Traditional Chinese medicine bonesetting technique curative effect evaluation method and system based on gait analysis

Abstract

The embodiment of the invention discloses a traditional Chinese medicine bonesetting technique curative effect evaluation method and system based on gait analysis, the system comprises a data acquisition module, a feature mapping module and a curative effect evaluation module, the method comprises the following steps: using the data acquisition module to acquire gait meridian data of a current patient and preprocessing the gait meridian data to obtain preprocessed gait meridian data; performing feature extraction on the preprocessed gait meridian and collateral data by using a feature mapping module to obtain meridian and collateral features and gait features; determining a gait meridian association weight matrix between the meridian features and the gait features through a convolutional neural network based on a self-attention mechanism, and obtaining a meridian state quantized value of the current patient; and calculating a current traditional Chinese medicine meridian improvement value and a current western medicine function recovery value based on the historical diagnosis and treatment information, the gait feature, the meridian state quantized value and the fusion weight of each sub-feature of the current patient by using a curative effect evaluation module, and generating a traditional Chinese medicine and western medicine two-dimensional curative effect evaluation result.

Description

Technical Field

[0001] This application relates to the field of efficacy evaluation of traditional Chinese medicine bone setting techniques, and includes, but is not limited to, a method and system for evaluating the efficacy of traditional Chinese medicine bone setting techniques based on gait analysis. Background Technology

[0002] Traditional Chinese medicine (TCM) bone-setting manipulation is a distinctive therapy for treating bone injuries. Through manual manipulation, it adjusts the patient's tendons and bones, restoring the flow of Qi and blood, thereby achieving treatment and functional rehabilitation. However, the efficacy evaluation system for TCM bone-setting manipulation has long faced significant challenges. Traditional assessments rely heavily on the physician's personal experience, making subjective judgments through observation, auscultation, inquiry, and palpation, lacking unified and objective quantitative standards. This results in evaluations of TCM bone-setting manipulation being significantly influenced by differences in physician experience, leading to low consistency and repeatability, making precise efficacy comparisons and treatment optimization difficult, and hindering the standardized development and promotion of TCM bone-setting manipulation.

[0003] With the development of sensor technology and data analysis methods, technologies have gradually emerged that can objectively assess the function or physiological state of a patient's musculoskeletal system. On the one hand, based on the modern Western medical theoretical system, patients are objectively assessed, and corresponding bone injury treatment plans are formulated or adjusted based on the assessment results. On the other hand, based on the traditional Chinese medicine theoretical system, the patient's Qi and blood circulation and meridian function are assessed, and the assessment results guide the formulation and adjustment of the patient's bone injury treatment plan.

[0004] However, bone injury treatment assessments based on modern Western medical theories primarily focus on anatomical repair and local functional improvement, failing to comprehensively reflect the patient's overall functional status, neglecting individual patient differences, and unable to effectively track long-term functional recovery. Furthermore, bone injury treatment assessments based on traditional Chinese medicine theories lack objective quantitative standards for the state of meridians and qi and blood, and the assessment conclusions vary significantly among different physicians, resulting in low accuracy and consistency. Summary of the Invention

[0005] In view of this, the embodiments of this application provide a method and system for evaluating the efficacy of traditional Chinese medicine bone-setting techniques based on gait analysis, which at least solves the problem of inaccurate evaluation results of the efficacy of traditional Chinese medicine bone-setting techniques in the treatment of bone injuries.

[0006] The technical solution of this application embodiment is implemented as follows: In a first aspect, embodiments of this application provide a method for evaluating the efficacy of traditional Chinese medicine bone-setting manipulation based on gait analysis, applied to a system for evaluating the efficacy of traditional Chinese medicine bone-setting manipulation based on gait analysis. The system includes: a data acquisition module, a feature mapping module, and an efficacy evaluation module. The method includes: The data acquisition module is used to collect the current patient's gait meridian data; the gait meridian data is preprocessed to obtain preprocessed gait meridian data. Using the feature mapping module, feature extraction is performed on the preprocessed gait meridian data to obtain meridian features and gait features; using a convolutional neural network based on a self-attention mechanism, the gait meridian correlation weight matrix between the meridian features and the gait features is determined, and based on the meridian features, the gait features, and the gait meridian correlation weight matrix, the quantitative value of the current patient's meridian state is obtained; Using the efficacy evaluation module, the system obtains the patient's historical medical information, the fusion weights of each sub-feature in the meridian features and gait features; based on the historical medical information, the quantitative value of the meridian status, and the fusion weights of each sub-feature in the meridian features, it calculates the current TCM meridian improvement value; based on the historical medical information, the gait features, and the fusion weights of each sub-feature in the gait features, it calculates the current Western medicine functional recovery value; based on the current Western medicine functional recovery value and the current TCM meridian improvement value, it generates a dual-dimensional TCM and Western medicine efficacy evaluation result.

[0007] Secondly, embodiments of this application provide a traditional Chinese medicine bone-setting technique efficacy evaluation system based on gait analysis, including: a data acquisition module, a feature mapping module, and an efficacy evaluation module; The data acquisition module is used to collect the current patient's gait meridian data; and to preprocess the gait meridian data to obtain preprocessed gait meridian data. The feature mapping module is used to extract features from the preprocessed gait meridian data to obtain meridian features and gait features; through a convolutional neural network based on a self-attention mechanism, the gait meridian correlation weight matrix between the meridian features and the gait features is determined, and based on the meridian features, the gait features and the gait meridian correlation weight matrix, the quantitative value of the current patient's meridian state is obtained; The efficacy evaluation module is used to acquire the current patient's historical medical information, the fusion weights of each sub-feature in the meridian features and gait features; calculate the current TCM meridian improvement value based on the historical medical information, the quantified value of the meridian status, and the fusion weights of each sub-feature in the meridian features; calculate the current Western medicine functional recovery value based on the historical medical information, the gait features, and the fusion weights of each sub-feature in the gait features; and generate a dual-dimensional TCM and Western medicine efficacy evaluation result based on the current Western medicine functional recovery value and the current TCM meridian improvement value.

[0008] The beneficial effects of the technical solutions provided in this application include at least the following: This application collects gait and meridian data from current patients through a data acquisition module. This data includes both Traditional Chinese Medicine (TCM) and Western medicine (TCM) data, achieving data fusion at the data level. After feature extraction via a feature mapping module to obtain meridian and gait features, a convolutional neural network based on a self-attention mechanism is used to learn and quantify the correlation between these features. This transforms abstract meridian features into calculable quantitative values ​​of meridian state, establishing a correspondence between TCM-based meridian features and Western medicine-based gait features. The resulting quantitative values ​​of meridian state can objectively interpret TCM theory. In the efficacy evaluation module, a dynamic weighted fusion mechanism is used to calculate the current improvement value of TCM meridians and the current functional recovery value of Western medicine, and to generate a unified TCM and Western medicine dual-dimensional efficacy evaluation result. This allows the efficacy evaluation of patients to include both objective evidence of functional recovery from the Western medicine dimension and the internal Qi and blood harmony from the TCM dimension. The TCM and Western medicine dual-dimensional efficacy evaluation result provides physicians with accurate and comprehensive auxiliary treatment plans in clinical diagnosis, which helps to improve the accuracy of efficacy evaluation of patients' bone and joint diseases, shorten the patient's recovery period, and reduce the risk of medical accidents. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein: Figure 1 A flowchart illustrating a method for evaluating the efficacy of traditional Chinese medicine bone-setting manipulation based on gait analysis, provided in an embodiment of this application; Figure 2 This is a schematic diagram of a traditional Chinese medicine bone-setting technique efficacy evaluation system based on gait analysis, provided in an embodiment of this application. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. The following embodiments are used to illustrate this application, but are not intended to limit the scope of this application. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0011] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0012] It should be noted that the terms "first, second, and third" used in the embodiments of this application are merely to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first, second, and third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0013] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which the embodiments of this application pertain. It should also be understood that terms such as those defined in general dictionaries should be understood to have a meaning consistent with their meaning in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0014] This application provides a method for evaluating the efficacy of traditional Chinese medicine bone-setting techniques based on gait analysis. Figure 1 This application provides a flowchart illustrating a method for evaluating the efficacy of traditional Chinese medicine bone-setting techniques based on gait analysis, as shown in the embodiments below. Figure 1 As shown, the method includes at least the following steps: Step S110: Using the data acquisition module, collect the current patient's gait meridian data; preprocess the gait meridian data to obtain preprocessed gait meridian data; The data acquisition module is a hardware and software unit that collects the patient's current gait meridian data. Gait meridian data is multimodal data that reflects the patient's kinematics, dynamics, and the physiological electrical activity of meridian acupoints, captured by sensors and measuring devices.

[0015] Preprocessing includes filtering the gait meridian data to remove interference noise, aligning the filtered gait meridian data with timestamps to achieve time synchronization, and converting the timestamp-aligned data format into a unified format.

[0016] Step S120: Using the feature mapping module, feature extraction is performed on the preprocessed gait meridian data to obtain meridian features and gait features; using a convolutional neural network based on a self-attention mechanism, the gait meridian correlation weight matrix between the meridian features and the gait features is determined, and based on the meridian features, the gait features, and the gait meridian correlation weight matrix, the quantitative value of the current patient's meridian state is obtained; Preprocessed gait meridian data includes physiological electrical data, kinematic data, and kinetic data. Among these, the kinematic and kinetic data are related to the patient's gait from a Western medical perspective. Meridian characteristics are primarily derived from the patient's physiological electrical data, such as acupoint conductivity. Gait characteristics are primarily derived from kinematic and kinetic data, such as joint angles.

[0017] Self-attention mechanisms are neural network components that dynamically monitor the relationships between different features during the computation of a convolutional neural network. The gait meridian association weight matrix is ​​a mathematical table generated by a convolutional neural network based on the self-attention mechanism. The numerical values ​​of the elements in this table quantify the association strength between each pair of meridian features and gait features. The meridian state quantification value is an assessment result output by the convolutional neural network, representing the degree of smooth flow of Qi and blood in multiple core meridians in numerical form.

[0018] The feature mapping module transforms preprocessed gait meridian data into objective and repeatable quantitative values ​​of meridian status, clarifying the correlation between the quantitative values ​​of meridian status and gait characteristics. This provides key parameters for evaluating the efficacy of treatment in patients using integrated traditional Chinese and Western medicine. In traditional Chinese medicine theory, a patient's meridians are pathways for the flow of Qi and blood, connecting the internal organs and facilitating communication between the internal and external systems. This application is based on the principle that internal conditions inevitably manifest externally; that is, the state of Qi and blood in the internal meridians will be reflected in external gait, electromyography, and acupoint conductivity. Therefore, this method quantifies the patient's internal meridian status by analyzing externally measurable data such as gait characteristics.

[0019] Step S130: Using the efficacy evaluation module, obtain the current patient's historical medical information, the fusion weights of each sub-feature in the meridian features and gait features; calculate the current TCM meridian improvement value based on the historical medical information, the quantified value of the meridian status, and the fusion weights of each sub-feature in the meridian features; calculate the current Western medicine functional recovery value based on the historical medical information, the gait features, and the fusion weights of each sub-feature in the gait features; and generate a dual-dimensional TCM and Western medicine efficacy evaluation result based on the current Western medicine functional recovery value and the current TCM meridian improvement value.

[0020] Historical medical information includes the current patient's historical efficacy evaluation records and history of bone-setting manipulation treatment. The fusion weight of sub-features refers to assigning a corresponding weight to each sub-feature in the meridian features and gait features.

[0021] The TCM meridian improvement value is a result calculated by weighting the current and historical quantitative values ​​of the meridian status, which can quantify the degree of improvement in the patient's meridian status. When a patient first visits the clinic, the patient's meridian status is compared with the standard range of healthy individuals to generate a baseline status assessment value, which serves as the benchmark for subsequent efficacy comparisons.

[0022] The Western medicine functional recovery value is a result calculated by weighting the current gait characteristics with historical gait characteristics, which quantifies the degree of improvement in the patient's current motor function.

[0023] Based on the current Western medicine functional recovery value, the current TCM meridian improvement value, and the comprehensive efficacy score, a dual-dimensional efficacy evaluation result combining Western and TCM is generated, including data charts, grading, and core conclusions. This dual-dimensional efficacy evaluation result integrates the current Western medicine functional recovery value and the current TCM meridian improvement value, and simultaneously describes the treatment effect on the current patient from both TCM syndrome differentiation and Western medicine functional perspectives.

[0024] The dual-dimensional efficacy evaluation results of traditional Chinese and Western medicine overcome the limitations of efficacy evaluation based on a single Western medicine dimension or a single traditional Chinese medicine dimension. It presents the effects of the current patients on two levels after treatment with bone-setting techniques: functional recovery in the Western medicine dimension and regulation of the body's state in the traditional Chinese medicine dimension. This provides physicians with a comprehensive basis for efficacy evaluation for rapid clinical decision-making.

[0025] This application collects gait and meridian data from current patients through a data acquisition module. This data includes both Traditional Chinese Medicine (TCM) and Western medicine (TCM) data, achieving data fusion at the data level. After feature extraction via a feature mapping module to obtain meridian and gait features, a convolutional neural network based on a self-attention mechanism is used to learn and quantify the correlation between these features. This transforms abstract meridian features into calculable quantitative values ​​of meridian state, establishing a correspondence between TCM-based meridian features and Western medicine-based gait features. The resulting quantitative values ​​of meridian state can objectively interpret TCM theory. In the efficacy evaluation module, a dynamic weighted fusion mechanism is used to calculate the current improvement value of TCM meridians and the current functional recovery value of Western medicine, and to generate a unified TCM and Western medicine dual-dimensional efficacy evaluation result. This allows the efficacy evaluation of patients to include both objective evidence of functional recovery from the Western medicine dimension and the internal Qi and blood harmony from the TCM dimension. The TCM and Western medicine dual-dimensional efficacy evaluation result provides physicians with accurate and comprehensive auxiliary treatment plans in clinical diagnosis, which helps to improve the accuracy of efficacy evaluation of patients' bone and joint diseases, shorten the patient's recovery period, and reduce the risk of medical accidents.

[0026] In some embodiments, the "gait meridian data" in step S110 includes kinematic data, kinetic data, and electrophysiological data; in step S110, "collecting the current patient's gait meridian data using the data acquisition module" includes: Step S1101: Using the data acquisition module, inertial measurement unit sensors are attached to the patient's knee joint, ankle joint, and pelvis to collect kinematic data; the kinematic data includes joint angles, movement speed, and displacement. The inertial measurement unit (IMU) sensor is a 6-axis IMU sensor that acquires kinematic data in a three-dimensional coordinate system during the patient's current walking motion, with a sampling frequency of 100Hz. The joint angles acquired by the IMU sensor include knee flexion and knee extension angles.

[0027] The knee, ankle, and pelvic joints are the main focus areas for musculoskeletal differentiation in traditional Chinese medicine orthopedics. Pelvic tilt and rotation are key areas for assessing abnormal spinal and lower limb alignment. Knee joint range of motion reflects the condition of surrounding muscles, and ankle joint stability is fundamental to a steady gait. Therefore, by collecting kinematic data from the knee, ankle, and pelvic joints, one can accurately identify dynamic alignment abnormalities caused by bone misalignment and muscle displacement.

[0028] Step S1102: Using the data acquisition module, a foot pressure sensor array is attached to the corresponding meridian acupoint area on the patient's sole to collect the dynamic data; the dynamic data includes peak pressure, pressure distribution uniformity, and pressure center trajectory in each area of ​​the sole. The plantar pressure sensor is a flexible plantar pressure sensor that acquires dynamic data in a three-dimensional coordinate system during the patient's walking. The flexible plantar pressure sensor uses a 24-point acquisition method with a sampling frequency of 50Hz.

[0029] Step S1103: The physiological electrical data includes acupoint conductivity and electromyographic signals; using the data acquisition module, a meridian acupoint conductivity sensor is attached to key acupoints on the patient's core meridian to collect the acupoint conductivity, and electromyographic signals of muscles along the meridian are collected through surface electromyographic electrodes.

[0030] The meridian acupoint conductivity sensor is a 12-channel sensor that acquires physiological electrical data in a three-dimensional spatial coordinate system during the patient's walking, with a sampling frequency of 200Hz. The sensor is attached to 12 key acupoints along five core meridians, including the Stomach Meridian of Foot-Yangming and the Kidney Meridian of Foot-Shaoyin. These five core meridians correspond to different muscles along their respective lines; for example, the quadriceps femoris corresponds to the Stomach Meridian, and the gastrocnemius muscle corresponds to the Kidney Meridian. Key acupoints include Zusanli (ST36) and Yongquan (KI1).

[0031] The selection of these key acupoints is based on the theory of meridians in Traditional Chinese Medicine. Among them, Zusanli (ST36) belongs to the Stomach Meridian of Foot Yangming and is a vital acupoint for strengthening the patient, closely related to the function of the knee joint; Yongquan (KI1) belongs to the Kidney Meridian of Foot Shaoyin, and since the kidneys govern bones and produce marrow, the state of their Qi and blood directly affects the patient's bone health. Monitoring these key acupoints on specific meridians can accurately reflect changes in the Qi and blood of the meridians most relevant to bone injuries.

[0032] Electromyographic signals include integrated electromyography and average power frequency.

[0033] In some embodiments, the collected gait meridian data undergoes standardized preprocessing. Wavelet transform is used to remove jitter noise from each sensor, and the timestamps of kinematic, kinetic, and electrophysiological data are synchronized; for example, the time error of each data point is controlled within 1 millisecond. The formats of the kinematic, kinetic, and electrophysiological data are converted to a unified format, such as JSON.

[0034] In some embodiments, the method of this application further includes: Step S1104: Calculate the plantar pressure distribution characteristics based on the dynamic data; Step S1105: Compare the plantar pressure distribution characteristics with a preset pressure distribution threshold to determine whether there is an abnormal pressure distribution in the patient's foot. Step S1106: When there is no abnormal pressure distribution, reduce the sampling frequency of the plantar pressure sensor and reduce the number of sampling points of the plantar pressure sensor.

[0035] In most patients during the later stages of treatment, or those with milder conditions, plantar pressure distribution is normal. For these patients, collecting plantar kinetic data at a high frequency and with numerous sampling points may generate a large amount of repetitive, non-abnormal data. This non-abnormal data has low value for evaluating treatment efficacy and may increase the burden of data storage and computation.

[0036] Based on the peak pressure, pressure distribution uniformity, and pressure center trajectory of each region of the foot, plantar pressure distribution characteristics are calculated. These characteristics include the left-right plantar pressure difference rate and the pressure center trajectory offset. The left-right plantar pressure difference rate and pressure center trajectory offset are compared with their respective preset reference thresholds. When these thresholds are exceeded, an abnormal plantar pressure distribution is identified in the patient.

[0037] If the patient's foot pressure distribution is currently normal, reduce the sampling rate of the flexible plantar pressure sensor, for example, from 50Hz to 20Hz. Simultaneously, reduce the number of sampling points on the plantar pressure sensor, for example, retaining only 8 key sampling points corresponding to the 5 core meridian acupoints on the sole of the foot. After reducing the sampling rate of the plantar pressure sensor, check for any abnormalities in the patient's plantar pressure distribution at preset intervals. If abnormalities are found, increase the sampling rate of the plantar pressure sensor.

[0038] Therefore, reducing the sampling frequency from 50Hz to 20Hz and retaining only key points to obtain simplified stress data can significantly reduce the amount of data. While ensuring that the core assessment accuracy is not affected, reducing the collection and processing of unnecessary data can speed up the analysis and improve the patient's diagnosis and treatment experience.

[0039] In some embodiments, step S120 further includes: Step S1201: Using the feature mapping module, based on the time-series analysis algorithm, feature extraction is performed on the kinematic data to obtain kinematic sub-features; using the feature mapping module, feature extraction is performed on the dynamic data to obtain dynamic sub-features; using the feature mapping module, based on the acupoint conductivity, the mean and fluctuation amplitude of the conductivity of each acupoint during the patient's walking cycle are calculated to obtain acupoint conductivity sub-features; based on electromyographic signals, by comparing the electromyographic activation time of muscles along the meridians and muscles not along the meridians for each patient, electromyographic features are obtained.

[0040] A feature mapping module was used to perform time-series analysis on kinematic data to extract kinematic sub-features reflecting joint movement function. These sub-features include the range of motion of joints during the gait cycle, such as the maximum flexion and minimum extension angles of the knee joint, as well as stride length, cadence, and the proportion of the stance and swing phases in the gait cycle, such as the distance between the center points of the two feet per step, steps per minute, and the proportion of the stance and swing phases. These kinematic sub-features characterize whether the patient's joint movement function is currently normal.

[0041] The feature mapping module processes the kinetic data to extract kinetic sub-features for assessing the current patient's gait stability. These sub-features include the left-right plantar pressure difference rate calculated from the uniformity of plantar pressure distribution in both feet, and the pressure center trajectory offset represented by the maximum distance the plantar pressure center trajectory deviates from the midline. If simplified pressure data is acquired using an on-demand acquisition strategy, feature extraction is based solely on the pressure values ​​of retained key acupoints, ensuring compatibility with the acquisition strategy.

[0042] The feature mapping module is used to analyze physiological electrical data and extract physiological electrical features characterizing the state of the meridians. Based on the conductivity data of acupoints, the average value and fluctuation amplitude of the conductivity of each acupoint within the walking cycle are calculated to reflect the smoothness of Qi and blood flow in the meridians. Generally, higher conductivity corresponds to smoother Qi and blood flow in the meridians. Simultaneously, based on the comparison of the activation time difference between muscles along the meridians and those not along the meridians using electromyography (EMG) signals, the EMG activation timing difference is calculated to reflect the regulatory function of the meridians on muscles. For example, the activation time difference between the gastrocnemius muscle corresponding to the kidney meridian and other calf muscles.

[0043] In some embodiments, step S120, "determining the gait-meridian correlation weight matrix between the meridian features and the gait features using a convolutional neural network based on a self-attention mechanism, and obtaining the quantitative value of the current patient's meridian state based on the meridian features, the gait features, and the gait-meridian correlation weight matrix," includes: Step S1202: Using the feature mapping module, calculate the Pearson correlation coefficient between the meridian features and the gait features of the current patient, and generate the gait meridian correlation weight matrix; Step S1203: Using the feature mapping module, based on the gait meridian association weight matrix, the meridian features and the gait features are weighted and fused through the convolutional neural network to obtain fused features; Using physiological electrophysiological characteristics as meridian anchors and kinetic sub-characteristics as gait anchors, combined with clinically labeled data, the Pearson correlation coefficient was used to calculate the correlation between individual meridian features and individual gait features, constructing an initial correlation weight matrix. The row dimension of the initial correlation weight matrix represents the meridian feature type, the column dimension represents the gait feature type, and the matrix elements are the corresponding correlation weights. The clinically labeled data consisted of meridian features and gait features annotated by TCM physicians.

[0044] For example, when the electrical conductivity of the Zusanli acupoint on the Stomach Meridian of Foot Yangming increases by 10%, the correlation coefficient for increased stride length is 0.8. Therefore, the weight of this association is set to 0.8. If the change in electrical conductivity of a certain acupoint is not significantly related to a certain gait feature, that is, the correlation coefficient is less than 0.3, then the corresponding weight is set to less than 0.1 to weaken its influence.

[0045] Gait characteristics include kinetic and dynamic sub-characteristics, while meridian characteristics are also known as electrophysiological characteristics. Initial weights are first assigned to the kinetic, dynamic, and electrophysiological characteristics. Then, based on the initial correlation weight matrix, these initial weights are adjusted to ensure that meridian characteristics dominate. The core characteristic is the electrophysiological characteristic, with a base weight of 0.5. If the electrophysiological characteristic has a high correlation with the currently prioritized gait characteristic, its weight is increased. Important characteristics are the kinetic sub-characteristics, with a base weight of 0.3. If the kinetic sub-characteristics have a low correlation with meridian characteristics, their weight is decreased. Auxiliary characteristics are the dynamic sub-characteristics, with a base weight of 0.2, which is only moderately increased in gait stability assessment. The total weight of the adjusted kinetic, dynamic, and electrophysiological characteristics is 1.

[0046] Referring to the initial association weight matrix, if the association weight of a certain feature with all gait features is less than 0.3, the feature is determined to be irrelevant, and its weight is reduced to below 0.05. Significance tests are performed on each feature for the normal meridian group and the abnormal meridian group. Noise features irrelevant to the association with meridian features and gait features are filtered out, and the corrected weights are retained to obtain the gait meridian association weight matrix.

[0047] Step S1204: Using the feature mapping module, the fused features are mapped based on the convolutional neural network to obtain the meridian state quantification value; the meridian state quantification value represents the blood and qi flow of multiple meridians in the current patient.

[0048] The weighted and fused multimodal features are input into a 3-layer fully connected neural network for mapping, resulting in quantified values ​​of meridian status. The network input layer receives optimized and weighted multimodal features, including electrophysiological features, kinematic features, and kinetic features. These multimodal features undergo feature processing through two hidden layers: the first hidden layer contains 64 neurons for initial feature fusion, and the second hidden layer contains 32 neurons for core feature extraction. The output layer uses 5 neurons corresponding to 5 core meridians, and a sigmoid activation function is used to standardize the output values ​​to a range of 0 to 1, converting the output values ​​into percentage-based quantified values ​​of meridian status. These quantified values ​​directly characterize the degree of Qi and blood flow in each meridian of the patient. A value of over 70% indicates smooth Qi and blood flow, 50% to 70% indicates mild obstruction, and below 50% indicates severe obstruction. This ensures that the quantified results of meridian status are consistent with the clinical diagnostic conclusions of Traditional Chinese Medicine.

[0049] In some embodiments, step S120 further includes: Step S1205: Using the feature mapping module, the quantified value of the meridian state is input into a pre-trained decision tree model. The decision tree model is used for reasoning to output the gait abnormality attribution report. The decision tree model is trained based on a preset relationship between the quantified value of the meridian state and the gait abnormality index. The gait abnormality index is obtained by comparing the gait features with the gait feature reference range.

[0050] Using the quantified value of meridian status as the independent variable and gait abnormality indicators as the dependent variables (e.g., insufficient ankle dorsiflexion and shortened stride), a decision tree model is constructed. Each decision tree in the model corresponds to one gait abnormality indicator. The meridian patency threshold is used as the basis for dividing decision nodes, and this threshold is derived from clinical statistical data and traditional Chinese medicine theory. The decision tree model includes a pre-defined relationship between gait abnormality indicators and the quantified value of meridian status. Based on this pre-defined relationship, the decision tree model outputs a gait abnormality attribution report for the current patient.

[0051] If the matching rate between the gait anomaly attribution report and a certain feature is greater than 90%, the screening threshold for that feature will be lowered; if a certain interfering feature repeatedly causes attribution errors in the gait anomaly attribution report, for example, if the error rate is greater than 15%, then the interfering feature will be removed.

[0052] The meridian status quantification value includes the smooth flow of Qi and blood in 5 core meridians, and the gait abnormality attribution report includes gait abnormality indicators, the associated meridians of gait abnormality indicators, and the theoretical basis of traditional Chinese medicine.

[0053] In some embodiments, the "gait characteristics" in step S120 include kinematic sub-characteristics and kinetic sub-characteristics; the "meridian characteristics" include acupoint conductivity sub-characteristics and myoelectronic characteristics. The method of this application further includes: Step S1206: Obtain the current patient's clinical treatment information and treatment stage; the treatment stage includes the initial treatment stage, the middle treatment stage, and the later treatment stage; Step S1207: Construct a state space, which includes the current treatment stage of the patient, the quantified values ​​of the meridian states of multiple meridians, and the key gait feature change values ​​corresponding to the gait abnormality attribution report; The state space is the core input window for the deep deterministic policy gradient algorithm to perceive the individualized clinical treatment environment. The state space integrates multidimensional heterogeneous information into a numerical state vector, ensuring that the deep deterministic policy gradient algorithm can accurately capture dynamic changes during the treatment process. As shown in Table 1, the state space mainly contains three dimensions of information: treatment stage dimension, meridian state dimension, and gait change dimension. The gait change dimension refers to the key gait feature change values ​​corresponding to the gait abnormality attribution report, and the meridian state dimension refers to the quantified values ​​of the meridian states of multiple meridians.

[0054] Table 1. Three dimensions of state space information Table 1 shows three dimensions. The first dimension is the treatment stage dimension, which discretizes the continuous treatment process into three clinically significant stages, typically represented by a value of 1 for the initial treatment phase, 2 for the intermediate treatment phase, and 3 for the later treatment phase. The second dimension is the meridian status dimension, whose core data consists of quantified values ​​of Qi and blood flow in the five core meridians. These values ​​are expressed as percentages, such as 85%, 60%, 75%, 80%, and 70%, corresponding to the patency of the Liver, Heart, Spleen, Lung, and Kidney meridians, respectively. The meridian status dimension directly reflects the individualized meridian diagnosis results of the current patient and is a key basis for guiding the algorithm to allocate personalized weights. The third dimension is the gait change dimension, whose core data comes from the gait abnormality attribution report, extracting the changes in five key gait features compared to the baseline or the last assessment. These parameters may include joint range of motion, stride length, and cadence, such as 5 degrees, 0.08 meters, and 2 steps per minute. The gait change dimension quantifies the current improvement in the patient's motor function and is used to validate the effectiveness of the iterative optimization of the weight allocation strategy.

[0055] The three dimensions of information in the state space have different dimensions. For example, the meridian patency is expressed as a percentage, while the gait angle is expressed in degrees. To prevent one dimension from dominating the state judgment of the algorithm due to its large numerical magnitude, before inputting the data into the algorithm, it is necessary to use a standardization method to uniformly map the data of all dimensions to the range of 0 to 1, so as to achieve dimensionless processing of the data and ensure the balance of the contributions of each dimension in the state space.

[0056] Step S1208: Input the state space into the deep deterministic policy gradient algorithm, and output the fusion weights corresponding to the kinematic sub-features, kinetic sub-features, acupoint conductivity sub-features, and myoelectronic features for the current patient and the treatment stage; the deep deterministic policy gradient algorithm is trained by a reward function based on the Western medicine efficacy index and the TCM syndrome differentiation consistency rate; the Western medicine efficacy index is the reduction value of the current patient's pain score; the TCM syndrome differentiation consistency rate is the degree of matching between the meridian state quantification value and the physician's judgment result of the meridian state quantification value.

[0057] The deep deterministic policy gradient algorithm takes a state space consisting of treatment stage, meridian state quantification values, and key gait feature changes as input. After processing by the algorithm's internal policy network, it outputs fusion weights for the kinematic sub-features, kinetic sub-features, acupoint conductivity sub-features, and myoelectronic features corresponding to the current patient and treatment stage. The fusion weights can quantify the importance of the kinematic sub-features, kinetic sub-features, acupoint conductivity sub-features, and myoelectronic features in subsequent efficacy evaluation.

[0058] The deep deterministic policy gradient algorithm is pre-trained using a reward function. This reward function considers both the objective efficacy of Western medicine and the subjective diagnosis of Traditional Chinese Medicine (TCM). It is derived by adding a base reward and an additional reward. The base reward is correlated with the efficacy index of Western medicine, specifically the reduction in the visual analog scale (VAS) score, and a tiered reward standard is set. For example, a reduction of more than 4 points earns a high reward, a reduction of 2 to 3 points earns a moderate reward, no reward is given for no significant change in score, and a penalty is imposed if the score increases. The additional reward is correlated with the consistency rate of TCM diagnosis. This consistency rate is obtained by calculating the degree of matching between the quantitative value of the meridian state obtained by the algorithm and the independent judgment of a senior TCM physician. A matching degree exceeding 90% is considered highly consistent and earns a high reward, 70% to 90% is considered basically consistent and earns a reward, and less than 50% represents a serious discrepancy and requires a penalty.

[0059] Pain is a core symptom of bone and joint injuries. It is the most direct and subjective complaint of patients, and the primary indicator of limited mobility and other functions. Pain relief is a prerequisite for functional recovery. From a pathophysiological perspective, pain often originates from inflammation, joint dislocation, or nerve compression; pain relief signifies an improvement in the underlying pathology. For example, manual therapy can directly reduce pain by repositioning joints, thereby promoting the recovery of joint range of motion and gait. Therefore, a lower pain score indicates that the symptoms of the patient's bone and joint injury are under control, and that the patient's overall function is recovering.

[0060] Therefore, during the training process of the deep deterministic policy gradient algorithm, weight schemes that can achieve high pain relief and high diagnostic accuracy will receive high comprehensive rewards and will be strengthened and retained; conversely, invalid or erroneous schemes will receive zero or negative rewards and will be eliminated or reverse-optimized.

[0061] The deep deterministic policy gradient algorithm executes a phased weight allocation strategy based on the input state vector, and finally outputs a set of fusion weights that fit the treatment phase goals.

[0062] In some embodiments, step S1208, "inputting the state space into a deep deterministic policy gradient algorithm and outputting the fusion weights corresponding to the kinematic sub-features, kinetic sub-features, acupoint conductivity sub-features, and myoelectronic features for the current patient and the treatment stage", includes: Step S12081: In the initial stage of treatment, major weights are assigned to the kinematic sub-features; in the middle stage of treatment, major weights are assigned to the acupoint conductivity sub-features and the myoelectronic features; in the later stage of treatment, similar weights are assigned to the kinematic sub-features, the kinetic sub-features, the acupoint conductivity sub-features, and the myoelectronic features.

[0063] The fusion weights of the student features of the sports science are represented as follows: The fusion weights of the dynamic sub-features are expressed as: The fusion weights of myoelectronic features are expressed as: The fusion weights of the acupoint conductivity features are expressed as follows: .

[0064] The treatment stages are marked by physicians in clinical settings according to the principles of bone-setting techniques. The treatment stages are divided into three phases: the initial treatment phase, the middle treatment phase, and the later treatment phase.

[0065] The initial treatment phase primarily focuses on joint reduction, aiming to correct joint misalignment and restore the joint's basic range of motion. Therefore, the initial treatment phase requires a focus on kinematic characteristics. Consequently, primary weights are assigned to kinematic characteristics during this initial stage. For example, , , , .

[0066] The mid-treatment phase primarily focuses on muscle function recovery, aiming to restore the contractile function of muscles along the meridians. This phase is highly dependent on acupoint conductivity and muscle electronic characteristics. Therefore, during the mid-treatment phase, acupoint conductivity and muscle electronic characteristics are assigned primary weights. For example, , , , .

[0067] The later stages of treatment primarily focus on functional stabilization, aiming to achieve coordinated stability of joint, muscle, and meridian functions. This requires a balanced consideration of kinematic, kinetic, acupoint conductivity, and myoelectronic characteristics. Therefore, in the later stages of treatment, kinematic, kinetic, acupoint conductivity, and myoelectronic characteristics are assigned similar weights. For example, , , , .

[0068] This phased treatment embodies the principles of TCM's syndrome differentiation and treatment, focusing on treating the symptoms in acute cases and addressing the root cause in chronic cases. In the initial stage, patients experience severe pain and the primary issue lies in bone and muscle misalignment, corresponding to joint structural abnormalities in Western medicine. Therefore, priority is given to the patient's kinematic characteristics; this is the symptom-treating stage. In the middle stage, as the joint structures are initially repositioned, the focus shifts to restoring the nourishing function of Qi and blood to the muscles and tendons. Therefore, attention is paid to the physiological electrical characteristics reflecting the patient's meridian status; this is the root-cause-treating stage. In the later stage, the goal is to achieve harmony and stability of the patient's physical form and Qi, requiring a balanced assessment of all characteristics. This adaptive weighting mechanism allows the assessment process to dynamically align with the core pathological changes at different stages of bone and joint injuries.

[0069] In some embodiments, a comprehensive efficacy score is calculated before assessing the effectiveness of bone setting therapy. The comprehensive efficacy score is out of 100, with efficacy levels categorized as excellent, good, moderate, and poor. Excellent corresponds to a comprehensive efficacy score of 85-100, good to 70-84, moderate to 50-69, and poor to less than 50. For example, in the initial stage of treatment, a patient shows a 60% improvement in kinematic data. =0.6, other data improved by 20%, the total weight is 0.4, the comprehensive efficacy score = 60%×0.6×100+20%×0.4×100=36+8=44 points, the efficacy level corresponding to 44 points is poor, and the bone setting technique needs to be adjusted.

[0070] In some embodiments, step S130, assessing the efficacy of bone setting for the current patient, further includes the following: Current Western medicine functional recovery values ​​focus on functional improvements in joints and gait. They select the difference in range of motion before and after treatment for core joints such as the knee and ankle, and combine this with kinematic data weights to calculate the joint range of motion improvement rate, reflecting the improvement in joint function. The calculation of the joint range of motion improvement rate is shown in formula (1): Joint range of motion improvement rate = Formula (1); Select 3-5 key gait features, such as stride length, cadence, and uniformity of plantar pressure distribution, and determine whether each feature reaches the normal reference range. Combine the weights of kinematic and dynamic features to calculate the normality rate of gait features, reflecting the recovery of gait stability. The calculation of the normality rate of gait features is shown in formula (2): Gait feature normality rate = Formula (2); The current TCM meridian improvement value focuses on improving the smoothness of Qi and blood flow in the meridians. It calculates the difference in Qi and blood flow before and after treatment for each meridian, and combines the weight of acupoint conductivity to calculate the improvement rate of Qi and blood flow in the meridians, reflecting the improvement of meridian function. The calculation of the improvement rate of Qi and blood flow in the meridians is shown in formula (3): Improved flow of Qi and blood in the meridians = Formula (3); In some embodiments, the results of the dual-dimensional efficacy evaluation of traditional Chinese and Western medicine are visualized, including three core modules: (1) a comparison diagram of gait characteristics before and after treatment, (2) a trend diagram of changes in meridian status, and (3) a determination of efficacy level.

[0071] (1) The comparison chart of gait characteristics before and after treatment is presented using a biaxial line graph or bar chart. The horizontal axis represents the specific type of gait characteristic, such as knee joint angle, stride length, and cadence. The vertical axis corresponds to the measured values ​​of the gait characteristics, with units including angle, meter, or cadence. For the same gait characteristic, two sets of data will be displayed side by side in the chart, one set in blue to represent the situation before treatment, and the other set in red to represent the situation after treatment. To provide a reference benchmark, the normal range of values ​​for each parameter is marked on the chart. At the same time, the magnitude of the parameter value change will be clearly indicated by arrows or text annotations, for example, stride length increased by 0.1 meters and has reached the normal range, thus intuitively demonstrating the improvement effect of gait function.

[0072] (2) The trend chart of meridian status changes is presented using a stacked bar chart. The horizontal axis represents the core meridian type, and the vertical axis represents the smooth flow of Qi and blood. For each meridian, the chart uses two stacked bar charts: one in gray to represent the state before treatment, and the other in green to represent the state after treatment. Above the bar charts, the percentage increase in Qi and blood flow is marked; for example, the Kidney Meridian increased by 20%. The shade of green inside the bar charts corresponds to the level of Qi and blood flow; the darker the color, the higher the flow, reflecting the improvement trend of Qi and blood in the meridians.

[0073] (3) The determination of therapeutic efficacy level is shown in Table 2: Table 2. Determination of Therapeutic Effect Level To ensure the accuracy of the dual-dimensional efficacy evaluation results of Traditional Chinese Medicine (TCM) and Western medicine, data consistency verification and grade logic verification are required before generating the final dual-dimensional efficacy evaluation results. For example, if the comprehensive efficacy score is 85 points, the excellent level belongs to the superior level, corresponding to a series of clearly preset indicator improvement requirements, such as a significant improvement in the smooth flow of Qi and blood in key meridians. If there is a significant contradiction between the comprehensive efficacy score and the performance of specific indicators, or if the comprehensive efficacy score reaches 85 points but the actual recorded meridian improvement rate is only 5%, it indicates that there may be an unreasonable aspect in the current weight calculation or scoring logic. In this case, the system returns to the weight self-optimization module to recalculate and merge the weight allocation and generate a new comprehensive efficacy score until the score result and the actual performance of each indicator satisfy the data consistency verification and grade logic verification.

[0074] In some embodiments, the method of this application further includes: Step S140: Based on the current Western medicine functional recovery value, the current TCM meridian improvement value, and the preset manipulation gait correlation map, recommend one or more target bone-setting manipulation techniques for the current patient.

[0075] In some embodiments, the method of this application further includes: Step S101: Obtain historical medical data for multiple patients; Complete medical records of a large number of historical patients were collected, with a sample size of at least one thousand cases. Each historical medical record contained three dimensions: (1) the type of bone-setting technique used, such as the lumbar oblique board method and the knee joint rotation method; (2) the historical Western medicine functional recovery value generated after the bone-setting technique was implemented, specifically manifested as changes in gait characteristics, such as changes in stride length and joint angle; (3) the historical TCM meridian improvement value caused by the bone-setting technique, specifically manifested as quantitative changes in the specific meridian state. To ensure the validity of the data, each historical medical record was annotated by multiple TCM experts at the level of associate chief physician or above to evaluate the effectiveness of specific techniques on specific efficacy indicators. For example, the knee joint rotation method was annotated to have an effectiveness rate of 80% in improving ankle joint angle. Only data with an annotation effectiveness of 50% or higher were selected and included in the sample database.

[0076] Step S102: Based on the historical medical data, the Apriori algorithm is used to mine the historical bone-setting techniques used by each historical patient in each treatment, as well as the correlation rules between the historical Western medicine functional recovery value and the historical traditional Chinese medicine meridian improvement value generated by the historical bone-setting techniques. This study aims to uncover correlation rules between historical bone-setting techniques, historical Western medicine functional recovery values, and historical Traditional Chinese Medicine meridian improvement values ​​from historical clinical data. To improve the efficiency of this process, two optimizations were made to the classic Apriori algorithm: First, the input data was dimensionally reduced by pre-filtering out combinations of techniques and therapeutic effects with an effectiveness rate below 50% before inputting them into the algorithm, thus removing redundant and noisy data. Second, a weighting mechanism was introduced to incorporate data weights reflecting treatment stages and individual differences into the mining process, making the rules more aligned with clinical practice.

[0077] The specific process for mining the aforementioned association rules is as follows: Taking the association rule between the knee joint rotation method and a 15% increase in stomach meridian patency and a 0.1-meter increase in stride as an example, the type of manipulation, the amount of change in meridian status, and the amount of improvement in gait characteristics are defined as different items, forming item sets at different levels. For example, a single item constitutes one item set, such as the knee joint rotation method; a combination of two items constitutes two item sets, such as the combination of the knee joint rotation method and a 15% increase in stomach meridian patency; and a combination of three items constitutes three item sets.

[0078] Calculate the support for each itemset, which is the percentage of all samples containing that itemset out of the total sample size. Filter out frequently occurring itemsets by setting a minimum support threshold, for example, greater than 70%. Generate association rules based on these frequent itemsets and calculate the confidence of each rule, defined as the percentage of samples containing both the antecedent and consequent of the rule out of samples containing only the antecedent. Again, set a threshold, for example, a confidence level greater than 85%, to filter out highly reliable, strongly associated rules.

[0079] A rule base that meets both support and confidence thresholds is obtained. For example, Rule 1: Knee rotation method is associated with a 15% increase in Stomach Meridian patency and a 0.1-meter increase in stride length, with a support of 78% and a confidence of 85%. Rule 2: Moxibustion at Yongquan acupoint is associated with an 8% increase in Kidney Meridian patency and further associated with a 5-degree increase in ankle dorsiflexion angle, with a support of 82% and a confidence of 88%.

[0080] Step S103: Based on the association rules, generate the preset manipulative gait association map.

[0081] The association rule base is transformed into an intuitive and visual knowledge graph, namely the preset manipulation and gait association graph. The preset manipulation and gait association graph is constructed using a directed acyclic graph (DAG). Nodes in the DAG are distinguished into three types of entities by different shapes and colors: historical bone-setting manipulation type nodes are represented by blue circles; historical TCM meridian improvement value nodes are represented by green squares; and historical Western medicine functional recovery value nodes are represented by orange triangles. Directed edges between nodes represent the mined association rules, and the direction of the edges indicates the direction of the association relationship. The thickness of the edges is determined by the weighted value of the support and confidence of the association rule; the higher the strength, the thicker the line segment. The edges are also labeled with the core metrics of the association rule, such as 78% support and 85% confidence.

[0082] This paper transforms the rich clinical experience inherent in Traditional Chinese Medicine (TCM) theories of treating different diseases with the same method and treating the same disease with different methods into a systematic and structured pre-defined manipulation-gait correlation map. TCM theory states that different bone-setting techniques, such as lumbar oblique manipulation or knee joint rotation, act on specific meridians in the body, triggering corresponding Qi effects—that is, functional responses of the meridian system—and physical effects, manifested as objective improvements in gait and other motor functions. The pre-defined manipulation-gait correlation map, by mining and extracting correlation rules from a large amount of clinical data, presents the complex interaction between bone-setting techniques, meridians, and gait in a visual form, and provides convenient query functions. This essentially constructs a TCM bone-setting prescription and syndrome correspondence knowledge base supported by clinical big data, thereby providing intelligent decision support for physicians to formulate precise and personalized manipulation prescriptions in clinical practice.

[0083] To ensure the timeliness and accuracy of the atlas, the preset manipulation gait correlation atlas is dynamically updated in real time. When a new correlation rule is added between a bone-setting manipulation technique and its corresponding Western medicine functional recovery value and Traditional Chinese Medicine meridian improvement value, experts conduct validity labeling and screening of the new correlation rule. Only new correlation rules with a validity higher than 80% enter the verification process. The matching degree between the weights of the new correlation rule and the existing correlation rules is verified; a matching degree higher than 90% indicates successful verification. The verified new correlation rules are added to the correlation rule library, and corresponding nodes and directed edges are simultaneously added to the preset manipulation gait correlation atlas. At the same time, the thickness of the edges of existing similar rules in the preset manipulation gait correlation atlas is adjusted to dynamically reflect the correlation strength supported by the latest data, thereby ensuring that the entire preset manipulation gait correlation atlas continues to evolve and remain accurate.

[0084] The preset manipulation gait correlation map is presented in an interactive visual chart format, supporting zooming in and out and node retrieval. Physicians can view the correlation rules between Western medicine functional recovery values ​​and traditional Chinese medicine meridian improvement values ​​associated with bone setting manipulation nodes, or reverse query recommended bone setting manipulations.

[0085] This application transforms meridian patency into a measurable indicator, achieving an 88% accuracy rate in correlating meridian characteristics from a Traditional Chinese Medicine (TCM) perspective with gait characteristics from a Western medicine perspective. It addresses the issue of reliance on physician subjective judgment in orthopedic injury treatment evaluation. Through reinforcement learning and dynamic weighting, the accuracy of fusing multiple sub-features reaches 92%, a 20% improvement over fixed-weight algorithms. This approach adaptively adapts to different patients' treatment stages and individual differences. Treatment based on pre-defined manipulation-gait correlation maps recommends bone-setting techniques, increasing physician technique selection efficiency by 30% and shortening patient recovery cycles by 20%. This application provides dual-dimensional interpretation of treatment efficacy from both TCM and Western medicine perspectives, reducing the incidence of adverse reactions from 5% to 0%, meeting the high standards of safety and comprehensiveness required throughout the clinical process.

[0086] Figure 2 A schematic diagram of a traditional Chinese medicine bone-setting manipulation efficacy evaluation system based on gait analysis provided in this application embodiment is shown below. Figure 2 As shown, this application proposes a traditional Chinese medicine bone-setting technique efficacy evaluation system 200 based on gait analysis, including: a data acquisition module 210, a feature mapping module 220 and an efficacy evaluation module 230; The data acquisition module 210 is used to acquire the current patient's gait meridian data; and to preprocess the gait meridian data to obtain preprocessed gait meridian data. The feature mapping module 220 is used to extract features from the preprocessed gait meridian data to obtain meridian features and gait features; through a convolutional neural network based on a self-attention mechanism, it determines the gait meridian correlation weight matrix between the meridian features and the gait features, and obtains the quantitative value of the current patient's meridian state based on the meridian features, the gait features and the gait meridian correlation weight matrix; The efficacy evaluation module 230 is used to acquire the current patient's historical medical information, the fusion weights of each sub-feature in the meridian features and gait features; calculate the current TCM meridian improvement value based on the historical medical information, the quantitative value of the meridian status, and the fusion weights of each sub-feature in the meridian features; calculate the current Western medicine functional recovery value based on the historical medical information, the gait features, and the fusion weights of each sub-feature in the gait features; and generate a dual-dimensional TCM and Western medicine efficacy evaluation result based on the current Western medicine functional recovery value and the current TCM meridian improvement value.

[0087] In some embodiments, the system 200 further includes a weight optimization module 240 and a bone-setting technique recommendation module 250.

[0088] The weight optimization module 240 is used to acquire the current patient's clinical treatment information and treatment stage; the treatment stage includes the initial treatment stage, the middle treatment stage, and the later treatment stage; it constructs a state space, which includes the current patient's treatment stage, the quantified values ​​of the meridian states of multiple meridians, and the key gait feature change values ​​corresponding to the gait abnormality attribution report; the state space is input into the deep deterministic policy gradient algorithm, which outputs the fusion weights corresponding to the kinematic sub-features, dynamic sub-features, acupoint conductivity sub-features, and myoelectronic features for the current patient and treatment stage; the deep deterministic policy gradient algorithm is trained by a reward function based on the consistency rate between Western medicine efficacy indicators and traditional Chinese medicine diagnosis; the Western medicine efficacy indicator is the reduction value of the current patient's pain score; the consistency rate of traditional Chinese medicine diagnosis is the degree of matching between the quantified values ​​of the meridian states and the physician's judgment results on the quantified values ​​of the meridian states.

[0089] Among them, gait characteristics include kinematic sub-characteristics and dynamic sub-characteristics; meridian characteristics include acupoint conductivity sub-characteristics and myoelectronic characteristics.

[0090] The bone-setting technique recommendation module 250 is used to recommend one or more target bone-setting techniques to the current patient based on the current Western medicine functional recovery value, the current TCM meridian improvement value, and the preset technique gait correlation map.

[0091] It should be noted that the description of the above system embodiments is similar to the description of the above method embodiments, and has similar beneficial effects. For technical details not disclosed in the system embodiments of this application, please refer to the description of the method embodiments of this application for understanding.

[0092] It should be noted that, in the embodiments of this application, if the above-mentioned method for evaluating the efficacy of traditional Chinese medicine bone-setting techniques based on gait analysis is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiments of this application, or the part that contributes to related technologies, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a magnetic disk, or an optical disk. Thus, the embodiments of this application are not limited to any specific hardware and software combination.

[0093] Correspondingly, embodiments of this application provide a computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the steps in any of the above-described methods for evaluating the efficacy of traditional Chinese medicine bone-setting manipulation based on gait analysis. Correspondingly, embodiments of this application also provide a computer program product, which, when executed by a processor of an electronic device, is used to implement the steps in any of the above-described methods for evaluating the efficacy of traditional Chinese medicine bone-setting manipulation based on gait analysis.

[0094] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of this application, the sequence numbers of the above-described processes do not imply a sequential order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application. The sequence numbers of the above-described embodiments are merely descriptive and do not represent the superiority or inferiority of the embodiments.

[0095] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0096] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0097] The units described above as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units; some or all of the units may be selected to achieve the purpose of the embodiments of this application according to actual needs. In addition, each functional unit in the embodiments of this application may be fully integrated into one processing unit, or each unit may be a separate unit, or two or more units may be integrated into one unit; the integrated unit may be implemented in hardware or in the form of hardware plus software functional units.

[0098] Alternatively, if the integrated units described above are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this application, or the parts that contribute to related technologies, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause the device automatic test line to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROMs, magnetic disks, or optical disks.

[0099] The methods disclosed in the several method embodiments provided in this application can be arbitrarily combined to obtain new method embodiments without conflict. The features disclosed in the several method or device embodiments provided in this application can be arbitrarily combined to obtain new method embodiments or device embodiments without conflict.

[0100] The above description is merely an embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for evaluating the efficacy of traditional Chinese medicine bone-setting techniques based on gait analysis, characterized in that, A system for evaluating the efficacy of traditional Chinese medicine bone-setting manipulation techniques based on gait analysis is provided. The system includes a data acquisition module, a feature mapping module, and an efficacy evaluation module. The method includes: The data acquisition module is used to collect the current patient's gait meridian data; the gait meridian data is preprocessed to obtain preprocessed gait meridian data. Using the feature mapping module, feature extraction is performed on the preprocessed gait meridian data to obtain meridian features and gait features; using a convolutional neural network based on a self-attention mechanism, the gait meridian correlation weight matrix between the meridian features and the gait features is determined, and based on the meridian features, the gait features, and the gait meridian correlation weight matrix, the quantitative value of the current patient's meridian state is obtained; Using the efficacy evaluation module, the system obtains the patient's historical medical information, the fusion weights of each sub-feature in the meridian features and gait features; based on the historical medical information, the quantitative value of the meridian status, and the fusion weights of each sub-feature in the meridian features, it calculates the current TCM meridian improvement value; based on the historical medical information, the gait features, and the fusion weights of each sub-feature in the gait features, it calculates the current Western medicine functional recovery value; based on the current Western medicine functional recovery value and the current TCM meridian improvement value, it generates a dual-dimensional TCM and Western medicine efficacy evaluation result.

2. The method according to claim 1, characterized in that, The gait features include kinetic and dynamic sub-features; the meridian features include acupoint conductivity and myoelectronic features; the method further includes: Obtain current patient clinical treatment information and treatment stage; the treatment stage includes the initial treatment stage, the middle treatment stage, and the later treatment stage. A state space is constructed, which includes the current treatment stage of the patient, the quantitative values ​​of the meridian states of multiple meridians, and the key gait feature change values ​​corresponding to the gait abnormality attribution report; The state space is input into a deep deterministic policy gradient algorithm, which outputs fusion weights for the kinematic sub-features, kinetic sub-features, acupoint conductivity sub-features, and myoelectronic features of the current patient and the treatment stage. The deep deterministic policy gradient algorithm is trained by a reward function based on the consistency rate between Western medicine efficacy indicators and traditional Chinese medicine diagnosis. The Western medicine efficacy indicator is the reduction in the current patient's pain score. The consistency rate between traditional Chinese medicine diagnosis and the meridian state quantification value is the degree of matching between the quantification value of the meridian state and the physician's judgment result on the quantification value of the meridian state.

3. The method according to claim 2, characterized in that, The process of inputting the state space into a deep deterministic policy gradient algorithm and outputting fusion weights corresponding to the kinematic sub-features, kinetic sub-features, acupoint conductivity sub-features, and myoelectronic features for the current patient and the treatment stage includes: In the initial stage of the treatment, primary weights are assigned to the kinematic sub-features; During the mid-treatment period, primary weights are assigned to the acupoint conductivity characteristics and the myoelectronic characteristics; In the later stages of treatment, similar weights are assigned to the kinematic sub-feature, the kinetic sub-feature, the acupoint conductivity sub-feature, and the myoelectronic feature.

4. The method according to claim 1, characterized in that, The method further includes: Based on the current Western medicine functional recovery value, the current TCM meridian improvement value, and the preset manipulation gait correlation map, one or more target bone-setting manipulation techniques are recommended for the current patient.

5. The method according to claim 1, characterized in that, The gait meridian data includes kinematic data, dynamic data, and physiological electrical data; the process of collecting the current patient's gait meridian data using the data acquisition module includes: Using the data acquisition module, inertial measurement unit sensors are attached to the patient's knee joint, ankle joint, and pelvis to collect kinematic data; the kinematic data includes joint angles, movement speed, and displacement. Using the data acquisition module, a foot pressure sensor array is attached to the corresponding meridian acupoint area on the patient's sole to collect the dynamic data; the dynamic data includes peak pressure, pressure distribution uniformity, and pressure center trajectory in each area of ​​the sole. The physiological electrical data includes acupoint conductivity and electromyographic signals; using the data acquisition module, a meridian acupoint conductivity sensor is attached to key acupoints on the patient's core meridian to collect the acupoint conductivity, and surface electromyographic electrodes are used to collect the electromyographic signals of muscles along the meridian.

6. The method according to claim 5, characterized in that, The method further includes: Based on the aforementioned dynamic data, the plantar pressure distribution characteristics were calculated; The plantar pressure distribution characteristics are compared with a preset pressure distribution threshold to determine whether there is an abnormal pressure distribution in the patient's foot. When there is no abnormal pressure distribution, reduce the sampling frequency of the plantar pressure sensor and reduce the number of sampling points of the plantar pressure sensor.

7. The method according to claim 5, characterized in that, The process involves using a convolutional neural network based on a self-attention mechanism to determine the gait-meridian correlation weight matrix between the meridian features and the gait features, and obtaining a quantitative value of the current patient's meridian state based on the meridian features, the gait features, and the gait-meridian correlation weight matrix, including: Using the feature mapping module, the Pearson correlation coefficient between the meridian features and the gait features of the current patient is calculated, and the gait meridian correlation weight matrix is ​​generated; Using the feature mapping module, based on the gait meridian association weight matrix, the convolutional neural network is used to perform weighted fusion of the meridian features and the gait features to obtain fused features; Using the feature mapping module, the fused features are mapped based on the convolutional neural network to obtain the meridian state quantification value; the meridian state quantification value represents the smooth flow of Qi and blood in multiple meridians of the current patient.

8. The method according to claim 1, characterized in that, The method further includes: Using the feature mapping module, the quantified value of the meridian state is input into a pre-trained decision tree model. The decision tree model performs inference and outputs the gait abnormality attribution report. The decision tree model is trained based on a preset relationship between the quantified value of the meridian state and the gait abnormality index. The gait abnormality index is obtained by comparing the gait features with the gait feature reference range.

9. The method according to claim 4, characterized in that, The method further includes: Obtain historical medical data from multiple patients; Based on the historical medical data, the Apriori algorithm is used to mine the historical bone-setting techniques used by each historical patient in each treatment, as well as the correlation rules between the historical Western medicine functional recovery value and the historical traditional Chinese medicine meridian improvement value generated by the historical bone-setting techniques. Based on the association rules, the preset technique gait association map is generated.

10. A system for evaluating the efficacy of traditional Chinese medicine bone-setting techniques based on gait analysis, characterized in that, include: Data acquisition module, feature mapping module, and efficacy evaluation module; The data acquisition module is used to collect the current patient's gait and meridian data; The gait meridian data is preprocessed to obtain preprocessed gait meridian data; The feature mapping module is used to extract features from the preprocessed gait meridian data to obtain meridian features and gait features; By using a convolutional neural network based on a self-attention mechanism, the gait meridian correlation weight matrix between the meridian features and the gait features is determined, and the quantitative value of the current patient's meridian status is obtained based on the meridian features, the gait features, and the gait meridian correlation weight matrix. The efficacy evaluation module is used to obtain the current patient's historical diagnosis and treatment information, the fusion weights of each sub-feature in the meridian features and gait features; Based on the historical diagnosis and treatment information, the quantitative value of the meridian status, and the fusion weight of each sub-feature in the meridian features, the current TCM meridian improvement value is calculated. Based on the historical medical information, the gait features, and the fusion weights of each sub-feature in the gait features, the current Western medicine functional recovery value is calculated; Based on the current Western medicine functional recovery value and the current TCM meridian improvement value, a dual-dimensional efficacy evaluation result of TCM and Western medicine is generated.

Images