Children and teenager foot spine deformity online movement intervention method based on multi-dimensional evaluation

Abstract

The invention relates to a multi-dimensional evaluation-based online motion intervention method for foot spine deformity of children and teenagers. The method comprises the following steps: acquiring three-dimensional foot data of children and adolescents, performing feature extraction and fusion operation on the data to obtain a foot fusion feature value, and performing violent matching based on the foot fusion feature value so as to obtain a corresponding feature parameter combination. And inputting the characteristic parameter combination into a trained foot evaluation model, and evaluating the foot spine condition of the children and adolescents to obtain a foot spine condition evaluation result. Then, a time sequence prediction model is constructed according to the time sequence prediction model, and a motion intervention scheme is generated. Based on the scheme, the correction effect of the plantar orthosis is evaluated by applying a multi-objective optimization algorithm, and design parameters of the orthosis are determined. And finally, according to the updated foot three-dimensional data and orthosis design parameters, optimizing and adjusting the exercise intervention scheme to obtain an updated exercise intervention scheme. The scientificity, accuracy and effectiveness of evaluation and intervention of the foot spine deformity of children and adolescents are greatly improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical rehabilitation, and particularly relates to an online exercise intervention method for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation. Background Art

[0002] With the development of medical rehabilitation technology, an online exercise intervention technology for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation has emerged. Children and adolescents are in a stage of rapid physical growth and development, and foot and spinal health is crucial. In recent years, the problem of foot and spinal deformities in children and adolescents has become increasingly prominent, which not only affects their physical appearance, but may also cause health problems such as pain and movement disorders, and even affect cardiopulmonary function and mental health in severe cases. Traditional means of evaluating and intervening in foot and spinal deformities mostly rely on manual examinations in offline professional medical institutions, which have limitations such as low efficiency, high cost, and difficulty in dynamically tracking the changes of the condition in real time. It is difficult to achieve accurate evaluation and efficient online intervention for the foot and spinal conditions of children and adolescents. Summary of the Invention

[0003] Based on this, in view of the above technical problems, it is necessary to provide an online exercise intervention method for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation that can achieve accurate evaluation and efficient online intervention for the foot and spinal conditions of children and adolescents.

[0004] In a first aspect, the present application provides an online exercise intervention method for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation, including:

[0005] Obtaining three-dimensional data of the feet of children and adolescents; the three-dimensional data of the feet includes three-dimensional information of the foot contour, joint range of motion, and plantar pressure distribution.

[0006] Performing feature extraction and fusion on the three-dimensional data of the feet to obtain a fused foot feature value; performing brute-force matching based on the fused foot feature value to obtain a corresponding combination of feature parameters.

[0007] Inputting the combination of feature parameters into a trained foot evaluation model to evaluate the foot and spinal conditions of children and adolescents, and obtaining an evaluation result of the foot and spinal conditions; the evaluation result of the foot and spinal conditions is used to determine whether there is a foot and spinal deformity and the type and degree of the deformity.

[0008] Constructing a time series prediction model according to the evaluation result of the foot and spinal conditions to predict the trend of the disease change, and generating an exercise intervention plan; the exercise intervention plan is used to supervise and guide the user to complete the online training content in real time through a video.

[0009] Evaluating the correction effect of the plantar orthosis based on the exercise intervention plan using a multi-objective optimization algorithm to obtain orthosis design parameters.

[0010] Optimize and adjust the exercise intervention plan according to the updated three-dimensional foot data and orthosis design parameters to obtain the updated exercise intervention plan.

[0011] In one embodiment, feature extraction and fusion are performed on the three-dimensional foot data to obtain foot fusion feature values, including:

[0012] Preprocess the three-dimensional foot data using a moving average filtering algorithm to obtain processed foot morphology data.

[0013] Perform enhancement processing on the processed foot morphology data based on the Laplacian operator algorithm to obtain enhanced foot morphology data.

[0014] Use a convolutional neural network to construct a foot feature model to locate the enhanced foot morphology data and obtain the target data region.

[0015] Perform feature extraction and marking on the target data region based on the Scale-Invariant Feature Transform (SIFT) algorithm to obtain foot feature points; the foot feature points include at least one of the morphology, movement, and pressure characteristics of the arch, heel, and metatarsophalangeal joints.

[0016] Perform feature fusion on the foot feature points using a feature fusion formula to obtain foot fusion feature values.

[0017] In one embodiment, perform feature fusion on the foot feature points using a feature fusion formula to obtain foot fusion feature values, including:

[0018] Use the following feature fusion formula to calculate and obtain foot fusion feature values:

[0019]

[0020] where H represents the foot fusion feature value, ω i represents the weight coefficient of the i-th foot feature, S i represents the value of the i-th foot feature, x represents the spatial position of the foot feature point, μ i represents the mean of the i-th foot feature, σ i represents the standard deviation of the i-th foot feature, and n represents the number of foot features.

[0021] In one embodiment, perform brute-force matching based on the foot fusion feature values to obtain corresponding feature parameter combinations, including:

[0022] Obtain the reference three-dimensional foot data of children and adolescents under normal conditions and preprocess the reference three-dimensional foot data.

[0023] Extract feature points from the foot fusion eigenvalue of the three-dimensional foot data and the reference three-dimensional foot data using the LDB descriptor to obtain the fusion descriptor corresponding to the three-dimensional foot data and the benchmark descriptor corresponding to the reference three-dimensional foot data.

[0024] Perform brute-force matching on the fusion descriptor and the benchmark descriptor to obtain candidate matching pairs.

[0025] Use the improved LMedS algorithm to remove the incorrect matches in each candidate matching pair to obtain the best matching pair.

[0026] Use the variation ratio to perform a secondary elimination on the matches with large feature differences in the best matching pair to obtain a feature parameter combination; the feature parameter combination includes the foot shape contour deviation parameter, the joint range of motion abnormality parameter, and the plantar pressure distribution difference parameter.

[0027] In one embodiment, input the feature parameter combination into the trained foot evaluation model to evaluate the foot and spine condition of children and adolescents, and obtain the foot and spine condition evaluation result, including:

[0028] Perform denoising and outlier processing on the feature parameter combination data to obtain matching training data.

[0029] Based on the matching training data, use a machine learning classifier for model training to obtain a foot evaluation model to be trained; the foot evaluation model is at least one of a support vector machine, a decision tree, a neural network, and a random forest.

[0030] Based on the historical training data, calculate the AUC value of the foot evaluation model to be trained using a formula to obtain the trained foot evaluation model; the foot evaluation model is a foot evaluation model whose performance meets the preset conditions.

[0031] Input the matching training data into the trained foot evaluation model to evaluate the foot and spine condition of children and adolescents, and obtain the foot and spine condition evaluation value.

[0032] Perform data annotation on the foot and spine condition evaluation value to obtain a foot and spine condition evaluation result for discriminating the foot and spine condition.

[0033] In one embodiment, construct a time series prediction model based on the foot and spine condition evaluation result to predict the disease progression trend and generate a motion intervention plan, including:

[0034] Input the feature parameters related to the disease condition in the foot and spine condition evaluation result into the time series prediction model and calculate using a formula to obtain the predicted disease deterioration probability value.

[0035] Use the following formula for calculation to obtain the predicted disease deterioration probability value:

[0036]

[0037] Among them, P(t) represents the probability value of disease deterioration at time t, n represents the number of characteristic parameters, represents the weight coefficient of the j-th characteristic parameter, x j (t) represents the value of the j-th characteristic parameter at time t, θ j represents the reference value of the j-th characteristic parameter, β j represents the attenuation coefficient, and α represents the normalization coefficient.

[0038] According to the predicted probability value of disease deterioration and combined with the key indicators of the foot arch condition, a preliminary exercise intervention plan is generated.

[0039] Using a machine learning algorithm to classify the preliminary exercise intervention plan to obtain the exercise intervention levels corresponding to different deformity types.

[0040] Among them, if the deformity type is flat feet, the relevant indicators of arch support are extracted to optimize the arch training content in the preliminary exercise intervention plan.

[0041] If the deformity type is scoliosis, the relevant indicators of spinal correction are extracted to optimize the spinal training content in the preliminary exercise intervention plan.

[0042] Based on the exercise intervention level, verify the matching degree between the preliminary exercise intervention plan and the probability value of disease deterioration to obtain the final exercise intervention plan.

[0043] In one of the embodiments, based on the exercise intervention plan, use a multi-objective optimization algorithm to evaluate the correction effect of the foot orthosis to obtain the orthosis design parameters, including:

[0044] Obtain the initial design parameters of the foot orthosis.

[0045] According to the initial design parameters and combined with the key indicators in the exercise intervention plan, construct a multi-objective optimization model.

[0046] According to the expected correction effect of the foot orthosis, use the multi-objective optimization model to construct the objective to obtain the optimization objective function; the objectives include arch support force and gait balance.

[0047] Based on the optimization objective function, use the genetic algorithm to iteratively optimize the design parameters of the foot orthosis to generate multiple groups of candidate design schemes.

[0048] According to the three-dimensional foot data, evaluate the correction effect of each group of candidate design schemes, and screen to obtain the optimal design scheme.

[0049] According to the optimal design scheme, use the test data to verify the matching degree of the correction effect of the optimized orthosis to obtain the orthosis design parameters.

[0050] In a second aspect, the present application also provides an online exercise intervention system for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation. The system includes:

[0051] A data acquisition module for acquiring three-dimensional data of the feet of children and adolescents; the three-dimensional data of the feet includes three-dimensional information on the external contour of the feet, the range of joint movement, and the plantar pressure distribution.

[0052] A foot and spine evaluation module for extracting and fusing features from the three-dimensional data of the feet to obtain a fused foot feature value; also for performing brute-force matching based on the fused foot feature value to obtain a corresponding combination of feature parameters; and also for inputting the combination of feature parameters into a trained foot evaluation model to evaluate the foot and spine condition of children and adolescents, obtaining an evaluation result of the foot and spine condition; the evaluation result of the foot and spine condition is used to determine whether there is a foot and spine deformity and the type and degree of the deformity.

[0053] An orthosis design module for constructing a time series prediction model based on the evaluation result of the foot and spine condition to predict the trend of the disease condition and generating an exercise intervention plan; the exercise intervention plan is used to supervise and guide the user to complete the online training content in real time through a video; and also for evaluating the correction effect of the plantar orthosis using a multi-objective optimization algorithm based on the exercise intervention plan to obtain orthosis design parameters.

[0054] An intervention generation module for optimizing and adjusting the exercise intervention plan according to the updated three-dimensional data of the feet and the orthosis design parameters to obtain an updated exercise intervention plan.

[0055] In a third aspect, the present application also provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the foregoing method is implemented.

[0056] In a fourth aspect, the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the foregoing method is implemented.

[0057] The above-mentioned online exercise intervention method for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation obtains three-dimensional data of children and adolescents' feet containing three-dimensional information of foot external contour, joint range of motion, and plantar pressure distribution, and then conducts feature extraction and fusion operations on the data to obtain foot fusion feature values, and performs brute-force matching based on this to obtain the corresponding combination of feature parameters. The combination of feature parameters is input into the trained foot evaluation model to evaluate the foot and spinal conditions of children and adolescents, and the foot and spinal condition evaluation results for judging whether there are foot and spinal deformities, their deformity types and degrees are obtained. Subsequently, a time series prediction model is constructed based on this evaluation result to predict the trend of disease changes, and an exercise intervention plan is generated accordingly. The exercise intervention plan is used to supervise and guide users to complete the online training content in real time through videos. Then, based on this exercise intervention plan, a multi-objective optimization algorithm is used to evaluate the correction effect of the foot orthosis to determine the orthosis design parameters. Finally, according to the subsequent updated three-dimensional data of the feet and the determined orthosis design parameters, the exercise intervention plan is optimized and adjusted to obtain a more perfect updated exercise intervention plan. It greatly improves the scientificity, accuracy, and effectiveness of the evaluation and intervention of children and adolescents' foot and spinal deformities, and helps to improve their foot and spinal health conditions. Description of the Drawings

[0058] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0059] Figure 1 It is a flowchart of an online exercise intervention method for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation provided by an embodiment of the present invention;

[0060] Figure 2 It is a flowchart of feature extraction and fusion of three-dimensional foot data to obtain foot fusion feature values provided by an embodiment of the present invention;

[0061] Figure 3 It is a flowchart of brute-force matching based on foot fusion feature values to obtain the corresponding combination of feature parameters provided by an embodiment of the present invention;

[0062] Figure 4 It is a flowchart of inputting the combination of feature parameters into the trained foot evaluation model to evaluate the foot and spinal conditions of children and adolescents to obtain the foot and spinal condition evaluation results provided by an embodiment of the present invention;

[0063] Figure 5The structural block diagram of the online exercise intervention system for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation provided by the embodiments of the present invention. Detailed implementation manners

[0064] In order to make the objectives, technical solutions and advantages of the present application more clear, the following further describes the present application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0065] First, the implementation environment of the embodiments of the present application is described. Exemplarily, the implementation environment includes a data acquisition device, a data processing and analysis device, and an exercise intervention device.

[0066] In the online exercise intervention method for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation, data acquisition devices, such as 3D scanners, plantar pressure detectors, motion sensors, and depth cameras, transmit the collected data related to the feet and spines of children and adolescents to the data processing and analysis device. After the data processing and analysis device completes the in-depth analysis of the foot and spine data and generates a personalized exercise intervention plan, through network connection methods such as the Internet, it transmits the corresponding instructions and training content to the exercise intervention device. At the same time, the exercise intervention device will also feedback the exercise execution situation back to the data processing and analysis device through the network to ensure the accurate implementation of the online exercise intervention method.

[0067] The data acquisition device includes a 3D scanner that can accurately perform three-dimensional modeling of the bodies of children and adolescents, and can carefully capture key data such as the shape, curvature of the spine, and the shape and arch height of the feet. The plantar pressure detector accurately collects the plantar pressure distribution of children during various daily activities through special pressure-sensitive materials and array sensors to reveal the structure and function status of the feet. Motion sensors, such as wearable devices like accelerometers and gyroscopes, can monitor the movements and posture changes of the limbs during exercise in real time, providing dynamic data support for subsequent analysis. The depth camera integrates color and infrared imaging technologies to achieve high-resolution scanning of the body posture, obtain detailed body contours and posture information, and provide a comprehensive and accurate data basis for the evaluation of foot and spinal deformities from multiple aspects.

[0068] Data processing and analysis devices are responsible for running complex data analysis software and artificial intelligence algorithms to preliminarily sort, clean, and extract features from the massive amount of foot and spine data transmitted by data collection devices. In large-scale application scenarios, they can securely store the foot and spine data of a large number of patients, and through a distributed computing architecture, efficiently run various data analysis models and algorithms to deeply explore the laws behind the foot and spine data, identify the types, degrees, and development trends of foot and spine deformities. The artificial intelligence chips integrated in some advanced devices, such as deep learning processors (DPUs) or graphics processing units (GPUs), further accelerate the operation process of artificial intelligence algorithms, significantly improving the efficiency of data processing and analysis, and providing a solid technical guarantee for formulating personalized exercise intervention programs.

[0069] Exercise intervention devices, such as smart wearable devices, rely on their portability and real-time interaction characteristics. On the one hand, they receive exercise instructions issued by data processing and analysis devices and accurately remind children and adolescents to carry out exercise training according to the plan. On the other hand, they can monitor physiological data such as exercise status, steps, and heart rate in real time and promptly feedback it to the data processing and analysis devices. Virtual reality (VR) / augmented reality (AR) devices create an immersive virtual training environment. Using technologies such as 3D modeling and spatial positioning, they enable children and adolescents to intuitively and accurately understand and execute correct exercise postures and movement essentials, greatly enhancing the fun and compliance of exercise training. Home fitness equipment provides the basic conditions for children and adolescents to carry out exercise intervention training at home. Cooperating with online guidance programs, it helps to enhance muscle strength, improve joint mobility and body balance ability, and comprehensively promote the rehabilitation and improvement of foot and spine functions.

[0070] Combined with the above implementation environment, the application scenarios of the embodiments of this application are described.

[0071] The online exercise intervention method for children and adolescents' foot and spine deformities based on multi-dimensional evaluation provided by the embodiments of this application uses advanced sensor technology and artificial intelligence algorithms to collect and analyze data from multiple dimensions such as body posture, bone development, and muscle strength, and accurately evaluate the foot and spine deformity conditions of children and adolescents. On this basis, the method customizes personalized exercise intervention programs for patients through an online platform and uses methods such as video guidance and real-time feedback to enable patients to receive professional exercise intervention guidance at home to help improve foot and spine deformities and promote healthy development. Exemplarily, the online exercise intervention method for children and adolescents' foot and spine deformities based on multi-dimensional evaluation provided by the embodiments of this application can be applied to at least one of the following scenarios including but not limited to.

[0072] First, the online exercise intervention method for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation is applied to the medical institution scenario. In medical institutions, data collection devices include professional 3D scanners that can accurately obtain detailed three-dimensional data of children and adolescents' spines and feet, assisting doctors in accurately judging the types and degrees of foot and spinal deformities. The plantar pressure detector can collect plantar pressure distribution data by simulating actions such as walking and standing of patients, providing strong evidence for diagnosis. The data is quickly transmitted to data processing and analysis devices, such as high-performance computers and professional servers. Doctors use advanced data analysis software and artificial intelligence algorithms to deeply analyze the data and accurately formulate personalized exercise intervention plans. Subsequently, exercise intervention devices become the key to plan implementation. Doctors use smart wearable devices to set exercise reminders and monitoring parameters for patients to ensure that patients can carry out rehabilitation training on time. At the same time, virtual reality devices can simulate professional rehabilitation scenarios, allowing patients to perform foot and spinal rehabilitation exercises in a realistic environment. Doctors can also observe the patients' exercise status in real time and adjust the plan according to the feedback data to ensure the effectiveness and safety of rehabilitation treatment.

[0073] Second, the online exercise intervention method for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation is applied to school physical examinations. Schools can deploy data collection devices to conduct foot and spinal health screenings on student groups. The combination of depth cameras and motion sensors can non-invasively collect students' body posture data in scenarios such as students' daily break activities and physical education classes. The data is aggregated to the school's data processing and analysis devices, and the overall foot and spinal health status of the student group is evaluated by means of big data analysis technology to identify students at potential risk of foot and spinal deformities. For these students, schools cooperate with parents and medical institutions to carry out personalized interventions using exercise intervention devices. For example, students are equipped with smart wearable devices to monitor their exercise training situations at school and at home to ensure the continuity and standardization of training. At the same time, schools can integrate exercise training content related to foot and spinal health into physical education courses by means of virtual reality or augmented reality technology to improve students' attention and participation in foot and spinal health, and prevent and improve foot and spinal deformities in children and adolescents at the group level.

[0074] In one embodiment, as Figure 1 shown, the present application provides an online exercise intervention method for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation, which may include the following steps:

[0075] Step S101, obtaining three-dimensional data of the feet of children and adolescents; the three-dimensional data of the feet includes three-dimensional information on the outer contour of the feet, the range of joint movement, and the plantar pressure distribution.

[0076] By using advanced laser scanning technology and high-precision pressure sensing equipment, the foot's contour can be recorded in detail, accurately presenting the overall shape and size details of the foot. With the help of motion capture systems and professional joint angle measurement instruments, the range of joint movement can be accurately obtained, and the movement trajectory and angle changes of each joint of the foot under different movements can be clearly defined. At the same time, using special pressure insoles with data collection software, the three-dimensional information of the plantar pressure distribution is collected in all directions, clearly presenting the pressure changes in each area of ​​the plantar under static and dynamic conditions.

[0077] Step S102, extracting and fusing the three-dimensional foot data to obtain foot fusion feature values; performing brute force matching based on the foot fusion feature values ​​to obtain corresponding feature parameter combinations.

[0078] First, the signal processing algorithm and machine learning model are used to extract features from the three-dimensional data including the foot contour, joint range of motion, and plantar pressure distribution. By mining the features of data in different dimensions, representative feature information is screened out. Subsequently, feature fusion technology is used to organically integrate the features extracted from each dimension to generate a foot fusion feature value. This fusion feature value comprehensively and concisely summarizes the key information of the three-dimensional foot data. Based on this foot fusion feature value, a brute force matching algorithm is used to compare it in a huge feature database, and the corresponding feature parameter combination is accurately found through one-by-one matching.

[0079] Step S103, input the characteristic parameter combination into the trained foot assessment model to assess the foot spine condition of children and adolescents, and obtain the foot spine condition assessment result; the foot spine condition assessment result is used to determine whether there is foot spine deformity and the type and degree of deformity.

[0080] The obtained characteristic parameter combination is imported as key input data into the foot assessment model that has been trained with a large amount of sample data and optimized for performance, and the foot spine condition assessment result is output. This result can not only clearly determine whether children and adolescents have foot spine deformity problems, but also rely on the model's ability to accurately identify various deformity characteristics to clearly define the specific type of deformity, such as differentiating between flat feet, high arched feet, scoliosis, etc., and at the same time use quantitative indicators to accurately measure the severity of the deformity.

[0081] Step S104, constructing a time series prediction model based on the results of the foot spine condition assessment to predict the disease change trend and generate an exercise intervention plan; the exercise intervention plan is used to supervise and guide users to complete online training content in real time through video.

[0082] After obtaining a detailed and accurate assessment result of the foot and spine condition, a targeted time series prediction model is constructed. This model uses advanced time series analysis algorithms to deeply learn and mine a large amount of historical data on the development of foot and spine diseases, so as to comprehensively and dynamically consider the changing characteristics of the disease over time. Through this model, the development trend of the foot and spine disease in the next period can be accurately predicted, including whether the disease tends to remit, stabilize or deteriorate. Based on the prediction results, considering various factors such as the physical function, motor ability and daily living habits of children and adolescents, a professional rehabilitation team formulates an individualized, scientific, reasonable and feasible exercise intervention plan. This plan clearly stipulates key elements such as the type, intensity, frequency and duration of exercise, aiming to effectively improve the foot and spine condition of children and adolescents through reasonable exercise training and promote the recovery of their foot and spine health.

[0083] Step S105, based on the exercise intervention plan, use a multi-objective optimization algorithm to evaluate the correction effect of the foot orthosis and obtain the orthosis design parameters.

[0084] First, according to the generated exercise intervention plan, fully consider elements such as the exercise type, intensity involved in the plan and the daily activity characteristics of children and adolescents. Then introduce a multi-objective optimization algorithm, which comprehensively weighs multiple key objectives such as correction effect, wearing comfort, and orthosis durability. For the foot orthosis, by constructing a detailed mathematical model, simulate its correction effect when implementing the exercise intervention plan under different combinations of design parameters (such as material properties, the shape and size of the support structure, the thickness of the insole buffer layer, etc.). The algorithm conducts intelligent search and iterative calculation in a large parameter combination space, and quantitatively evaluates the correction effect corresponding to each parameter combination. After multiple rounds of optimization operations, select the parameter combination that can achieve the optimal balance of multiple objectives, and finally obtain accurate orthosis design parameters to ensure that the designed foot orthosis can play the maximum correction role during the exercise intervention process.

[0085] Step S106, optimize and adjust the exercise intervention plan according to the updated three-dimensional foot data and orthosis design parameters to obtain an updated exercise intervention plan.

[0086] The three-dimensional foot data collected and updated regularly will reflect the dynamic changes in the foot's shape and function in real time. At the same time, the established orthosis design parameters will also be adjusted due to actual usage feedback and technological improvements. Based on the updated three-dimensional foot data, it is possible to accurately observe the latest conditions of the foot's external contour, joint range of motion, and plantar pressure distribution. The changes in the orthosis design parameters are directly related to the way it supports and corrects the foot. Considering these two aspects comprehensively and using professional data analysis tools and sports rehabilitation knowledge, the initial exercise intervention plan is comprehensively reviewed and optimized. From the selection of sports items, the reasonable increase or decrease of exercise intensity, to the fine adjustment of exercise frequency and duration, ensure that the updated exercise intervention plan closely fits the actual needs of children and adolescents' feet and spine that are constantly changing, and continuously provide the most appropriate and effective exercise guidance for improving their foot and spine health conditions.

[0087] The above-mentioned online exercise intervention method for children and adolescents with foot and spine deformities based on multi-dimensional evaluation, after obtaining the three-dimensional foot data of children and adolescents containing three-dimensional information such as the foot's external contour, joint range of motion, and plantar pressure distribution, performs feature extraction and fusion operations on the data to obtain the foot fusion feature value, and based on this, conducts brute-force matching to further obtain the corresponding combination of feature parameters. Input this combination of feature parameters into the trained foot evaluation model to evaluate the foot and spine conditions of children and adolescents, and obtain the foot and spine condition evaluation results for judging whether there are foot and spine deformities, their deformity types and degrees. Subsequently, based on this evaluation result, a time series prediction model is constructed to predict the trend of the disease, and an exercise intervention plan is generated accordingly. Then, based on this exercise intervention plan, the exercise intervention plan is used to supervise and guide users to complete the online training content in real time through videos. Use a multi-objective optimization algorithm to evaluate the correction effect of the plantar orthosis and determine the orthosis design parameters. Finally, according to the subsequent updated three-dimensional foot data and the determined orthosis design parameters, the exercise intervention plan is optimized and adjusted to obtain a more perfect updated exercise intervention plan.

[0088] In one embodiment, as Figure 2 shown, the feature extraction and fusion of the three-dimensional foot data to obtain the foot fusion feature value may include the following steps:

[0089] Step S201, preprocess the three-dimensional foot data using the moving average filtering algorithm to obtain the processed foot shape data.

[0090] Step S202, perform enhancement processing on the processed foot shape data based on the Laplacian operator algorithm to obtain the enhanced foot shape data.

[0091] Step S203, use a convolutional neural network to construct a foot feature model to locate the enhanced foot shape data to obtain the target data area.

[0092] Step S204: Based on the Scale Invariant Feature Transform (SIFT) algorithm, perform feature extraction and marking on the target data region to obtain foot feature points. The foot feature points include at least one of the morphology, movement, and pressure characteristics of the arch, heel, and metatarsophalangeal joints.

[0093] Step S205: Use the feature fusion formula to fuse the foot feature points to obtain the foot fusion feature value.

[0094] Specifically, first, use the moving average filtering algorithm to preprocess the three-dimensional foot data containing the three-dimensional information of the foot outline, joint range of motion, and plantar pressure distribution. By smoothing the data sequence, effectively remove noise interference, so as to obtain stable and reliable processed foot morphology data. Immediately afterwards, based on the Laplacian operator algorithm, perform enhancement processing on the processed data to highlight the edges and detail features in the data, improve the recognition rate of key information, and thus obtain the enhanced foot morphology data. Subsequently, use a convolutional neural network to construct a foot feature model, conduct in-depth analysis and positioning on the enhanced foot morphology data, and accurately identify the target data region containing important features. Then, based on the Scale Invariant Feature Transform (SIFT) algorithm, perform feature extraction and marking operations on the target data region, and successfully obtain foot feature points covering at least one of the morphology, movement, and pressure characteristics of the arch, heel, and metatarsophalangeal joints. Finally, use a carefully designed feature fusion formula to organically fuse the foot feature points, and finally obtain the foot fusion feature value that comprehensively reflects the foot condition.

[0095] In one embodiment, using the feature fusion formula to fuse the foot feature points to obtain the foot fusion feature value may include the following steps:

[0096] Use the following feature fusion formula to calculate and obtain the foot fusion feature value:

[0097]

[0098] where H represents the foot fusion feature value, ω i represents the weight coefficient of the i-th foot feature, S i represents the value of the i-th foot feature, x represents the spatial position of the foot feature point, μ i represents the mean of the i-th foot feature, σ i represents the standard deviation of the i-th foot feature, and n represents the number of foot features.

[0099] Calculate the foot fusion feature value by applying a specific feature fusion formula. This formula can fully consider the importance, specific values, spatial positions, and data distribution characteristics of each foot feature, comprehensively and scientifically fuse the information. The generated foot fusion feature value can accurately and comprehensively reflect the foot condition, providing a highly reliable and representative data basis for the subsequent foot and spine condition assessment and intervention measure formulation in the online exercise intervention method for children and adolescents' foot and spine deformities based on multi-dimensional evaluation, greatly improving the scientificity and accuracy of the entire intervention process.

[0100] In one embodiment, as Figure 3 shown, perform brute-force matching based on the foot fusion feature value to obtain the corresponding feature parameter combination, which may include the following steps:

[0101] Step S301: Obtain the reference foot three-dimensional data of children and adolescents under normal conditions and preprocess the reference foot three-dimensional data.

[0102] Step S302: Use the LDB descriptor to extract feature points from the foot fusion feature values of the foot three-dimensional data and the reference foot three-dimensional data to obtain the fusion descriptor corresponding to the foot three-dimensional data and the reference descriptor corresponding to the reference foot three-dimensional data.

[0103] Step S303: Perform brute-force matching on the fusion descriptor and the reference descriptor to obtain candidate matching pairs.

[0104] Step S304: Use the improved LMedS algorithm to eliminate the incorrect matches in each candidate matching pair to obtain the best matching pair.

[0105] Step S305: Use the variation ratio to perform a secondary elimination on the matches with large feature differences in the best matching pair to obtain the feature parameter combination; the feature parameter combination includes the foot shape contour deviation parameter, the joint range of motion abnormality parameter, and the plantar pressure distribution difference parameter.

[0106] First, collect the reference three-dimensional foot data of children and adolescents under normal conditions. This data covers comprehensive information such as the foot's external contour, joint range of motion, and plantar pressure distribution, and perform preprocessing operations on it. Through means such as filtering and noise reduction, ensure the accuracy and reliability of the reference data. Subsequently, use the LDB descriptor to extract feature points for the foot fusion feature values of the three-dimensional foot data obtained through feature fusion and the foot fusion feature values of the reference three-dimensional foot data, respectively obtaining the fusion descriptor corresponding to the three-dimensional foot data and the reference descriptor corresponding to the reference three-dimensional foot data. The descriptors highly condense the key feature information of their respective data. Then, perform brute-force matching on the fusion descriptor and the reference descriptor. By comprehensively comparing all possible combinations, select the candidate matching pairs. To improve the matching accuracy, use the improved LMedS algorithm to eliminate the incorrect matches in each candidate matching pair, thereby obtaining a more accurate best matching pair. Finally, with the help of the variation ratio, perform secondary screening and elimination on the matches with large feature differences in the best matching pair, and successfully obtain a feature parameter combination including the foot external contour deviation parameter, joint range of motion abnormality parameter, and plantar pressure distribution difference parameter.

[0107] In one embodiment, as Figure 4 shown, input the feature parameter combination into the trained foot evaluation model to evaluate the foot and spine condition of children and adolescents, and obtain the foot and spine condition evaluation result. The steps may include the following:

[0108] Step S401, perform denoising and outlier processing on the feature parameter combination data to obtain the matching training data.

[0109] Step S402, based on the matching training data, use a machine learning classifier for model training to obtain the foot evaluation model to be trained; the foot evaluation model is at least one of a support vector machine, a decision tree, a neural network, and a random forest.

[0110] Step S403, calculate the AUC value of the foot evaluation model to be trained based on the historical training data using a formula to obtain the trained foot evaluation model; the foot evaluation model is a foot evaluation model whose performance meets the preset conditions.

[0111] Step S404, input the matching training data into the trained foot evaluation model to evaluate the foot and spine condition of children and adolescents, and obtain the foot and spine condition evaluation value.

[0112] Step S405, perform data annotation on the foot and spine condition evaluation value to obtain the foot and spine condition evaluation result for discriminating the foot and spine condition.

[0113] For the characteristic parameter combination data, professional data processing algorithms are used for denoising operations to remove the noise interference mixed in the data acquisition and processing process, and at the same time identify and eliminate outliers, so as to obtain higher-quality matching training data that better meets the requirements of model training. Immediately afterwards, based on the matching training data, at least one of the machine learning classifiers such as support vector machines, decision trees, neural networks, or random forests is selected to carry out model training work, and a foot evaluation model to be trained is initially constructed. The classifier can learn the complex association patterns between the foot arch conditions and characteristic parameters from the data. With the help of historical training data, the AUC value of the model to be trained is calculated according to a specific formula. By continuously optimizing the model parameters, a foot evaluation model with performance meeting the preset conditions is selected to ensure that the model has good accuracy and reliability. Subsequently, the matching training data is input into the trained foot evaluation model to evaluate the foot arch conditions of children and adolescents. The model outputs the foot arch condition evaluation value through calculation. Finally, data annotation is performed on the evaluation value. According to the established annotation rules and standards, the evaluation value is converted into a clear foot arch condition evaluation result to accurately determine whether children and adolescents have foot arch deformities, the types and degrees of deformities.

[0114] In one of the embodiments, a time series prediction model is constructed based on the foot arch condition evaluation result to predict the trend of disease progression, and a motion intervention plan can be generated, which may include the following steps:

[0115] Step S501, input the characteristic parameters related to the disease condition in the foot arch condition evaluation result into the time series prediction model and calculate using the formula to obtain the predicted probability value of disease deterioration.

[0116] The following formula is used for calculation to obtain the predicted probability value of disease deterioration:

[0117]

[0118] Among them, P(t) represents the probability value of disease deterioration at time t, n represents the number of characteristic parameters, represents the weight coefficient of the jth characteristic parameter, x j (t) represents the value of the jth characteristic parameter at time t, θ j represents the reference value of the jth characteristic parameter, β j represents the attenuation coefficient, and α represents the normalization coefficient.

[0119] Step S502, generate a preliminary motion intervention plan according to the predicted probability value of disease deterioration in combination with the key indicators of the foot arch condition.

[0120] Step S503, classify the preliminary motion intervention plan using machine learning algorithms to obtain the motion intervention levels corresponding to different deformity types.

[0121] Among them, if the deformity type is flat feet, extract the arch support-related indicators and optimize the arch training content in the preliminary exercise intervention plan.

[0122] If the deformity type is scoliosis, extract the spinal correction-related indicators and optimize the spinal training content in the preliminary exercise intervention plan.

[0123] Step S504, verify the matching degree between the preliminary exercise intervention plan and the disease deterioration probability value based on the exercise intervention level to obtain the final exercise intervention plan.

[0124] First, extract the characteristic parameters closely related to the disease from the foot and spine condition assessment results, input them into a carefully constructed time series prediction model, and calculate the predicted disease deterioration probability value using a specific formula. Subsequently, based on the predicted disease deterioration probability value, combined with the key indicators of the foot and spine condition, such as the key data in aspects such as the foot shape contour, joint range of motion, and plantar pressure distribution, formulate a preliminary exercise intervention plan. Then, use machine learning algorithms to classify the preliminary exercise intervention plan. For different deformity types, such as flat feet, focus on extracting the arch support-related indicators and optimizing the arch training content in the plan; if it is scoliosis, extract the spinal correction-related indicators and optimize the spinal training content, so as to obtain the exercise intervention levels corresponding to different deformity types. Finally, based on the exercise intervention level, verify the matching degree between the preliminary exercise intervention plan and the disease deterioration probability value, and through continuous adjustment and improvement, obtain the most suitable final exercise intervention plan.

[0125] This embodiment not only realizes the scientific prediction of the development of foot and spine deformity diseases, but also can accurately formulate personalized exercise intervention plans according to different deformity types, effectively improve the pertinence and effectiveness of the intervention plan, greatly increase the possibility of improving the foot and spine health status of children and adolescents, and provide a practical, scientific and efficient method for solving the problem of foot and spine deformities in children and adolescents.

[0126] In one of the embodiments, based on the exercise intervention plan, use a multi-objective optimization algorithm to evaluate the correction effect of the foot orthosis to obtain the orthosis design parameters, which may include the following steps:

[0127] Step S601, obtain the initial design parameters of the foot orthosis.

[0128] Step S602, construct a multi-objective optimization model according to the initial design parameters combined with the key indicators in the exercise intervention plan.

[0129] Step S603, construct the objective according to the expected correction effect of the foot orthosis using the multi-objective optimization model to obtain the optimization objective function; the objectives include arch support force and gait balance.

[0130] Step S604, iteratively optimizing the design parameters of the plantar orthosis using a genetic algorithm based on the optimization objective function to generate multiple groups of candidate design solutions.

[0131] Step S605, evaluating the correction effect of each group of candidate design solutions according to the three-dimensional foot data, and selecting the optimal design solution.

[0132] Step S606, verifying the matching degree of the correction effect of the optimized orthosis by using the test data according to the optimal design solution, and obtaining the design parameters of the orthosis.

[0133] First, the initial design parameters of the plantar orthosis are obtained. Then, the initial design parameters are combined with the key indicators in the exercise intervention plan. Through scientific construction, a multi-objective optimization model is generated, which can comprehensively consider the influence of multiple factors on the design of the orthosis. Subsequently, according to the expected correction effect of the plantar orthosis, the multi-objective optimization model is used to construct the target, and the optimization objective function including important objectives such as arch support and gait balance is obtained to clarify the direction of design optimization. Based on the optimization objective function, the design parameters of the plantar orthosis are optimized by multiple rounds of iterative optimization using a genetic algorithm, and multiple groups of candidate design schemes are generated by continuous search in the parameter space. With the help of three-dimensional foot data, the correction effect of each group of candidate design schemes is comprehensively evaluated, and the optimal design scheme is determined through rigorous screening. Finally, according to the optimal design scheme, the test data is used to verify the matching degree between the correction effect of the optimized orthosis and the expected one. After this verification link, accurate and reliable orthosis design parameters are finally obtained.

[0134] Second, as Figure 5 As shown, the present application also provides an online exercise intervention system for children and adolescents with foot and spine deformity based on multi-dimensional assessment, the system comprising:

[0135] The data acquisition module 701 is used to acquire the three-dimensional data of the feet of children and adolescents; the three-dimensional data of the feet include three-dimensional information of the foot shape contour, joint range of motion and plantar pressure distribution.

[0136] The foot spine assessment module 702 is used to extract and fuse the features of the three-dimensional foot data to obtain the foot fusion feature value; it is also used to perform brute force matching based on the foot fusion feature value to obtain the corresponding feature parameter combination; it is also used to input the feature parameter combination into the trained foot assessment model to assess the foot spine condition of children and adolescents and obtain the foot spine condition assessment result; the foot spine condition assessment result is used to determine whether there is a foot spine deformity and the type and degree of the deformity.

[0137] The orthosis design module 703 is used to construct a time series prediction model based on the foot ridge condition assessment result to predict the disease change trend and generate an exercise intervention plan; the exercise intervention plan is used to supervise and guide the user to complete the online training content in real time through a video; it is also used to evaluate the correction effect of the foot orthosis by using a multi-objective optimization algorithm based on the exercise intervention plan to obtain the orthosis design parameters.

[0138] The intervention generation module 704 is used to optimize and adjust the exercise intervention plan according to the updated three-dimensional foot data and the orthosis design parameters to obtain an updated exercise intervention plan.

[0139] For the above-mentioned online exercise intervention system for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation, the data acquisition module acquires the three-dimensional foot data of children and adolescents, which covers the three-dimensional information of the foot external contour, joint range of motion, and plantar pressure distribution. The foot and spinal evaluation module receives this data, obtains the foot fusion feature value through feature extraction and fusion operations, and obtains the corresponding feature parameter combination through brute-force matching. Subsequently, this combination is input into the trained foot evaluation model to evaluate the foot and spinal condition of children and adolescents, and output the foot and spinal condition evaluation result to determine whether there is a foot and spinal deformity and the type and degree of the deformity. The orthosis design module constructs a time series prediction model based on the foot and spinal condition evaluation result to predict the disease change trend, and then generates an exercise intervention plan. The exercise intervention plan is used to supervise and guide the user to complete the online training content in real time through a video. At the same time, based on this plan, a multi-objective optimization algorithm is used to evaluate the correction effect of the foot orthosis to determine the orthosis design parameters. Finally, the intervention generation module optimizes and adjusts the exercise intervention plan according to the updated three-dimensional foot data and the determined orthosis design parameters to obtain an updated exercise intervention plan to achieve effective intervention for children and adolescents with foot and spinal deformities. It greatly improves the scientificity, accuracy, and effectiveness of the evaluation and intervention of children and adolescents' foot and spinal deformities, and helps to improve their foot and spinal health status.

[0140] It should be understood that although the steps in the flowcharts involved in the above-mentioned embodiments are shown in sequence according to the arrows, these steps do not necessarily need to be executed in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments may include multiple steps or multiple stages. These steps or stages do not necessarily need to be executed at the same time, but can be executed at different times. The execution order of these steps or stages does not necessarily need to be sequential, but can be executed alternately or alternately with at least a part of other steps or steps in other steps.

[0141] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the online exercise intervention method for children and adolescents with foot and spinal deformities based on multi-dimensional evaluation as described above are implemented.

[0142] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0143] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial descriptions of the method embodiments. The device embodiments described above are only illustrative. The components described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the present disclosure solution. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0144] The above embodiments only represent several implementation manners of the embodiments of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the embodiments of the application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the embodiments of the present application, several deformations and improvements can still be made, and these all belong to the protection scope of the embodiments of the present application.

Claims

1. An online exercise intervention method for foot ridge deformity in children and adolescents based on multi-dimensional assessment, characterized by: The method comprises: Acquire three-dimensional foot data of children and adolescents; the three-dimensional foot data includes three-dimensional information of foot shape contour, joint range of motion and plantar pressure distribution; Extracting and fusing the three-dimensional foot data to obtain foot fusion feature values; performing brute force matching based on the foot fusion feature values ​​to obtain corresponding feature parameter combinations; The characteristic parameter combination is input into a trained foot assessment model to assess the foot spine condition of children and adolescents, and obtain a foot spine condition assessment result; the foot spine condition assessment result is used to determine whether there is a foot spine deformity and the type and degree of the deformity; A time series prediction model is constructed based on the foot spine condition assessment results to predict the disease change trend and generate an exercise intervention plan; the exercise intervention plan is used to supervise and guide users to complete online training content in real time through video; Based on the exercise intervention plan, a multi-objective optimization algorithm is used to evaluate the correction effect of the plantar orthosis to obtain the design parameters of the orthosis; The exercise intervention plan is optimized and adjusted according to the updated three-dimensional foot data and the orthosis design parameters to obtain an updated exercise intervention plan.

2. An online exercise intervention method for foot and spine deformity in children and adolescents based on multi-dimensional assessment, characterized by: The extracting and fusing the foot three-dimensional data to obtain the foot fusion feature value includes: Preprocessing the three-dimensional foot data using a sliding average filtering algorithm to obtain processed foot morphology data; Performing enhancement processing on the processed foot morphology data based on a Plaus operator algorithm to obtain enhanced foot morphology data; The foot feature model is constructed using a convolutional neural network to locate the enhanced foot morphology data and obtain the target data area; Extracting and marking the target data area based on a velocity-invariant feature transformation algorithm to obtain foot feature points; the foot feature points include at least one of the morphology, movement and pressure features of the arch, heel and metatarsophalangeal joint; The feature fusion formula is used to perform feature fusion on the foot feature points to obtain foot fusion feature values.

3. The method according to claim 2, characterized in that The step of using a feature fusion formula to fuse the foot feature points to obtain a foot fusion feature value includes: The foot fusion feature value is calculated using the following feature fusion formula: Where H represents the foot fusion characteristic value, ω i represents the weight coefficient of the i-th foot feature, S i represents the value of the i-th foot feature, x represents the spatial position of the foot feature point, μ i represents the mean of the i-th foot feature, σ i represents the standard deviation of the i-th foot feature, and n represents the number of foot features.

4. The method according to claim 1, characterized in that: The brute force matching based on the foot fusion feature value to obtain a corresponding feature parameter combination includes: Acquire reference three-dimensional foot data of children and adolescents under normal conditions, and preprocess the reference three-dimensional foot data; Extracting feature points from the foot fusion feature values ​​of the foot three-dimensional data and the reference foot three-dimensional data using an LDB descriptor to obtain a fusion descriptor corresponding to the foot three-dimensional data and a reference descriptor corresponding to the reference foot three-dimensional data; Performing brute force matching on the fusion descriptor and the reference descriptor to obtain a candidate matching pair; Using the improved LMedS algorithm to eliminate the wrong matches in each of the candidate matching pairs, to obtain the best matching pair; The best matching pairs with large feature differences are eliminated twice using the difference-in-number ratio to obtain a feature parameter combination; the feature parameter combination includes a foot shape contour deviation parameter, a joint range of motion abnormality parameter, and a plantar pressure distribution difference parameter.

5. The method according to claim 1, characterized in that The characteristic parameter combination is input into the trained foot assessment model to assess the foot spine condition of children and adolescents, and the foot spine condition assessment result is obtained, including: De-noising and outlier processing are performed on the characteristic parameter combination data to obtain matching training data; Based on the matching training data, a machine learning classifier is used to perform model training to obtain a foot assessment model to be trained; the foot assessment model is at least one of a support vector machine, a decision tree, a neural network and a random forest; Calculating the AUC value of the foot evaluation model to be trained using a formula based on historical training data to obtain a trained foot evaluation model; the foot evaluation model is a foot evaluation model whose performance meets preset conditions; Inputting the matching training data into the trained foot assessment model to assess the foot spine condition of children and adolescents to obtain a foot spine condition assessment value; The foot spine condition assessment value is data labeled to obtain a foot spine condition assessment result for determining the foot spine condition.

6. The method according to claim 1, characterized in that The method of constructing a time series prediction model based on the foot spine condition assessment result to predict the disease change trend and generate an exercise intervention plan includes: Inputting characteristic parameters related to the disease condition in the foot spine condition assessment result into a time series prediction model and calculating using a formula to obtain a predicted disease deterioration probability value; The predicted probability value of the disease worsening is calculated using the following formula: Where P(t) represents the probability of disease progression at time t, n represents the number of characteristic parameters, represents the weight coefficient of the jth feature parameter, x j (t) represents the value of the jth characteristic parameter at time t, θ j represents the reference value of the jth characteristic parameter, β j represents the attenuation coefficient, α represents the normalization coefficient; Generate a preliminary exercise intervention plan based on the predicted probability value of the disease worsening and the key indicators of the foot spine condition; Using a machine learning algorithm to classify the preliminary exercise intervention plan, and obtaining exercise intervention levels corresponding to different deformity types; If the deformity type is flat feet, extract the arch support related indexes to optimize the arch training content in the preliminary exercise intervention program; If the deformity type is scoliosis, extracting spinal correction related indicators and optimizing the spinal training content in the preliminary exercise intervention program; The matching degree between the preliminary exercise intervention plan and the probability value of disease progression is verified based on the exercise intervention level to obtain a final exercise intervention plan.

7. The method according to claim 1, characterized in that The method of evaluating the correction effect of the plantar orthosis by using a multi-objective optimization algorithm based on the exercise intervention plan to obtain the orthosis design parameters includes: Obtain initial design parameters of the plantar orthosis; A multi-objective optimization model is constructed based on the initial design parameters combined with the key indicators in the exercise intervention program; According to the expected correction effect of the plantar orthosis, a multi-objective optimization model is used to construct an objective to obtain an optimization objective function; the objectives include arch support and gait balance; Iteratively optimizing the design parameters of the plantar orthosis using a genetic algorithm based on the optimization objective function to generate multiple sets of candidate design solutions; Evaluate the correction effect of each group of candidate design solutions according to the three-dimensional foot data, and select the optimal design solution; According to the optimal design scheme, the matching degree of the correction effect of the optimized orthosis is verified by using the test data to obtain the orthosis design parameters.

8. An online exercise intervention system for children and adolescents with foot ridge deformity based on multi-dimensional assessment, characterized by: The system comprises: A data acquisition module, used to acquire three-dimensional foot data of children and adolescents; the three-dimensional foot data includes three-dimensional information of foot shape contour, joint range of motion and plantar pressure distribution; A foot spine assessment module is used to extract and fuse the foot three-dimensional data to obtain foot fusion feature values; it is also used to perform brute force matching based on the foot fusion feature values ​​to obtain corresponding feature parameter combinations; it is also used to input the feature parameter combinations into a trained foot assessment model to assess the foot spine conditions of children and adolescents to obtain foot spine condition assessment results; the foot spine condition assessment results are used to determine whether there is foot spine deformity and the type and degree of deformity; An orthosis design module is used to construct a time series prediction model based on the foot spine condition assessment results to predict the disease change trend and generate an exercise intervention plan; the exercise intervention plan is used to supervise and guide users to complete online training content in real time through video; and is also used to evaluate the correction effect of the plantar orthosis using a multi-objective optimization algorithm based on the exercise intervention plan to obtain orthosis design parameters; The intervention generation module is used to optimize and adjust the movement intervention plan according to the updated three-dimensional foot data and the orthosis design parameters to obtain an updated movement intervention plan.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images