation sub-region can be divided based on traditional Chinese medicine syndrome theory or data-driven clustering methods. For example, the syndrome differentiation sub-region may include, but is not limited to, one or more of liver qi stagnation sub-regions, kidney yang deficiency sub-regions, and phlegm-dampness obstruction sub-regions.

[0039] The number of symptom features can be the total number of identifiable and countable independent symptoms or signs within a single syndrome differentiation sub-region. It can reflect the pathological activity level of that sub-region and serve as the basis for density calculation. In an exemplary embodiment, the number of symptom features may include, but is not limited to, one or more of the following: the number of primary symptom features, the number of secondary symptom features, and the number of tongue and pulse related features. The symptom feature density value can be a comprehensive quantitative index obtained by weighted summation and averaging of the number of symptom features in each syndrome differentiation sub-region. It can be used to characterize the clustering intensity and distribution pattern of symptoms in spatial structure. Furthermore, the symptom feature density value can be obtained by using a weighted average method to fuse the feature numbers of different sub-regions, with the weights reflecting clinical importance. The standard syndrome density value can be a standardized reference value for the symptom feature density of each syndrome differentiation sub-region, derived from statistical analysis of typical syndrome clinical data. It can be used as a benchmark for ratio calculation, measuring the degree of matching between the current patient and the standard syndrome. In a specific embodiment, the standard syndrome density value may include, but is not limited to, one or more of the following: the standard density value for kidney deficiency syndrome, the standard density value for liver stagnation syndrome, and the standard density value for blood stasis syndrome.

[0040] Dividing symptom regions in the diagnostic data into several sub-regions can be achieved through structured segmentation based on TCM syndrome classification logic or machine learning clustering results. Further, this operation can be implemented through rule-based segmentation based on the syndrome type guidelines of "Traditional Chinese Medicine Gynecology," or by using unsupervised clustering algorithms (such as K-means) to discover sub-regions from the symptom co-occurrence matrix of historical cases. This preserves the spatial heterogeneity of symptom distribution and avoids information loss due to averaging. Obtaining the number of symptom features within each sub-region can be achieved by counting the number of symptom entries or sign markers falling into each sub-region. Further, this operation can be achieved through counting by matching structured electronic medical record fields, or by extracting and classifying symptoms into sub-regions from free text using natural language processing, thereby quantifying the pathological intensity of each sub-region. Finally, the symptom feature density value is obtained by weighted summation and averaging of the number of symptom features within each sub-region. This can be achieved by multiplying the feature counts of each sub-region by preset weights, summing the results, and taking the arithmetic mean. Furthermore, this operation can be achieved by setting the weight of the primary symptom sub-region to 1.0 and the weight of the secondary symptom sub-region to 0.5, or by determining the optimal weights through physician expert scoring or backpropagation learning, thereby generating a single density index reflecting the overall symptom distribution pattern. The initial diagnostic parameters are obtained by calculating the ratio between the symptom feature density values ​​and the standard syndrome density values. This can be done by performing item-by-item or overall ratio calculations to map the patient's density values ​​to the standard syndrome space. Further, this operation can be achieved by calculating the ratio vector of the patient's density values ​​to the density values ​​of each standard syndrome as the initial diagnostic parameters, or by using the ratio normalized by cosine similarity as the parameter component, thus realizing the transformation from experience-based description to data-driven diagnostics and providing comparable initial parameters.

[0041] For example, in the initial diagnosis of patients with complex syndromes (mixed liver stagnation and kidney deficiency), the AI ​​intelligent diagnosis and acupoint positioning assistance system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility in this embodiment can be as follows: After receiving the diagnosis data, the system divides the symptom area into three sub-regions: "emotional regulation," "reproductive endocrinology," and "lumbar and knee function." In the "emotional regulation" sub-region, four symptoms such as irritability and chest and rib distension are identified; in the "reproductive endocrinology" sub-region, three symptoms such as delayed menstruation and scanty and pale menstrual flow are identified; and in the "lumbar and knee function" sub-region, two symptoms such as lower back pain and frequent urination at night are identified. Based on the clinical weights (emotions 0.4, reproductive 0.5, lumbar and knee 0.1), the weighted summation and mean value yield a symptom feature density value of 3.2. This value is compared with the standard density value of liver stagnation syndrome (4.0) and the standard density value of kidney deficiency syndrome (3.5) to obtain the initial diagnosis parameter vector [0.8, 0.91], indicating that it is closer to kidney deficiency syndrome but has obvious liver stagnation components. This parameter is then input into the syndrome comparison module, triggering an in-depth analysis process for the complex syndrome.

[0042] In one embodiment, the symptom deviation degree is obtained as follows: Obtain the primary symptom offset component and the secondary symptom offset component, and then perform vector synthesis on the primary symptom offset component and the secondary symptom offset component to obtain the symptom deviation degree.

[0043] The primary symptom offset component can be a vector-based quantitative indicator reflecting the changes in key primary symptoms (such as menstrual cycle abnormalities, ovulation disorders, etc.) relative to the baseline state during the tracking period. It includes the direction and magnitude of change and can be used to characterize the dynamic evolution of the core pathological state, serving as a major component of symptom deviation. In this embodiment, the primary symptom offset component can be obtained by mapping the changes in primary symptoms to a preset clinical dimension coordinate system (such as a three-dimensional space of cycle-ovulation-body temperature) in conjunction with the context to generate a vector. For example, the primary symptom offset component can include, but is not limited to, one or more of the following: cycle disorder offset component, basal body temperature offset component, and LH peak absence offset component. The secondary symptom offset component can be a vector-based quantitative indicator reflecting the changes in auxiliary symptoms (such as mood fluctuations, sleep quality, breast tenderness, etc.) relative to the baseline state during the tracking period. It can be used to supplement primary symptom information and enhance the perception dimension of overall syndrome fluctuations. In an exemplary embodiment, the secondary symptom offset component can be obtained by clustering by symptom category and projecting it to an auxiliary subspace in conjunction with the context. Further, the secondary symptom offset component can include, but is not limited to, one or more of the following: emotion-related offset component, digestion-related offset component, and physical discomfort offset component.

[0044] Obtaining the primary and secondary symptom offset components can be achieved by separately calculating the changes in the primary and secondary symptoms relative to an initial or standard baseline within the tracking period and assigning directional attributes to form vectors. Further, this operation can be implemented by assigning a unit directional basis vector to the primary symptom and then superimposing the weighted and scaled secondary symptom components into the same vector space, thus enabling structured separation modeling of the dynamic changes in primary and secondary symptoms. The primary and secondary symptom offset components are then vector-synthesized to obtain the symptom deviation degree. This can be achieved by performing a mathematical synthesis operation on the two vectors to generate a single comprehensive deviation degree vector or its magnitude. In a specific embodiment, this operation can be implemented by calculating the Euclidean norm of the weighted vector sum as a scalar deviation degree, or by retaining the synthesized multidimensional vector form for direct use by subsequent neural network modules, thus preserving the directionality and hierarchical differences of symptom changes and improving the clinical sensitivity of deviation assessment. The symptom deviation degree, obtained by vector synthesis of the primary and secondary symptom offset components, can be used to quantify the overall intensity and direction of dynamic fluctuations in the condition and to trigger signals indicating the level of influence of syndrome differentiation. In this embodiment, the symptom deviation can be synthesized from the two types of offset components by combining the context and using methods such as vector addition, weighted Euclidean norm, or tensor fusion.

[0045] Taking the dynamic assessment of patients with ovulation disorders during the mid-term treatment as an example, the AI ​​intelligent syndrome differentiation and acupoint positioning assistance system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility in this embodiment can be as follows: At the initial consultation, the patient's main symptoms are delayed menstruation (45-day cycle) and anovulation, and secondary symptoms include irritability and insomnia. During the follow-up on the 10th day of the second course of treatment, the basal body temperature is still monophasic (main symptom deviation component: [−0.8, −1.0], indicating that ovulation function has not recovered and the cycle continues to prolong), but irritability is reduced and sleep is improved (secondary symptom deviation component: [+0.3, +0.4], indicating a positive deviation in the emotional dimension). The system assigns a high weight (e.g., 1.0) to the main symptom deviation component and a low weight (e.g., 0.4) to the secondary symptoms. After vector synthesis, the symptom deviation modulus is 1.12, which exceeds the deviation threshold of 1.0, triggering a high-level syndrome differentiation signal. Although the secondary symptoms improved, the primary symptoms worsened, causing the overall trend to deviate. Based on this, the system maintained the diagnosis of kidney yang deficiency and strengthened the combination of warming and tonifying acupoints to avoid misjudging the efficacy due to the improvement of secondary symptoms.

[0046] In one embodiment, the primary symptom offset component is obtained as follows: The ratio of the symptom change interval value and the symptom characteristic quantification value within the same tracking period is calculated to obtain the symptom synchronization coefficient. Periods with symptom synchronization coefficients that meet preset conditions are selected and marked as valid diagnosis periods. The duration of the effective diagnostic period is obtained, and the ratio of the duration of the effective diagnostic period to the total duration of the tracking period is calculated to obtain the main symptom offset component.

[0047] The tracking period can be several continuous or discrete time windows divided within the tracking cycle, used for dynamic analysis of local symptoms. In this embodiment, the tracking period serves as the basic time unit for calculating the symptom synchronization coefficient. For example, the tracking period can include, but is not limited to, one or more of the following: the menstrual follicular phase, the ovulation phase, and the luteal phase. The symptom change interval value can be a quantitative representation of the time interval between two occurrences or significant changes of the primary symptom within the same tracking period. Further, the symptom change interval value can reflect the temporal frequency characteristics of the primary symptom's evolution. In an exemplary embodiment, the symptom change interval value can include, but is not limited to, periodic interval values, sudden interval values, and gradual interval values. The symptom characteristic quantification value can be a numerical measure of the severity, intensity, or manifestation of the primary symptom within the same tracking period. In a specific embodiment, the symptom characteristic quantification value can characterize the pathological activity level of the primary symptom within that period. For example, the symptom characteristic quantification value can include, but is not limited to, menstrual flow score, basal body temperature fluctuation amplitude, and LH peak absence confidence value.

[0048] The symptom synchronization coefficient can be the ratio of the symptom change interval value to the quantified value of the symptom characteristics, used to measure the synergistic consistency of the main symptoms in terms of time and intensity. In this embodiment, the symptom synchronization coefficient can identify clinically effective time periods with stable evolution patterns. Further, the symptom synchronization coefficient can be generated through ratio calculation; a higher value indicates more concentrated and regular symptoms. In a specific embodiment, the symptom synchronization coefficient can be obtained by dividing the quantified value by the interval value. Preset conditions can be a set of screening rules used to determine whether the symptom synchronization coefficient has clinical diagnostic value. For example, preset conditions can filter noise or occasional fluctuations, retaining reliable diagnostic data. In an exemplary embodiment, preset conditions may include, but are not limited to, lower threshold conditions, trend consistency conditions, and periodic repetition conditions. The effective diagnostic time period can be the tracking time period in which the symptom synchronization coefficient meets the preset conditions, marked as a reliable time period that can be used for diagnostic analysis. In this embodiment, the effective diagnostic time period serves as the data basis for calculating the main symptom offset component. Further, the effective diagnostic time period may include, but is not limited to, high synchronization time periods, moderately stable time periods, and low noise interference time periods. The duration of the effective diagnostic time period can be the sum of the durations of all time periods marked as effective diagnostic time periods. In one specific embodiment, the duration of the effective diagnostic period can quantify the cumulative time of stable evolution of the main symptom. For example, the duration of the effective diagnostic period may include, but is not limited to, continuous effective duration, cumulative effective duration, weighted effective duration, etc.

[0049] The total duration of the tracking period can be the length of the complete time window during which the symptom tracking module performs monitoring. In this embodiment, the total duration of the tracking period serves as the normalization benchmark for calculating the primary symptom offset component. Furthermore, the total duration of the tracking period can include, but is not limited to, the duration of the entire menstrual cycle, the duration of the treatment phase, or the total duration of daytime monitoring. The symptom synchronization coefficient is obtained by calculating the ratio between the symptom change interval value and the symptom characteristic quantification value within the same tracking period. This can be achieved by dividing the symptom characteristic quantification value by the symptom change interval value (or its reciprocal) for each tracking period. Further, this operation can be implemented by directly calculating the quantification value divided by the interval value (unit: days); if the interval is 0, it is set to the maximum value, or the ratio of the smoothed quantification value to the average interval can be used, thereby constructing a time-intensity dual-dimensional synergy indicator to identify the regularity of symptom evolution.

[0050] The process involves selecting time periods where the symptom synchronization coefficient meets preset criteria and marking them as valid diagnostic periods. This can be achieved by comparing the synchronization coefficient of each tracking period with preset criteria (such as thresholds or trend slopes), and marking those that meet the criteria as valid. Furthermore, this operation can be achieved by setting a synchronization coefficient greater than 0.7 as valid, or by combining the moving average trend to determine whether the synchronization coefficient consistently meets the standard, thereby excluding sporadic, noisy, or unreliable data and retaining time periods with diagnostic significance. The duration of valid diagnostic periods can be obtained by summing the durations of all tracked periods marked as valid. This operation can also be achieved by directly summing the seconds or days of each valid period, or by merging overlapping periods before calculating the total duration, thereby quantifying the total time of stable manifestation of the main symptom. The ratio of the duration of valid diagnostic periods to the total duration of the tracking period is calculated to obtain the main symptom offset component. This can be achieved by performing a division operation where the valid duration is divided by the total duration, outputting the ratio value within the interval [0, 1]. Furthermore, this operation can be achieved by directly using the ratio as a scalar offset component, or by mapping the ratio to a preset clinical level (e.g., 0–0.3 for low offset, 0.3–0.7 for medium offset, and >0.7 for high offset) and then encoding it into a vector, thereby converting the persistence and stability of the main symptom into an offset component that can participate in vector synthesis.

[0051] For example, in the scenario of assessing the stability of the primary symptom during the treatment of patients with oligomenorrhea, the AI-powered intelligent syndrome differentiation and acupoint positioning assistance system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility in this embodiment can be as follows: The system collects data every 2 days within a 28-day tracking cycle. On days 6–10, the patient reports delayed menstruation and a persistent monophasic basal body temperature, with a symptom change interval of 4 days (from the last record to this one), and a symptom characteristic quantification value of 0.9 (high confidence anovulation). The synchronization coefficient = 0.9 / 4 = 0.225. On days 18–26, three consecutive records show monophasic basal body temperature and no menstruation, with the interval shortened to 2 days, the quantification value maintained at 0.85, and the synchronization coefficient = 0.85 / 2 = 0.425. The preset condition requires a synchronization coefficient ≥ 0.4, therefore only days 18–26 (a total of 8 days) are marked as the effective syndrome differentiation period. The effective duration of 8 days ÷ the total duration of 28 days ≈ 0.286, which is used as the input vector synthesis module for the primary symptom offset component. Although the main symptoms are present throughout the entire process, only the periods of high synchronicity are included to avoid interference from early uncertain data in the diagnosis, so that subsequent deviation assessments can focus more on reliable evolution stages.

[0052] In one embodiment, the symptom change interval value and the symptom feature quantification value are obtained as follows: Within the tracking period, the tracking period is divided into several time segments. A time-symptom coordinate system is established, with the horizontal dimension representing the time segment and the vertical dimension representing the symptom feature quantification value corresponding to each time segment. The obtained symptom feature quantification value is substituted into the time-symptom coordinate system to draw the symptom change curve. Based on the symptom change curve, the peak points and valley points of symptoms are obtained. The difference between the coordinates of adjacent peak points and valley points is calculated to obtain the quantitative value of symptom characteristics. Within the tracking period, the tracking period is divided into several time segments, and a time-symptom change coordinate system is established. The horizontal dimension represents the time segment, and the vertical dimension represents the frequency of symptom changes corresponding to each time segment. The obtained symptom change frequency is substituted into the time-symptom change coordinate system to draw the symptom change frequency curve. Based on the symptom change frequency curve, the points of increasing and decreasing frequency of change are obtained. The difference between the coordinates of adjacent points of increasing and decreasing frequency of change is calculated to obtain the symptom change interval value.

[0053] In this system, a time segment can be a continuous time unit divided according to fixed or adaptive rules of the tracking period, serving as the basic time granularity for constructing the time-symptom coordinate system and the time-symptom change coordinate system. In this embodiment, the time segment can divide the tracking period according to a preset time granularity or an event-driven approach, thereby providing a unified time reference for dual-coordinate system modeling. The time-symptom coordinate system can be a two-dimensional coordinate system constructed with time segments as the horizontal axis and symptom feature quantification values ​​as the vertical axis, which can be used to visualize and analyze the evolution trend of symptom intensity over time. Furthermore, the time-symptom coordinate system can be implemented by mapping structured time-series data to the Cartesian coordinate plane. In an exemplary embodiment, the time-symptom coordinate system can adopt a standard Cartesian coordinate system or a logarithmic vertical axis to compress the range of high-intensity symptoms. The symptom change curve can be a continuous or discrete curve formed by connecting the symptom feature quantification values ​​corresponding to each time segment in the time-symptom coordinate system, which can be used to characterize the dynamic trajectory of the main symptom intensity and support extreme point extraction. In a specific embodiment, the symptom change curve can be formed by mapping the quantification value of each time segment to coordinate points and connecting them into a line to form an analyzable continuous dynamic trajectory. For example, the symptom change curve may include, but is not limited to, one or more of the following: smooth interpolation curve, stepped discrete curve, spline fitting curve, etc.

[0054] Symptom peaks can be the coordinate points corresponding to local maxima on the symptom change curve, and can be used to identify key time points when symptoms worsen or become active. Furthermore, symptom peaks can be identified by comparing the sign of the first derivative or using a sliding window to extract key turning points in symptom evolution from local extrema. In one embodiment, symptom peaks can include, but are not limited to, one or more of global peaks, intra-cycle peaks, and treatment response peaks. Symptom troughs can be the coordinate points corresponding to local minima on the symptom change curve, and can be used to identify key time points when symptoms subside or return to rest. Furthermore, symptom troughs can be identified by comparing the sign of the first derivative or using a sliding window to extract key turning points in symptom evolution from local extrema. In one specific embodiment, symptom troughs can include, but are not limited to, one or more of baseline troughs, post-treatment troughs, and cyclic troughs. The difference between the coordinates of adjacent symptom peaks and symptom troughs is calculated to obtain a quantified symptom characteristic value, which can be the absolute value of the difference between adjacent extreme points on the vertical axis (intensity) or the horizontal axis (time). Furthermore, this operation can be achieved by taking the difference in the vertical axis as the intensity fluctuation value or taking the difference in the horizontal axis as the duration of symptom duration / interval, thereby quantifying the amplitude or cycle length of symptom fluctuations as feature inputs.

[0055] The time-symptom change coordinate system can be a two-dimensional coordinate system constructed with time segments as the horizontal axis and symptom change frequency as the vertical axis, which can be used to analyze the temporal patterns of symptom fluctuation frequency. Furthermore, the time-symptom change coordinate system can be implemented by statistically analyzing the number of symptom state switching events within a unit time segment and mapping it to the coordinate plane. In a specific embodiment, the time-symptom change coordinate system can normalize the frequency to the [0, 1] interval, or retain the original count to preserve absolute activity information. Symptom change frequency can be the number of times the symptom state (e.g., appearance / disappearance, aggravation / relief) changes within a single time segment, and can be used to reflect symptom instability or fluctuation activity. In an exemplary embodiment, symptom change frequency can include, but is not limited to, one or more of the following: primary symptom switching frequency, secondary symptom fluctuation frequency, and compound symptom alternation frequency.

[0056] A symptom change frequency curve can be a curve formed by connecting the symptom change frequencies corresponding to each time segment in a time-symptom change coordinate system. It can be used to reveal the regularity or disorder of symptom fluctuation rhythm. Furthermore, the symptom change frequency curve can be visualized by mapping the frequency values ​​of each time segment to points and connecting them. For example, the symptom change frequency curve can be one or more of the following: a monotonically increasing curve, a periodic oscillation curve, and a random fluctuation curve. A frequency increase point can be a turning point on the symptom change frequency curve where the slope changes from negative to positive or continues to rise, which can be used to identify the starting moment when symptom fluctuations become more active. Furthermore, the frequency increase point can identify the trend reversal and locate the start and end moments of symptom fluctuation activity by changing the slope sign or the second derivative. In a specific embodiment, the frequency increase point can be one or more of the following: an acceleration point, a plateau breakout point, and a threshold crossing point.

[0057] The frequency decrease point can be the inflection point on the symptom frequency curve where the slope changes from positive to negative or continues to decline, and can be used to identify the starting moment when symptom fluctuations tend to stabilize. Furthermore, the frequency decrease point can identify the trend reversal and locate the start and end points of symptom fluctuation activity by the change in the slope sign or the second derivative. In an exemplary embodiment, the frequency decrease point can be one or more of the following, including but not limited to deceleration inflection points, stable transition points, and intervention response points. The symptom change interval value is obtained by calculating the difference between the coordinates of adjacent frequency increase points and frequency decrease points; this can be achieved by calculating the distance between adjacent increase-decrease point pairs on the time axis. Further, this operation can be achieved by taking the difference in the horizontal coordinates as the interval value or by combining it with frequency amplitude weighting to calculate the effective interval, thereby quantifying the complete cycle length of symptom fluctuations from active to subsided.

[0058] Taking dynamic symptom modeling of patients with premenstrual tension syndrome (PMS) as an example, the AI-powered intelligent diagnosis and acupoint positioning assistance system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility in this embodiment can be as follows: The system divides the 28-day tracking cycle into 14 two-day time segments. In the time-symptom coordinate system, the mood irritability score (0–5 points) forms a fluctuation curve, identifying extreme points such as day 6 (peak 4.2), day 12 (trough 1.8), and day 20 (peak 3.9); the difference between the vertical coordinates of adjacent peaks and troughs is 2.4 and 2.1, respectively, and the average value is taken to obtain the symptom feature quantification value of 2.25. In the time-symptom change coordinate system, the number of mood state switching times every two days (e.g., "calm → irritable → calm" is counted as 2 times) is counted, and a frequency curve is plotted. It is found that the frequency increases from 1 to 3 on days 5–7 (increasing point), and decreases to 1 on days 8–10 (decreasing point), with a time difference of 4 days, which is used as the symptom change interval value. The interval value and the quantification value are then used to calculate the synchronization coefficient to determine whether PMS symptoms show regular periodic fluctuations, thereby distinguishing between physiological fluctuations and pathological aggravation.

[0059] In one embodiment, the secondary symptom offset component is obtained as follows: The symptom deviation value of a sub-region is compared with the allowable deviation value of the sub-region. If the symptom deviation value of a sub-region is greater than or equal to the allowable deviation value of the sub-region, the sub-region is marked as a symptom abnormal sub-region. If the symptom deviation value of a sub-region is less than the allowable deviation value of the sub-region, the sub-region is marked as a symptom normal sub-region. The correlation value of the sub-region's vital signs is compared with the correlation threshold of the sub-region. If the correlation value of the sub-region's vital signs is less than the correlation threshold of the sub-region, the sub-region is marked as a sub-region with abnormal vital signs correlation. If the correlation value of the sub-region's vital signs is greater than or equal to the correlation threshold of the sub-region, the sub-region is marked as a sub-region with normal vital signs correlation. Obtain the symptom abnormal sub-regions, perform spatial overlap analysis on the symptom abnormal sub-regions and the sign-related abnormal sub-regions to obtain overlapping abnormal sub-regions, obtain the number of overlapping abnormal sub-regions, sum the number of overlapping abnormal sub-regions to obtain the total number of abnormal regions, calculate the ratio of the total number of abnormal regions to the total number of syndrome differentiation sub-regions to obtain the secondary symptom offset component.

[0060] The sub-region symptom deviation value can be a quantitative deviation between the current symptom presentation and the expected symptoms of a standard syndrome within a single syndrome sub-region, and can be used to measure whether local symptoms exceed the normal fluctuation range. In an exemplary embodiment, the sub-region symptom deviation value may include, but is not limited to, one or more of the following: emotional sub-region deviation value, digestive sub-region deviation value, sleep sub-region deviation value, etc. The sub-region deviation allowable value can be a preset upper limit threshold used to determine whether the sub-region symptom deviation is acceptable, and can be used as a baseline for judging symptom abnormality. For example, the sub-region deviation allowable value can be the allowable value of the primary symptom-related sub-region, the allowable value of the secondary symptom-related sub-region, the allowable value of the constitution-sensitive sub-region, etc. The symptom abnormal sub-region can be a syndrome sub-region where the sub-region symptom deviation value is greater than or equal to the sub-region deviation allowable value, and can be used to identify areas where the symptom presentation deviates significantly from the expectation. Further, the symptom abnormal sub-region may include a high-deviation emotional area, a persistent insomnia area, an abnormally irritable area, etc.

[0061] A symptom-normal sub-region can be a sub-region whose symptom deviation value is less than the allowable deviation value, and can be used to identify areas where symptoms are within a reasonable fluctuation range. Comparing the symptom deviation value of a sub-region with the allowable deviation value can be performed on each sub-region. Further, this operation can be achieved by performing if-else logic judgments on a sub-region-by-sub-region basis or by generating a Boolean mask after batch vectorized comparison, thereby initially identifying areas with abnormal symptom presentation. If the symptom deviation value of a sub-region is greater than or equal to the allowable deviation value, then the sub-region is marked as a symptom-abnormal sub-region; this can be achieved by assigning an abnormal label when the deviation value meets the standard. In a specific embodiment, this operation can be achieved by setting an abnormal flag bit or adding the sub-region ID to an abnormal list, thereby establishing a set of symptom-abnormal regions.

[0062] If the symptom deviation value of a sub-region is less than the allowable deviation value for that sub-region, then that sub-region is marked as a normal symptom sub-region. This can be achieved by assigning a "normal" label when the deviation value does not exceed the limit. For example, this operation can be implemented by defaulting to a normal state without explicit marking, or by explicitly recording the normal region ID, thus preserving information about non-abnormal regions. The sub-region sign correlation value can be a quantitative indicator measuring the consistency between the symptom description and the corresponding objective signs such as tongue appearance and pulse within a certain diagnostic sub-region, and can be used to assess the credibility of symptom-sign synergy. In this embodiment, the sub-region sign correlation value can be calculated using a multimodal alignment model or rule matching to determine the semantic / numerical consistency between symptoms and signs. Further, the sub-region sign correlation value can include tongue-pulse-emotion correlation values, pulse-sleep correlation values, tongue coating-digestion correlation values, etc. The sub-region correlation threshold can be a preset minimum correlation standard for determining whether symptoms and signs have sufficient consistency, and can be used to screen symptom regions with insufficient sign support. For example, the sub-region correlation threshold can be a strong correlation threshold, a weak correlation threshold, a clinically acceptable correlation threshold, etc.

[0063] An abnormally correlated sub-region of physical signs can be a sub-region whose physical sign correlation value is less than the sub-region correlation threshold. It can be used to identify areas where symptoms lack physical sign support or are contradictory. In one specific embodiment, an abnormally correlated sub-region of physical signs can include areas of emotional abnormality without pulse support, areas of digestive discomfort with inconsistent tongue appearance, etc. A normally correlated sub-region of physical signs can be a sub-region whose physical sign correlation value is greater than or equal to the sub-region correlation threshold. It can be used to identify areas where symptoms and signs are consistent and highly reliable. Comparing the physical sign correlation value of a sub-region with the sub-region correlation threshold can be a threshold judgment performed on the correlation value of each sub-region. Further, this operation can be implemented by using a sigmoid function to smooth the threshold transition or by using a hard threshold for binary classification, thereby assessing the consistency between symptoms and signs. If the physical sign correlation value of a sub-region is less than the sub-region correlation threshold, the sub-region is marked as an abnormally correlated sub-region of physical signs; this can be done when the correlation is insufficient. In an exemplary embodiment, this operation can be implemented by generating a mask of abnormal physical signs or recording a log of low-correlation sub-regions, thereby identifying symptom areas lacking objective support.

[0064] If the correlation value of a sub-region's vital signs is greater than or equal to the correlation threshold of that sub-region, then that sub-region is marked as a sub-region with normal vital sign correlation. This can be achieved when the correlation meets the threshold, thus confirming a symptom-sign synergistic region. Overlapping abnormal sub-regions can be dialectical sub-regions that are simultaneously marked as symptom-abnormal sub-regions and vital sign-correlation-abnormal sub-regions, and can be used to represent secondary pathological regions with high confidence. For example, overlapping abnormal sub-regions can include areas of dual abnormal emotions, symptom-sign disconnect, unreliable complaints, etc. The total number of abnormal regions can be the sum of the number of all overlapping abnormal sub-regions, and can be used to quantify the severity of the overall abnormality of the secondary symptom. In a specific embodiment, the total number of abnormal regions can be expressed as a cumulative number of overlapping abnormalities, a weighted abnormality count, etc.

[0065] The total number of syndrome differentiation sub-regions can be the total number of all syndrome differentiation sub-regions divided in the current syndrome differentiation task, and can be used as the normalized denominator for calculating the secondary symptom offset component. Obtaining symptom-abnormal sub-regions involves spatially overlapping them with sign-related abnormal sub-regions to obtain overlapping abnormal sub-regions. This can be achieved by performing an intersection operation on two abnormal sets. Further, this operation can be implemented by finding the intersection of sets based on sub-region IDs or by using a bitmap mask bitwise AND operation, thereby filtering out high-confidence problem areas with dual abnormalities. Obtaining the number of overlapping abnormal sub-regions involves summing the numbers of overlapping abnormal sub-regions to obtain the total number of abnormal regions. This can be achieved by counting the number of overlapping abnormal sub-regions. In an exemplary embodiment, this operation can be implemented by directly calling the len() function to count or by summing the overlap mask, thereby quantifying the overall scale of secondary symptom abnormalities. The ratio of the total number of abnormal regions to the total number of syndrome differentiation sub-regions is calculated to obtain the secondary symptom offset component, which can be achieved by performing a division operation: total number of abnormal regions ÷ total number of syndrome differentiation sub-regions. For example, this operation can be achieved by directly outputting the ratio value as a scalar component or mapping it to a preset clinical level and then encoding it into a vector, thereby generating a standardized subsymptom abnormality ratio in the [0, 1] interval for vector synthesis.

[0066] Taking the secondary symptom reliability assessment of a patient with anxiety and premenstrual breast tenderness as an example, the AI ​​intelligent syndrome differentiation and acupoint positioning assistance system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility in this embodiment can be as follows: The system divides the symptom area into 5 syndrome differentiation sub-regions: emotion, breast, sleep, digestion, and physical strength. The patient reports "severe anxiety" (symptom deviation value of the emotion sub-region = 0.85), but the pulse is normal, the tongue is pale red with a thin white coating, and the emotion-tongue-pulse correlation value is only 0.3 (below the correlation threshold of 0.6), so this sub-region is marked as symptom abnormal and sign correlation abnormal; while the symptom deviation value of the "breast tenderness" sub-region = 0.78 (exceeding the allowable value of 0.7), and is highly correlated with the wiry pulse (correlation value = 0.82 > 0.6), it is only marked as symptom abnormal and is not included in the sign abnormality set. The remaining sub-regions are normal. There is only 1 overlapping abnormal sub-region in the emotion area. The total number of abnormal regions = 1, the total number of syndrome differentiation sub-regions = 5, and the secondary symptom offset component = 1 / 5 = 0.2. Although the patient's subjective anxiety was strong, its contribution to the overall deviation was suppressed due to the lack of physical signs. This prevented the selection of acupoints from being overly biased towards soothing the liver and relieving depression, thus ignoring the actual breast distension and pain (although there was no double abnormality, it still needs attention).

[0067] In one embodiment, the sub-region symptom deviation value is obtained as follows: The diagnostic data is divided into regions to obtain diagnostic sub-region data. Based on the diagnostic sub-region data, the actual quantitative value of symptoms in each diagnostic sub-region is extracted and marked as the sub-region actual quantitative value. The difference between the sub-region actual quantitative value and the sub-region standard quantitative value is calculated to obtain the sub-region symptom deviation value. The sub-region data can be a structured set of symptoms and signs categorized into different sub-regions based on TCM syndrome logic or anatomical function. This data can serve as the basic unit for extracting the measured quantitative values ​​and correlations of sub-regions. In an exemplary embodiment, the sub-region data may include, but is not limited to, one or more of the following: emotional sub-region data, reproductive endocrine sub-region data, and spleen and stomach digestive sub-region data. The measured quantitative values ​​of a sub-region can be numerical indicators of symptom severity or intensity extracted from the actual collected syndrome data within a specific sub-region. These values ​​can reflect the patient's true symptom level in that sub-region. For example, the measured quantitative values ​​of a sub-region may include, but are not limited to, anxiety scores, breast distension frequency, and insomnia duration. The standardized quantitative values ​​of a sub-region can be reference values ​​representing the expected symptom intensity of the corresponding sub-region in the standard syndrome model. These values ​​can be used as a benchmark for calculating symptom deviation. Furthermore, the standardized quantitative values ​​of a sub-region may include, but are not limited to, the standardized values ​​for the emotional sub-region of liver stagnation syndrome, the standardized values ​​for the waist and knee sub-region of kidney deficiency syndrome, and the standardized values ​​for the digestive sub-region of phlegm-dampness syndrome.

[0068] The sub-region symptom deviation value can be the difference between the measured quantified value and the standardized quantified value of a sub-region, representing the degree to which local symptoms deviate from the standard syndrome type, and can be used to identify sub-regions with abnormal symptoms. In a specific embodiment, the sub-region symptom deviation value can be obtained by subtraction (measured minus standard) or by calculating the absolute difference. Dividing the syndrome differentiation data into regions to obtain syndrome differentiation sub-region data can be achieved by allocating the overall syndrome differentiation data to each syndrome differentiation sub-region according to preset syndrome zoning rules. Furthermore, dividing the syndrome differentiation data into regions to obtain syndrome differentiation sub-region data can be achieved by rule mapping based on a TCM syndrome type knowledge graph or by automatically dividing regions based on symptom co-occurrence patterns using clustering algorithms, thereby realizing the spatial structured organization of symptoms and signs.

[0069] Based on the data from the sub-regions of syndrome differentiation, the actual quantitative values ​​of symptoms within each sub-region are extracted and labeled as the sub-region measured values. This can be achieved by parsing and numerically quantifying the intensity of symptom manifestations from the sub-region data. Further, this operation can be implemented by directly reading scoring values ​​from structured questionnaire fields or by extracting and mapping them to a 0–1 quantification range from free text using natural language processing techniques, thereby generating comparable local symptom indicators. The difference between the measured values ​​and the standard quantification values ​​of the sub-regions is calculated to obtain the sub-region symptom deviation value. This can be achieved by performing an arithmetic operation of subtracting the standard value from the measured value. In an exemplary embodiment, this operation can be achieved by calculating the absolute difference |measured − standard| or by retaining the sign to distinguish between positive and negative deviations, thereby quantifying the degree to which local symptoms deviate from the standard syndrome type. The method for obtaining the sub-regional vital sign correlation value is as follows: the vital sign correlation detection is performed on the syndrome differentiation data, the syndrome differentiation data is divided into several syndrome differentiation sub-regions, the vital sign correlation degree of each syndrome differentiation sub-region is obtained and marked as the sub-region measured correlation degree, and the difference between the sub-region measured correlation degree and the sub-region standard correlation degree is calculated to obtain the sub-regional vital sign correlation value.

[0070] Physical sign correlation detection can be an analytical process that assesses the semantic or numerical consistency between symptom descriptions and objective physical signs such as tongue appearance and pulse. It can be used to generate sub-region measured correlation degrees to support the collaborative judgment of physical signs and symptoms. For example, physical sign correlation detection may include, but is not limited to, tongue-pulse-symptom matching detection, multimodal consistency scoring, and clinical logic verification. The measured correlation degree of a sub-region can be a quantitative value of the degree of consistency actually observed between symptoms and corresponding physical signs (such as tongue appearance and pulse) within a specific syndrome differentiation sub-region. It can be used to measure whether symptoms are supported by objective physical signs. In a specific embodiment, the measured correlation degree of a sub-region may include, but is not limited to, the correlation degree between emotion and wiry pulse, the correlation degree between tongue coating and digestive symptoms, and the correlation degree between basal body temperature and menstruation. The standard correlation degree of a sub-region can be a reference value of the typical correlation strength that symptoms and physical signs should possess within a specific syndrome differentiation sub-region in a standard syndrome model. It can be used as a benchmark for judging abnormal correlation of physical signs. Furthermore, the standard correlation degree of a sub-region may include, but is not limited to, the standard correlation degree between emotion and wiry pulse in liver stagnation syndrome, and the standard correlation degree between aversion to cold and deep and slow pulse in kidney yang deficiency syndrome.

[0071] The sub-regional sign association value can be the difference between the measured association degree and the standard association degree of the sub-region, representing the degree of deviation in symptom-sign synergy, and can be used to identify sub-regions with abnormal sign association. In an exemplary embodiment, the sub-regional sign association value can be obtained by calculating the difference (measured minus standard), with negative values ​​indicating insufficient association. Sign association detection on diagnostic data can involve analyzing the semantic or numerical matching relationship between symptoms and signs such as tongue and pulse. Further, this operation can be achieved by calculating similarity using a pre-trained multimodal alignment model or by performing logical consistency checks based on a clinical rule base, thereby generating a basis for symptom-sign consistency assessment. Dividing the diagnostic data into several diagnostic sub-regions can be achieved by repeatedly applying the region division logic to ensure that sign association analysis and symptom analysis use the same sub-region structure. Exemplarily, this operation can be achieved by reusing the sub-region division results of previous steps or by performing it independently while maintaining logical consistency in partitioning, thereby ensuring that symptoms and signs are compared within the same spatial unit.

[0072] Obtaining the correlation degree of physical signs within each diagnostic sub-region, labeled as the measured correlation degree of the sub-region, can be achieved by calculating the correlation score between symptoms and corresponding physical signs within each sub-region. In a specific embodiment, this operation can be implemented by calculating the Pearson correlation coefficient or using the correlation strength normalized by attention weights, thereby quantifying the local symptom-sign synergy level. The difference between the measured correlation degree and the standard correlation degree of the sub-region is calculated to obtain the physical sign correlation value of the sub-region, which can be achieved by subtracting the standard correlation degree from the measured correlation degree. Further, this operation can be implemented by taking the negative value of the difference as the abnormality degree (the smaller the value, the more abnormal) or by setting a threshold to truncate and highlight significant deviations, thereby identifying abnormal regions with insufficient or excessive physical sign support.

[0073] For example, in the scenario of assessing the consistency of secondary symptoms and signs in patients with liver stagnation and spleen deficiency syndrome, the AI-powered intelligent syndrome differentiation and acupoint positioning assistance system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility in this embodiment can be as follows: The system divides the syndrome differentiation data into three sub-regions: "emotion," "digestion," and "physical strength." In the "emotion" sub-region, the patient's self-rated anxiety score is 0.85 (measured value of the sub-region), while the standard value for liver stagnation syndrome is 0.75, with a deviation of 0.10; simultaneously, the pulse is wiry, the measured correlation between emotion and pulse is 0.82, the standard correlation is 0.80, and the sign correlation is +0.02, indicating that the symptoms and signs are consistent. In the "digestion" sub-region, the patient reports abdominal distension of 0.60, the standard value is 0.55, and the deviation is 0.05; however, the tongue coating is thin and white (not greasy), the measured correlation is only 0.30, the standard correlation is 0.70, and the sign correlation is -0.40, suggesting that the symptoms lack physical evidence support. In subsequent calculations of secondary symptom deviation components, only the "digestion" sub-region is included in the overlapping abnormal area because it simultaneously meets the conditions of exceeding the symptom deviation limit (assuming an allowable value of 0.04) and having a sign correlation value below the threshold (e.g., <0). This avoids misjudging genuine liver stagnation emotions as unreliable complaints and also suppresses the interference of isolated digestive symptoms on the overall syndrome differentiation.

[0074] In one embodiment, the physical fitness compensation value is obtained as follows: Obtain the quantitative values ​​of symptom characteristics and the interval values ​​of symptom changes within the effective diagnostic period. Calculate the ratio between the quantitative values ​​of symptom characteristics and the interval values ​​of symptom changes to obtain the symptom change rate value. Sum the symptom change rate values ​​within each effective diagnostic period and take the average value to obtain the constitution compensation value.

[0075] The symptom change rate value can be the ratio of the quantified value of symptom characteristics to the symptom change interval value within a single effective diagnostic period. It can reflect an individual's response speed to pathological stimuli and serve as a dynamic representation of constitutional differences. In this embodiment, the symptom change rate value can be obtained by dividing the quantified value of symptom characteristics by the symptom change interval value, with units of "intensity / day". Furthermore, the symptom change rate value can be a basic parameter for synergistic effects with other subjects, used to generate subsequent constitutional compensation values. The constitutional compensation value can be the summation and averaging of the symptom change rate values ​​across all effective diagnostic periods. It can be used to introduce individualized constitutional factors in diagnostic correction, improving the adaptability of diagnostic conclusions. In an exemplary embodiment, the constitutional compensation value is obtained by arithmetically averaging the rate values ​​across multiple effective diagnostic periods. In a specific embodiment, the constitutional compensation value serves as a quantitative indicator of the influence of individual constitution on the evolution of syndromes, forming a hierarchical data relationship with the symptom change rate value.

[0076] Obtaining the quantified values ​​of symptom characteristics and symptom change intervals within the effective diagnostic period can be achieved by extracting previously calculated quantified values ​​and symptom change intervals from the marked effective diagnostic periods. Furthermore, this operation can be implemented by directly querying the historical calculation cache or re-extracting extreme points from the original time-series curve and recalculating, thereby ensuring that the rate calculation is based on highly reliable clinical data. The symptom change rate value is obtained by calculating the ratio of the quantified values ​​of symptom characteristics and symptom change intervals, which can be achieved by performing a division operation: quantified value of symptom characteristics ÷ symptom change interval value. In a specific embodiment, this operation can be achieved by setting the interval value to a maximum value or skipping the period if it is 0, or by normalizing the quantified values ​​and interval values ​​before calculating the ratio, thereby generating a rate index reflecting the dynamic characteristics of disease progression. The constitution compensation value is obtained by summing and averaging the symptom change rate values ​​within each effective diagnostic period, which can be achieved by calculating the arithmetic mean of the rate values ​​over all effective periods. For example, this operation can be improved by using a weighted average (weighted by the length of the effective time period) or by averaging after removing outlier rate values, thereby stably capturing individual-specific disease response patterns and forming a body-weighted representation.

[0077] For example, in a scenario comparing the dynamic responses of patients with Yin deficiency and Yang deficiency constitutions, the AI-powered intelligent syndrome differentiation and acupoint positioning assistance system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility in this embodiment can be as follows: Patient A (Yin deficiency and Yang deficiency constitution) has the following symptom characteristic quantification values ​​in two effective syndrome differentiation periods: Period 1: symptom characteristic quantification value = 0.9, interval = 2 days, rate = 0.45; Period 2: quantification value = 0.8, interval = 3 days, rate ≈ 0.27. The mean constitution compensation value = 0.36. Patient B (Yang deficiency constitution) has the following symptom characteristics in the same period: Period 1: quantification value = 0.6, interval = 6 days, rate = 0.10; Period 2: quantification value = 0.5, interval = 8 days, rate = 0.0625, mean = 0.081. Based on this, the system assigns a higher constitution compensation value to A, indicating that their syndrome is prone to acute symptoms and requires clearing and moistening; while B has a lower compensation value, indicating that the disease is slow and requires warming and tonifying. In the syndrome differentiation and correction module, even if the initial syndrome differentiation of both is "postmenstrual period", the comprehensive syndrome differentiation and correction value of A will shift towards "deficient heat" and B will be strengthened towards "yang deficiency and cold coagulation". The final acupoints will focus on Taixi, Zhaohai and Guanyuan, Mingmen respectively, to achieve personalized treatment that is adapted to the constitution.

[0078] In one embodiment, the method for obtaining the certificate type conversion coefficient is as follows: Based on the symptom synchronization coefficient corresponding to the effective differentiation period, it is marked as the effective synchronization coefficient. The difference between all effective synchronization coefficients in the tracking period and the standard value of the synchronization coefficient is calculated to obtain the synchronization coefficient deviation value. The variance of the synchronization coefficient deviation value is calculated to obtain the syndrome conversion coefficient.

[0079] The effective synchronization coefficient can be the symptom synchronization coefficient obtained within the effective diagnostic period, representing a clinically reliable indicator of symptom time-intensity synergy, and can be used as high-quality input data for syndrome stability analysis. In this embodiment, the effective synchronization coefficient can be extracted from the corresponding calculated symptom synchronization coefficient within the marked effective diagnostic period. The standard value of the synchronization coefficient can be a reference benchmark value of the symptom synchronization coefficient derived from the statistical analysis of clinical data of typical stable syndromes, and can be used to measure whether the patient's synchronicity deviates from the standard syndrome pattern. For example, the standard value of the synchronization coefficient can be the standard value of the synchronization coefficient for kidney deficiency syndrome, liver stagnation syndrome, phlegm-dampness syndrome, etc. The synchronization coefficient deviation value can be a sequence of differences between the effective synchronization coefficient and the standard value of the synchronization coefficient, reflecting the degree of deviation of synchronicity within each effective period, and can be used to characterize the local fluctuation characteristics of the syndrome in the time dimension. In an exemplary embodiment, the synchronization coefficient deviation value can be obtained by performing a subtraction operation on each effective synchronization coefficient: effective synchronization coefficient − standard value of synchronization coefficient. Further, the synchronization coefficient deviation value can include, but is not limited to, one or more of the following: positive deviation value, negative deviation value, zero-centered deviation value, etc.

[0080] The syndrome transformation coefficient can be the variance of the synchronization coefficient deviation value sequence, used to quantify the degree of dispersion of the syndrome around the standard pattern within the tracking period. It can reflect the strength and instability of the syndrome evolution trend; a larger value indicates a higher probability of syndrome transformation. In one embodiment, the syndrome transformation coefficient is obtained by performing statistical variance calculation on the synchronization coefficient deviation value. Based on the symptom synchronization coefficient corresponding to the effective syndrome differentiation period, it is marked as the effective synchronization coefficient, which can be extracted from the corresponding calculated symptom synchronization coefficient of the marked effective syndrome differentiation period. Furthermore, this operation can be implemented by associating the synchronization coefficient table with the period ID index or by synchronously caching the synchronization coefficient value when the effective period is marked, thereby ensuring that subsequent analysis is based only on high-confidence, low-noise synchronization data.

[0081] The synchronization coefficient deviation value is obtained by calculating the difference between all effective synchronization coefficients and the standard value of the synchronization coefficient within the tracking period. This can be achieved by subtracting each effective synchronization coefficient: effective synchronization coefficient - standard value of synchronization coefficient. Furthermore, this operation can be implemented by calculating the signed deviation to retain directional information or by taking the absolute value for subsequent variance calculation, thereby generating a deviation sequence reflecting local synchronization deviations from the standard. The variance of the synchronization coefficient deviation value is then calculated to obtain the pattern transformation coefficient. This can be achieved by applying a statistical variance formula (such as sample variance or population variance) to the synchronization coefficient deviation value sequence. Furthermore, this operation can be implemented using unbiased sample variance (n-1 denominator) or by employing a sliding window variance to capture the phased transformation trend, thereby quantifying the instability of the pattern during the dynamic observation period and serving as a criterion for whether the pattern is transforming.

[0082] For example, in the early identification of the transformation from kidney yang deficiency to yin-yang deficiency, the AI ​​intelligent syndrome differentiation and acupoint positioning assistance system for the thirteen acupuncture points for regulating menstruation and promoting fertility in this embodiment can be as follows: the patient's initial diagnosis is kidney yang deficiency, and the standard value of the synchronization coefficient is set to 0.5. During the 28-day tracking period, the system identifies three effective syndrome differentiation periods, with corresponding effective synchronization coefficients of 0.48, 0.35, and 0.22, respectively. The calculated deviation value sequence is: [−0.02, −0.15, −0.28]. The variance is 0.018, which is significantly higher than the typical variance of stable kidney yang deficiency patients (<0.005), so the syndrome conversion coefficient is high. This result suggests a continuous decrease in symptom synchronicity (such as a persistent monophasic basal body temperature but reduced aversion to cold, and the appearance of signs of yin deficiency such as dry mouth). Based on this, the system introduces the conversion path weight of "kidney yang deficiency → yin-yang deficiency" into the syndrome differentiation correction module and adds yin-nourishing acupoints (such as Taixi) in advance to achieve pre-intervention.

[0083] In one embodiment, the verification method for the comprehensive dialectical correction value is as follows: Obtain the constitution compensation value and syndrome conversion coefficient within the effective diagnosis period, calculate the ratio between the constitution compensation value and the syndrome conversion coefficient to obtain the correction verification coefficient, and confirm the validity of the comprehensive diagnosis correction value if the correction verification coefficient is within the preset verification range.

[0084] The correction verification coefficient can be the ratio of the constitution compensation value to the syndrome transformation coefficient within the effective diagnosis period, used to measure the consistency and rationality of the diagnosis correction logic. In this embodiment, the correction verification coefficient can serve as a criterion for the validity of the comprehensive diagnosis correction value, preventing over-correction or logical conflicts. Furthermore, the correction verification coefficient can be generated through ratio calculation, reflecting the degree of coordination between individual constitution and syndrome evolution. For example, the correction verification coefficient can be calculated by dividing the constitution compensation value by the syndrome transformation coefficient; when the denominator is close to zero, it can be set as an upper limit to maintain numerical stability. The preset verification range can be a reasonable range of the correction verification coefficient set based on clinical data statistics or traditional Chinese medicine theory, used to determine whether the diagnosis correction is credible. In an exemplary embodiment, the preset verification range can provide an objective threshold to screen effective correction results, ensuring the clinical rationality of the diagnosis logic. Furthermore, the preset verification range can include, but is not limited to, one or more of the following: low coordination tolerance range, medium coordination standard range, and high coordination preferred range.

[0085] Obtaining the constitution compensation value and syndrome conversion coefficient within the valid diagnosis period can be achieved by extracting the corresponding constitution compensation value and syndrome conversion coefficient from the diagnosis periods that have been marked as valid. Furthermore, this operation can be implemented by directly calling the correction factor associated with the valid periods marked in the symptom tracking module, thereby ensuring that the data used for verification comes from a stable and reliable diagnosis window, improving verification accuracy. In a specific embodiment, the average of constitution and syndrome parameters from multiple valid periods can also be used for verification. The ratio of the constitution compensation value to the syndrome conversion coefficient is calculated to obtain the corrected verification coefficient, which can be achieved by performing a division operation between the two to generate a single numerical index. Further, this operation can be achieved by dividing the constitution compensation value by the syndrome conversion coefficient; if the denominator is close to zero, it is set as an upper limit value, or a logarithmic ratio is used to compress the dynamic range, thereby quantifying the matching degree between the constitution basis and the dynamics of the syndrome and identifying potential logical conflicts. If the corrected verification coefficient is within the preset verification range, the comprehensive diagnosis correction value is confirmed to be valid. This can be achieved by performing an interval inclusion judgment between the corrected verification coefficient and the preset verification range; if the condition is met, the correction value is marked as valid. Furthermore, this operation can be implemented by setting a Boolean flag to indicate validity, or by writing the verification result into the audit log and triggering the acupoint optimization process or the re-evaluation process, thereby introducing a quality control mechanism to ensure the clinical credibility and safety of the syndrome differentiation correction results.

[0086] For example, in a scenario where a patient with Yang deficiency constitution shows a tendency towards liver stagnation transforming into fire, the AI-powered intelligent syndrome differentiation and acupoint positioning assistance system for the Thirteen Acupuncture Points for regulating menstruation and promoting fertility in this embodiment could be as follows: The patient is initially diagnosed with kidney Yang deficiency syndrome, with a constitution compensation value of +0.6 (Yang deficiency tendency). In the third week of treatment, the patient experiences increased irritability, bitter taste in the mouth, and premenstrual breast tenderness, and the system calculates a syndrome transformation coefficient of -0.5 (evolution towards liver stagnation transforming into fire). Within the effective syndrome differentiation period, the correction verification coefficient = 0.6 / |−0.5| = 1.2. The preset verification range is [0.3, 1.0]. Because 1.2 exceeds the upper limit, the system determines that the current constitution and the direction of syndrome evolution are inconsistent (it is illogical for a Yang deficiency constitution to rapidly transform into a heat syndrome in a short period of time), rejects the comprehensive syndrome differentiation correction value, and triggers a re-collection or expert review process. This avoids misjudging a fundamental change in syndrome due to short-term emotional fluctuations and ensures the robustness of syndrome differentiation.

[0087] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. An AI-powered intelligent diagnosis and acupoint location assistance system for thirteen acupuncture points used to regulate menstruation and promote fertility, characterized in that, The system includes: The four diagnostic methods data acquisition module is used to collect information from patients using the four diagnostic methods of traditional Chinese medicine, and to obtain diagnostic data including symptoms and signs. The diagnostic data includes pulse characteristics, tongue characteristics and symptom distribution information. Initial diagnostic parameters are obtained by identifying the feature distribution information. The dialectical comparison module is used to compare the initial dialectical parameters with the standard syndrome parameter range based on the initial dialectical parameters. If the initial dialectical parameters exceed the standard syndrome parameter range, a dialectical deviation signal is generated. The symptom tracking module is used to track the dynamic changes of symptoms in the diagnostic data based on the diagnostic deviation signal. Within the tracking period, the symptom change sequence is obtained, the symptom deviation degree is acquired, and it is compared with the deviation degree threshold to generate a diagnostic influence level signal. If the symptom deviation degree is greater than or equal to the deviation threshold, a high-level diagnostic signal is generated. The syndrome differentiation correction module is used to obtain the syndrome type conversion coefficient and the constitution compensation value based on the high-level syndrome differentiation signal, and to calculate the comprehensive syndrome differentiation correction value by superimposing the syndrome type conversion coefficient and the constitution compensation value. The acupoint optimization module is used to obtain dynamic acupoint coefficients based on comprehensive syndrome differentiation correction values. It integrates standard acupoint parameters with dynamic acupoint coefficients to calculate the required acupoint parameters for adjustment, thereby optimizing the positioning of the thirteen acupoints for regulating menstruation and promoting fertility. The dynamic acupoint coefficients are obtained by weighting and summing all symptom deviations and taking the average value.

2. The AI-powered intelligent diagnosis and acupoint location assistance system for the thirteen acupuncture points used to regulate menstruation and promote fertility as described in claim 1, characterized in that, The initial dialectical parameters are obtained as follows: The symptom regions in the syndrome differentiation data are divided into several syndrome differentiation sub-regions. The number of symptom features in each syndrome differentiation sub-region is obtained. The number of symptom features in each syndrome differentiation sub-region is weighted, summed, and averaged to obtain the symptom feature density value. The initial diagnostic parameters are obtained by calculating the ratio between the symptom feature density value and the standard syndrome density value.

3. The AI-powered intelligent diagnosis and acupoint location assistance system for the thirteen acupuncture points used to regulate menstruation and promote fertility as described in claim 2, characterized in that, The method for obtaining the symptom deviation degree is as follows: Obtain the primary symptom offset component and the secondary symptom offset component, and then perform vector synthesis on the primary symptom offset component and the secondary symptom offset component to obtain the symptom deviation degree.

4. The AI-powered intelligent diagnosis and acupoint positioning assistance system for the thirteen acupuncture points used to regulate menstruation and promote fertility as described in claim 3, characterized in that, The method for obtaining the main symptom offset component is as follows: The ratio of the symptom change interval value and the symptom characteristic quantification value within the same tracking period is calculated to obtain the symptom synchronization coefficient. Periods with symptom synchronization coefficients that meet preset conditions are selected and marked as valid diagnosis periods. The duration of the effective diagnostic period is obtained, and the ratio of the duration of the effective diagnostic period to the total duration of the tracking period is calculated to obtain the main symptom offset component.

5. The AI-powered intelligent diagnosis and acupoint location assistance system for the thirteen acupuncture points used to regulate menstruation and promote fertility as described in claim 4, characterized in that, The symptom change interval value and symptom feature quantification value are obtained in the following way: Within the tracking period, the tracking period is divided into several time segments. A time-symptom coordinate system is established, with the horizontal dimension representing the time segment and the vertical dimension representing the symptom feature quantification value corresponding to each time segment. The obtained symptom feature quantification value is substituted into the time-symptom coordinate system to draw the symptom change curve. Based on the symptom change curve, the peak points and valley points of symptoms are obtained. The difference between the coordinates of adjacent peak points and valley points is calculated to obtain the quantitative value of symptom characteristics. Within the tracking period, the tracking period is divided into several time segments, and a time-symptom change coordinate system is established. The horizontal dimension represents the time segment, and the vertical dimension represents the frequency of symptom changes corresponding to each time segment. The obtained symptom change frequency is substituted into the time-symptom change coordinate system to draw the symptom change frequency curve. Based on the symptom change frequency curve, the points of increasing and decreasing frequency of change are obtained. The difference between the coordinates of adjacent points of increasing and decreasing frequency of change is calculated to obtain the symptom change interval value.

6. The AI-powered intelligent diagnosis and acupoint positioning assistance system for the thirteen acupuncture points used to regulate menstruation and promote fertility as described in claim 5, characterized in that, The method for obtaining the secondary symptom offset component is as follows: The symptom deviation value of a sub-region is compared with the allowable deviation value of the sub-region. If the symptom deviation value of a sub-region is greater than or equal to the allowable deviation value of the sub-region, the sub-region is marked as a symptom abnormal sub-region. If the symptom deviation value of a sub-region is less than the allowable deviation value of the sub-region, the sub-region is marked as a symptom normal sub-region. The correlation value of the sub-region's vital signs is compared with the correlation threshold of the sub-region. If the correlation value of the sub-region's vital signs is less than the correlation threshold of the sub-region, the sub-region is marked as a sub-region with abnormal vital signs correlation. If the correlation value of the sub-region's vital signs is greater than or equal to the correlation threshold of the sub-region, the sub-region is marked as a sub-region with normal vital signs correlation. Obtain the symptom abnormal sub-regions, perform spatial overlap analysis on the symptom abnormal sub-regions and the sign-related abnormal sub-regions to obtain overlapping abnormal sub-regions, obtain the number of overlapping abnormal sub-regions, sum the number of overlapping abnormal sub-regions to obtain the total number of abnormal regions, calculate the ratio of the total number of abnormal regions to the total number of syndrome differentiation sub-regions to obtain the secondary symptom offset component.

7. The AI-powered intelligent diagnosis and acupoint location assistance system for the thirteen acupuncture points used to regulate menstruation and promote fertility as described in claim 6, characterized in that, The method for obtaining the sub-region symptom deviation value is as follows: The diagnostic data is divided into regions to obtain diagnostic sub-region data. Based on the diagnostic sub-region data, the actual quantitative value of symptoms in each diagnostic sub-region is extracted and marked as the sub-region actual quantitative value. The difference between the sub-region actual quantitative value and the sub-region standard quantitative value is calculated to obtain the sub-region symptom deviation value. The method for obtaining the sub-regional vital sign correlation value is as follows: the vital sign correlation detection is performed on the syndrome differentiation data, the syndrome differentiation data is divided into several syndrome differentiation sub-regions, the vital sign correlation degree of each syndrome differentiation sub-region is obtained and marked as the sub-region measured correlation degree, and the difference between the sub-region measured correlation degree and the sub-region standard correlation degree is calculated to obtain the sub-regional vital sign correlation value.

8. The AI-powered intelligent diagnosis and acupoint location assistance system for the thirteen acupuncture points used to regulate menstruation and promote fertility as described in claim 7, characterized in that, The method for obtaining the physical compensation value is as follows: Obtain the quantitative values ​​of symptom characteristics and the interval values ​​of symptom changes within the effective diagnostic period. Calculate the ratio between the quantitative values ​​of symptom characteristics and the interval values ​​of symptom changes to obtain the symptom change rate value. Sum the symptom change rate values ​​within each effective diagnostic period and take the average value to obtain the constitution compensation value.

9. The AI-powered intelligent diagnosis and acupoint positioning assistance system for the thirteen acupuncture points used to regulate menstruation and promote fertility as described in claim 7, characterized in that, The method for obtaining the certificate type conversion coefficient is as follows: Based on the symptom synchronization coefficient corresponding to the effective differentiation period, it is marked as the effective synchronization coefficient. The difference between all effective synchronization coefficients in the tracking period and the standard value of the synchronization coefficient is calculated to obtain the synchronization coefficient deviation value. The variance of the synchronization coefficient deviation value is calculated to obtain the syndrome conversion coefficient.

10. The AI-powered intelligent diagnosis and acupoint location assistance system for the thirteen acupuncture points used to regulate menstruation and promote fertility as described in claim 4, characterized in that, The verification method for the comprehensive dialectical correction value is as follows: Obtain the constitution compensation value and syndrome conversion coefficient within the effective diagnosis period, calculate the ratio between the constitution compensation value and the syndrome conversion coefficient to obtain the correction verification coefficient, and confirm the validity of the comprehensive diagnosis correction value if the correction verification coefficient is within the preset verification range.

Images