A method and system for preventive evaluation of knee joint injury based on kinematic analysis
Patent Information
- Application Number
- CN202511517082.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-23
- Publication Date
- 2026-10-27
AI Technical Summary
目前，对于膝关节损伤的评估方法主要包括影像学检查、临床体格检查等，但这些方法大多是在损伤发生后进行诊断，对于预防性评估存在不足
本发明提出了一种基于运动力学分析的膝关节损伤预防性评估方法，包括：采集受试者运动过程中的待测数据；通过所述待测数据构建生物力学模型并进行校准和验证；根据待测数据和生物力学模型进行特征参数提取；基于提取的特征参数，建立膝关节损伤风险评估指标体系和预测模型；根据风险评估指标体系和预测模型对受试者的膝关节损伤风险进行评估。本发明通过数据采集公式精准获取膝关节运动、受力及肌肉状态数据；模型建立公式通过有限元等方法模拟力学行为，并经验证确保可靠；特征提取公式从海量数据中提炼关键风险特征；风险评估公式结合指标阈值与机器学习，精准划分风险等级，并为高风险者定制预防方案。实现精准数据处理、科学模型构建和有效风险判断，为膝关节损伤风险评估提供可靠方案，兼具科学与应用价值。
Smart Images
Figure CN122991570A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of sports medicine technology, and more specifically to a method and system for preventive assessment of knee joint injuries based on biomechanical analysis. Background Technology
[0002] The knee joint is one of the most easily injured joints in the human body. During sports such as basketball, soccer, and running, the knee joint bears significant pressure and impact. Currently, assessment methods for knee joint injuries mainly include imaging examinations and clinical physical examinations. However, these methods mostly diagnose after an injury has occurred, and are insufficient for preventative assessment. Assessment methods based on biomechanical analysis, on the other hand, can analyze the mechanical characteristics of the knee joint during movement before an injury occurs, identifying potential injury risks in advance and allowing for appropriate preventative measures.
[0003] Therefore, how to propose a method and system for preventive assessment of knee joint injuries based on biomechanical analysis, and how to identify potential injury risks in advance by analyzing the mechanical characteristics of the knee joint during movement, is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention
[0004] In view of this, the present invention provides a method and system for preventive assessment of knee joint injuries based on biomechanical analysis. By analyzing the biomechanical characteristics of the knee joint during movement, potential injury risks can be identified in advance. To achieve the above objectives, the present invention adopts the following technical solution: A method for preventive assessment of knee joint injuries based on biomechanical analysis, comprising: Collect test data during the subject's movement; A biomechanical model was constructed using the test data and then calibrated and validated. Feature parameters are extracted based on the test data and biomechanical model; Based on the extracted feature parameters, a knee joint injury risk assessment index system and prediction model were established. The risk of knee injury to the subjects was assessed based on a risk assessment index system and a prediction model.
[0005] Optionally, the data to be collected during the subject's movement includes: kinematic data, dynamic data, and electromyographic data. A motion capture system is used to collect kinematic data of the knee joint when the subject performs a specific movement task; force sensors are used to measure the ground reaction force and the internal and external forces borne by the knee joint; miniature force sensors are installed at the muscle attachment points around the knee joint to measure the effect of muscle contraction on the knee joint; and a surface electromyography device is used to collect electrical activity signals of the muscles around the knee joint.
[0006] Optionally, it also includes: calculating the knee joint angle in three-dimensional space based on the three-dimensional coordinate changes of reflective markers placed on key parts of the body, calculating the angular velocity through the rate of change of the joint angle with time, calculating the angular acceleration through the rate of change of the angular velocity with time, processing the raw electromyographic signals collected by the surface electromyography device, and calculating the root mean square value to represent the degree of muscle activation.
[0007] Optionally, the construction of the biomechanical model includes: In constructing the multibody dynamics model, the equations of motion are established based on the Newton-Euler equations: ; Where M is the generalized mass matrix and q is the generalized coordinate vector. and These are the generalized velocity and generalized acceleration vectors, respectively. The matrix of Coriolis force and centrifugal force. The gravity vector As a generalized force vector, this equation is used to simulate the mechanical behavior of the knee joint under different motion states.
[0008] Optionally, the calibration and verification process includes: using the mean squared error (MSE) to evaluate the difference between the model prediction results and the actual measurement data, as shown in the formula: ; in, These are actual measured values. The model predicts the value, and n is the number of data samples. By adjusting the model parameters, the MSE is minimized to reflect the actual mechanical properties of the knee joint.
[0009] Optionally, the feature parameter extraction based on the test data and biomechanical model includes: extracting the maximum flexion-extension angle and peak acceleration from kinematic features, extracting the maximum force and torque from dynamic features, and extracting muscle activation time and peak electromyographic value from electromyographic features.
[0010] Optionally, the knee joint injury risk assessment index system includes: Set the maximum flexion-extension angle threshold of the knee joint as The threshold for inward and outward folding torque is The muscle activation delay threshold is A risk level is determined to be high when the following conditions are met: Maximum knee flexion and extension angle: ; Inward and outward turning torque: ; Delayed muscle activation: ,in, This represents the actual measured muscle activation delay time.
[0011] Optionally, the prediction model includes: Constructing a Vector Machine (SVM) for a binary classification problem, its decision function is: ; in For the weight vector, For the input feature vector, For bias terms, The sign function is optimized by training on data from subjects with known injuries and no injuries. and The value of .
[0012] Optionally, it also includes: constructing a judgment matrix using the analytic hierarchy process, determining the relative importance of different risk indicators through expert scoring, and calculating the weight vector; The extracted feature parameters are combined with the weight vector to calculate the comprehensive risk score; Risk levels are determined based on the range of values in the comprehensive risk score. The calculated comprehensive risk score is used as the input feature and is fed into the support vector machine (SVM) along with the original feature parameters. The model's prediction results are compared with the actual damage situation, and the risk index weights, risk level thresholds, and machine learning model parameters are adjusted through cross-validation.
[0013] Optionally, a knee injury prevention assessment system based on biomechanical analysis includes: Data acquisition module: Used to collect the data to be measured during the subject's movement; Biomechanical model construction module: used to construct, calibrate, and validate a biomechanical model using the test data; Feature parameter extraction module: used to extract feature parameters based on the test data and biomechanical model; Model building module: used to establish a knee joint injury risk assessment index system and prediction model based on the extracted feature parameters; Assessment module: Used to assess the risk of knee injury in subjects based on a risk assessment index system and a prediction model.
[0014] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a method and system for preventive assessment of knee joint injuries based on biomechanical analysis, which has the following beneficial effects: This invention proposes a method for preventive assessment of knee joint injuries based on biomechanical analysis, comprising: collecting test data during the subject's movement; constructing a biomechanical model using the test data and calibrating and validating it; extracting feature parameters based on the test data and the biomechanical model; establishing a knee joint injury risk assessment index system and prediction model based on the extracted feature parameters; and assessing the subject's knee joint injury risk based on the risk assessment index system and prediction model. This invention accurately acquires knee joint movement, force, and muscle state data through data acquisition formulas; the model establishment formulas simulate mechanical behavior using methods such as the finite element method and have been verified to ensure reliability; the feature extraction formulas extract key risk features from massive amounts of data; and the risk assessment formulas combine index thresholds with machine learning to accurately classify risk levels and customize prevention plans for high-risk individuals. This achieves accurate data processing, scientific model construction, and effective risk judgment, providing a reliable solution for knee joint injury risk assessment, possessing both scientific and practical value. Attached Figure Description
[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0016] Figure 1 This invention provides a schematic flowchart of a method for preventive assessment of knee joint injuries based on biomechanical analysis.
[0017] Figure 2 This invention provides a structural framework diagram of a knee joint injury preventive assessment system based on biomechanical analysis. Detailed Implementation
[0018] 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 some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0019] This invention discloses a method for preventive assessment of knee joint injuries based on biomechanical analysis, such as... Figure 1 As shown, it includes: Collect test data during the subject's movement; A biomechanical model was constructed using the test data and then calibrated and validated. Feature parameters are extracted based on the test data and biomechanical model; Based on the extracted feature parameters, a knee joint injury risk assessment index system and prediction model were established. The risk of knee injury to the subjects was assessed based on a risk assessment index system and a prediction model.
[0020] Furthermore, the data to be collected during the subject's movement includes: kinematic data, dynamic data, and electromyographic data. A motion capture system is used to collect kinematic data of the knee joint when the subject performs a specific movement task; force sensors are used to measure the ground reaction force and the internal and external forces borne by the knee joint; miniature force sensors are installed at the muscle attachment points around the knee joint to measure the effect of muscle contraction on the knee joint; and a surface electromyography device is used to collect electrical activity signals of the muscles around the knee joint.
[0021] Furthermore, it also includes: calculating the knee joint angle in three-dimensional space based on the three-dimensional coordinate changes of reflective markers placed on key parts of the body, calculating the angular velocity through the rate of change of the joint angle with respect to time, calculating the angular acceleration through the rate of change of the angular velocity with respect to time, processing the raw electromyographic signals collected by the surface electromyography instrument, and calculating the root mean square value to represent the degree of muscle activation.
[0022] In a specific embodiment, the knee joint angle is calculated in three-dimensional space based on the changes in the three-dimensional coordinates of reflective markers placed on key parts of the body.
[0023] (1) Obtaining joint points through motion capture system , , The coordinates, in vector form and Calculate joint angles based on According to the dot product formula... We can obtain: ;in, , , , .
[0024] (2) Calculation of angular velocity: angular velocity The rate of change of joint angle over time is calculated, and approximated using the finite difference method: ;in, and They are time points The joint angle with t, For time intervals.
[0025] (3) Calculation of angular acceleration: angular acceleration The rate of change of angular velocity with respect to time is calculated using the finite difference method as well: ;in and They are time points And the angular velocity of t.
[0026] (4) Ground reaction force analysis: Let the components of the force collected by the shoe sole force sensor in the x, y, and z directions be F respectively. x F y F z The magnitude of the ground reaction force F is: .
[0027] The direction can be determined by the ratio of the components, such as the angle with the x-axis. : .
[0028] (5) Calculation of intra-articular forces: Miniature force sensors are installed at the muscle attachment points around the knee joint. Let the forces measured by each sensor be F. i (i=1,2,...,n, where n is the number of sensors), then the total internal force borne by the knee joint.
[0029] (6) Electromyography (EMG) data processing: The raw EMG signals acquired by the surface electromyography (SEMG) instrument need to be processed, and the root mean square (RMS) value is calculated to reflect the degree of muscle activation. The formula is as follows: , where x i Let N be the electromyographic signal value at the i-th sampling point, and N be the number of sampling points.
[0030] Furthermore, the construction of the biomechanical model includes: In constructing the multibody dynamics model, the equations of motion are established based on the Newton-Euler equations: ; Where M is the generalized mass matrix and q is the generalized coordinate vector. and These are the generalized velocity and generalized acceleration vectors, respectively. The matrix of Coriolis force and centrifugal force. The gravity vector As a generalized force vector, this equation is used to simulate the mechanical behavior of the knee joint under different motion states.
[0031] Furthermore, the calibration and verification process includes: using the mean squared error (MSE) to evaluate the difference between the model prediction results and the actual measurement data, as shown in the formula: ; in, These are actual measured values. The model predicts the value, and n is the number of data samples. By adjusting the model parameters, the MSE is minimized to reflect the actual mechanical properties of the knee joint.
[0032] Furthermore, the feature parameter extraction based on the test data and biomechanical model includes: extracting the maximum flexion-extension angle and peak acceleration from kinematic features, extracting the maximum force and torque from dynamic features, and extracting muscle activation time and peak electromyographic value from electromyographic features.
[0033] In a specific embodiment, the feature parameter extraction includes: Maximum flexion-extension angle: data collected from the knee joint flexion-extension angle. In the process, the maximum flexion-extension angle is obtained by maximizing the value. .
[0034] Peak acceleration: data on knee joint acceleration The peak acceleration is obtained by finding the maximum value. .
[0035] Maximum force: Data on the force exerted on the knee joint In the process, the maximum force is obtained by finding the maximum value. : .
[0036] Torque Calculation: Assuming the position vector from the point of application of force F to the joint center is r, then the torque M is: Its size ,in Let r be the angle between r and F. The maximum torque M can be obtained by finding the maximum value. max .
[0037] Muscle activation time: Set the electromyographic signal threshold T, when the electromyographic signal... Time when the threshold is first exceeded This refers to the muscle activation time, i.e.: .
[0038] Peak electromyography (EMG) values: from processed EMG data In the process, the peak electromyographic value is obtained by finding the maximum value. .
[0039] Furthermore, the knee joint injury risk assessment index system includes: Set the maximum flexion-extension angle threshold of the knee joint as The threshold for inward and outward folding torque is The muscle activation delay threshold is A risk level is determined to be high when the following conditions are met: Maximum knee flexion and extension angle: ; Inward and outward turning torque: ; Delayed muscle activation: ,in, This represents the actual measured muscle activation delay time.
[0040] Furthermore, the prediction model includes: Constructing a Vector Machine (SVM) for a binary classification problem, its decision function is: ; in For the weight vector, For the input feature vector, For bias terms, The sign function is optimized by training on data from subjects with known injuries and no injuries. and The value of .
[0041] In a specific embodiment, the specific formula steps for risk assessment include: S1: Using the analytic hierarchy process (AHP), a judgment matrix is constructed. Expert scoring determines the relative importance of different risk indicators (such as maximum knee flexion / extension angle, varus / valgus moment, and muscle activation delay), and a weight vector is calculated. Assuming there are m risk indicators, the weight vector is represented as follows: And satisfy .
[0042] S2: Combine the extracted feature parameters with the weight vector to calculate the comprehensive risk score; the extracted feature parameters (Each risk indicator) is combined with its weight vector w to calculate the comprehensive risk score S: For different types of indicators, normalization may be necessary to ensure that all indicators are calculated under the same dimensions. For example, the maximum flexion-extension angle of the knee joint... It can be done through the formula Normalization is performed, where and These are the minimum and maximum values of the maximum flexion and extension angle of the knee joint in the training dataset, respectively.
[0043] S3: Risk levels are determined based on the numerical range of the comprehensive risk score S; when At that time, it was determined to be low risk; when At that time, it was determined to be a medium-risk area; when At that time, it was determined to be high risk.
[0044] in, and The pre-set threshold can be determined by analyzing the training dataset, combined with clinical experience or statistical methods.
[0045] S4: The calculated comprehensive risk score S is used as one of the input features, along with other original feature parameters x, and fed into the Support Vector Machine (SVM) model. The model's predictions are compared with the actual damage situation, and the risk indicator weights w and risk level thresholds are adjusted using methods such as cross-validation. and In addition, the parameters of the machine learning model (such as the kernel function parameters of SVM) are continuously optimized to improve the accuracy and reliability of the risk assessment model.
[0046] In a specific implementation, the SVM risk assessment model is integrated with an attention mechanism. Building upon the existing method of optimizing the SVM model by inputting the comprehensive risk score S and the original feature parameters x, an attention mechanism is introduced. The specific steps are as follows: (1) Construct the attention weight matrix Combine the comprehensive risk score S and the original feature parameters Composition of joint feature vectors Through a learnable weight matrix (where n is the defined attention dimension) Perform matrix multiplication with the joint feature vector, and then pass it through an activation function (such as the ReLU function) to obtain the attention score vector. Then, the attention score vector is transformed into an attention weight vector using the Softmax function. Make each weight between 0 and 1 and sum to 1: ,in, , which correspond to the comprehensive risk score S and the attention weight of each original feature parameter, respectively. The larger the weight, the more important the feature is for assessing the risk of knee injury.
[0047] (2) Attention-based feature fusion
[0048] Using the calculated attention weight vector For joint eigenvectors We perform a weighted summation to obtain the feature vector of the fused attention mechanism. This feature vector highlights more critical features for assessing the risk of knee injury, providing a more discriminative input for subsequent SVM models.
[0049] (3) SVM model training and optimization
[0050] Feature vectors that incorporate attention mechanisms The input is fed into a Support Vector Machine (SVM) model for training. During training, the error between the model's predictions and the actual damage (e.g., using mean squared error, MSE) is used as the loss function L. Where N is the number of samples, For the actual damage label of the k-th sample, The damage labels predicted by the model.
[0051] A 5-fold cross-validation approach was employed, dividing the dataset into five subsets. Four subsets were used alternately as the training set, and the remaining subset as the test set, for multiple training and evaluation cycles. During each training iteration, stochastic gradient descent (SGD) was used to update the risk indicator weights w and risk level thresholds based on the gradient calculated from the loss function L. and Furthermore, by continuously reducing the loss function value and optimizing the accuracy and reliability of the risk assessment model, the kernel function parameter gamma of the SVM model is improved.
[0052] This embodiment incorporates an attention mechanism, enabling the model to more intelligently focus on key risk features and improve the accuracy of knee injury risk assessment.
[0053] In a specific implementation, the recommended exercise posture adjustment includes: analyzing the kinematic characteristics of high-risk subjects, such as the maximum flexion-extension angle and varus-valgus angle of the knee joint. If the maximum flexion-extension angle of the knee joint exceeds a threshold... Calculate the difference between the actual angle and the threshold. Based on the magnitude of the difference, suggestions for adjusting movement posture are generated. When the knee is small, it is recommended to actively control the range of knee flexion and extension during exercise, reducing the range of motion with each flexion and extension. The range; when When the condition is more severe, it is recommended to seek guidance from a professional sports coach for targeted posture correction training. The training intensity can be set at 3-5 times per week, 30-60 minutes each time.
[0054] Specific muscle training recommendations include: assessing the functional status of muscles around the knee joint by analyzing characteristics such as muscle activation time, peak EMG value, and muscle synergistic contraction patterns in electromyography (EMG) data. If the synergistic contraction ratio of the quadriceps and hamstrings deviates significantly from the normal range, the deviation value should be calculated. A specific muscle training plan will be developed based on the deviation value. If the quadriceps are relatively underactivated, strength training for the quadriceps, such as squats, can be recommended. The training intensity can be adjusted according to the subject's physical condition, following a formula. (k1 is an empirical coefficient, set to 0.5-1 according to the actual situation) The number of training sessions is 2-3 times per week, each training session contains 3-4 sets, and each set is repeated 8-12 times.
[0055] Recommendations for reducing high-risk exercise include: analyzing the subjects' exercise data, combined with kinetic parameters such as the duration and impulse of the ground reaction force, to identify exercise patterns that may lead to high stress on the knee joint. For example, if the impulse of the ground reaction force during jumping... Exceeding the threshold Calculate the excess ratio Based on the proportion exceeding the limit, recommendations are given to reduce high-risk activities. When p is small, it is recommended to reduce the frequency of the high-risk activity, such as reducing the number of jumping exercises per week from n times to [missing information]. When p is high, it is recommended to temporarily stop the high-risk exercise until knee joint function improves.
[0056] In a specific implementation, a knee injury prevention assessment system based on biomechanical analysis, such as... Figure 2 As shown, it includes: Data acquisition module: Used to collect the data to be measured during the subject's movement; Biomechanical model construction module: used to construct, calibrate, and validate a biomechanical model using the test data; Feature parameter extraction module: used to extract feature parameters based on the test data and biomechanical model; Model building module: used to establish a knee joint injury risk assessment index system and prediction model based on the extracted feature parameters; Assessment module: Used to assess the risk of knee injury in subjects based on a risk assessment index system and a prediction model.
[0057] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.
[0058] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for preventive assessment of knee joint injuries based on biomechanical analysis, characterized in that, include: Collect the data to be measured during the subject's movement; A biomechanical model was constructed using the test data and then calibrated and validated. Feature parameters are extracted based on the test data and biomechanical model; Based on the extracted feature parameters, a knee joint injury risk assessment index system and prediction model were established. The risk of knee injury to the subjects was assessed based on a risk assessment index system and a prediction model.
2. The method for preventive assessment of knee joint injury based on biomechanical analysis according to claim 1, characterized in that, The data collected during the subject's movement includes: kinematic data, kinetic data, and electromyographic data. A motion capture system is used to collect kinematic data of the knee joint when the subject performs a specific movement task; force sensors are used to measure ground reaction force and the internal and external forces borne by the knee joint; miniature force sensors are installed at the muscle attachment points around the knee joint to measure the effect of muscle contraction on the knee joint; and a surface electromyography device is used to collect electrical activity signals of the muscles around the knee joint.
3. The method for preventive assessment of knee joint injury based on biomechanical analysis according to claim 2, characterized in that, Also includes: In three-dimensional space, the knee joint angle is calculated based on the three-dimensional coordinate changes of reflective markers placed on key parts of the body. The angular velocity is calculated by the rate of change of the joint angle with respect to time, and the angular acceleration is calculated by the rate of change of the angular velocity with respect to time. The raw electromyography signals collected by the surface electromyography instrument are processed, and the root mean square value is calculated to represent the degree of muscle activation.
4. The method for preventive assessment of knee joint injury based on biomechanical analysis according to claim 1, characterized in that, The construction of the biomechanical model includes: In constructing the multibody dynamics model, the equations of motion are established based on the Newton-Euler equations: ； Where M is the generalized mass matrix and q is the generalized coordinate vector. and These are the generalized velocity and generalized acceleration vectors, respectively. The matrix of Coriolis force and centrifugal force. The gravity vector As a generalized force vector, this equation is used to simulate the mechanical behavior of the knee joint under different motion states.
5. The method for preventive assessment of knee joint injury based on biomechanical analysis according to claim 1, characterized in that, The calibration and verification process includes: using the mean squared error (MSE) to evaluate the difference between the model prediction results and the actual measurement data, as shown in the formula: ； in, These are actual measured values. The model predicts the value, and n is the number of data samples. By adjusting the model parameters, the MSE is minimized to reflect the actual mechanical properties of the knee joint.
6. The method for preventive assessment of knee joint injury based on biomechanical analysis according to claim 1, characterized in that, The feature parameter extraction based on the test data and biomechanical model includes: extracting the maximum flexion-extension angle and peak acceleration from kinematic features, extracting the maximum force and torque from dynamic features, and extracting muscle activation time and peak electromyographic value from electromyographic features.
7. The method for preventive assessment of knee joint injury based on biomechanical analysis according to claim 1, characterized in that, The knee joint injury risk assessment index system includes: Set the maximum flexion-extension angle threshold of the knee joint as The threshold for inward and outward folding torque is The muscle activation delay threshold is A risk level is determined to be high when the following conditions are met: Maximum knee flexion and extension angle: ; Inward and outward turning torque: ; Delayed muscle activation: ,in, This represents the actual measured muscle activation delay time.
8. The method for preventive assessment of knee joint injury based on biomechanical analysis according to claim 1, characterized in that, The prediction model includes: Constructing a Vector Machine (SVM) for a binary classification problem, its decision function is: ； in For the weight vector, For the input feature vector, For bias terms, The sign function is optimized by training on data from subjects with known injuries and no injuries. and The value of .
9. A method for preventive assessment of knee joint injury based on biomechanical analysis according to claim 8, characterized in that, Also includes: A judgment matrix is constructed using the analytic hierarchy process (AHP), and the relative importance of different risk indicators is determined through expert scoring, and a weight vector is calculated. The extracted feature parameters are combined with the weight vector to calculate the comprehensive risk score; Risk levels are determined based on the range of values in the comprehensive risk score. The calculated comprehensive risk score is used as the input feature and is fed into the support vector machine (SVM) along with the original feature parameters. The model's prediction results are compared with the actual damage situation, and the risk index weights, risk level thresholds, and machine learning model parameters are adjusted through cross-validation.
10. A knee joint injury preventive assessment system based on biomechanical analysis, characterized in that, include: Data acquisition module: Used to collect the data to be measured during the subject's movement; Biomechanical model construction module: used to construct, calibrate, and validate a biomechanical model using the test data; Feature parameter extraction module: used to extract feature parameters based on the test data and biomechanical model; Model building module: used to establish a knee joint injury risk assessment index system and prediction model based on the extracted feature parameters; Assessment module: Used to assess the risk of knee injury in subjects based on a risk assessment index system and a prediction model.