Spine posture abnormity three-level screening and correcting system and method

Abstract

The invention relates to the technical field of spine health, solves the problems that in the prior art, a spine posture abnormity screening mode is single, a screening process lacks grading progressive performance and a correction strategy lacks individuation, and provides a spine posture abnormity three-level screening and correction system and method. The system comprises a data acquisition module used for acquiring first original data, second original data and third original data; the primary screening module is used for obtaining a primary screening result; the secondary screening module is used for obtaining a secondary screening result according to the second original data; the third-stage screening module is used for obtaining a third-stage screening result according to the third original data; the comprehensive evaluation module is used for obtaining a comprehensive evaluation result; the correction management module is used for generating a personalized correction strategy; and the control module is used for managing the working process of each module. The invention provides a spine posture abnormity screening and correction system capable of generating a personalized spine correction strategy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of spine health, and in particular to a spine posture abnormality three-stage screening and correction system and method. BACKGROUND

[0002] Spine posture abnormality is a common health problem, mainly manifested as scoliosis, humpback, pelvic tilt, head and neck forward tilt and other bad postures. Long-term spine posture abnormality not only causes abnormal appearance, but also may cause chronic pain, motor dysfunction and cardiopulmonary limitation and other health hazards, especially in adolescents and long-term sitting population. Therefore, early screening and scientific correction of spine posture abnormality are necessary for preventing disease development, improving quality of life and promoting physical health. In the prior art, screening and correction of spine posture abnormality mainly rely on single-level detection and intervention means. For example, some studies use static posture photos for two-dimensional angle measurement to identify abnormal postures; some technologies use motion capture or inertial sensors to evaluate dynamic motion trajectories; and some medical scenarios rely on X-ray, MRI and other medical imaging means for structural lesion detection. Common correction methods include physical therapy, wearing braces or single exercise prescription guidance. These technical solutions realize monitoring and intervention of spine posture to some extent, but still have limitations. The main problems existing in the prior art are: first, single data source screening method cannot comprehensively reflect the spine posture state of the user, and it is difficult to simultaneously consider comprehensive analysis of static posture, dynamic function and medical image; second, the existing screening process is mostly single evaluation, lacking a hierarchical progressive process from primary screening to professional diagnosis, which is easy to cause missed judgment or misjudgment; third, the existing correction method is relatively general, lacking individualized correction strategy based on individual characteristics and comprehensive risk results, resulting in large differences in correction effect.

[0003] Therefore, how to construct a spine posture abnormality screening and correction system capable of realizing multi-level progressive screening and multi-modal data fusion analysis and generating individualized correction strategies has become a technical problem to be solved. SUMMARY

[0004] Therefore, the present application provides a spine posture abnormality three-stage screening and correction system and method to solve the problems of single screening method, lack of hierarchical progression in screening process and lack of individualization in correction strategy in the prior art.

[0005] The technical scheme adopted by the present application is: In a first aspect, the present application provides a spine posture abnormality three-stage screening and correction system, which comprises: The control module and the data collection module, the first screening module, the second screening module, the third screening module, the comprehensive evaluation module and the correction management module connected therewith; The data collection module is configured to collect first raw data for first screening, second raw data for second screening and third raw data for third screening related to the posture of the user's spine; The first screening module is configured to perform preliminary analysis on the first raw data to identify the abnormal risk of the user's spine posture and obtain a first screening result; The second screening module is configured to perform further analysis and quantitative evaluation on the medium and high risk users in the first screening result according to the second raw data to obtain a second screening result; The third screening module is configured to perform professional evaluation on the medium and high risk users in the second screening result according to the third raw data to obtain a third screening result; The comprehensive evaluation module is configured to perform comprehensive analysis on the first screening result, the second screening result and the third screening result to obtain a comprehensive evaluation result; The correction management module is configured to generate a personalized correction strategy corresponding to the user according to the comprehensive evaluation result and the first raw data. The control module is configured to manage the workflow of each module, coordinate the data collection, screening, evaluation and correction management process, monitor the system state and optimize the screening and correction strategy.

[0006] Preferably, the data collection module is further configured to: obtain preset first screening accuracy requirements, second screening accuracy requirements and third screening accuracy requirements; collect the basic information, static posture images and dynamic posture images of the user according to the first screening accuracy requirements to obtain the first raw data; collect three-dimensional images of the user according to the second screening accuracy requirements, and determine the second raw data according to the three-dimensional images and the dynamic posture images; collect medical image data and historical disease information of the user according to the third screening accuracy requirements to obtain the third raw data.

[0007] Preferably, the first screening module is further configured to: statistically analyze and classify the height, weight, age and gender of the user according to the basic information to obtain user body type feature data; extract image features from the front view static image, the rear view static image and the side view static image of the user according to the static posture images to obtain static posture analysis results; According to the dynamic posture image, action trajectory analysis and joint angle calculation are performed to obtain a dynamic posture analysis result; According to the static posture analysis result, the dynamic posture analysis result, and user body shape feature data, a user spine posture abnormality risk is identified and preliminarily scored to obtain the first-level screening result.

[0008] Preferably, the first-level screening module extracts image features from a front-view static image, a back-view static image, and a side-view static image of the user according to the static posture image, and obtains a static posture analysis result according to the extracted image feature information, and the specific process includes: Human key point recognition is performed on the front-view static image to extract shoulder, neck, and pelvic position features to obtain front-view image feature information; Spine midline extraction and left-right symmetry analysis are performed on the back-view static image to extract shoulder blade, spine midline, and waist position features to obtain back-view image feature information; Spine curvature fitting and head and neck position recognition are performed on the side-view static image to extract thoracic vertebra kyphosis angle, lumbar vertebra lordosis angle, and head forward inclination degree to obtain side-view image feature information; According to the front-view image feature information, the back-view image feature information, and the side-view image feature information, posture parameters in each view are normalized and feature fused to obtain multi-view comprehensive feature information; According to the multi-view comprehensive feature information, a static posture of the user is comprehensively evaluated to identify shoulder height difference, pelvic tilt angle, and spine curvature abnormality to obtain the static posture analysis result; The specific process in which the first-level screening module obtains a dynamic posture analysis result according to the dynamic posture image includes: According to the collected dynamic posture image, time-series arrangement is performed on continuous dynamic posture image frames to obtain a dynamic image sequence of user actions; According to the dynamic image sequence, human key point recognition is performed to extract head, shoulder joint, hip joint, knee joint, and ankle joint key point coordinates to obtain key point time series data; According to the key point time series data, trajectory fitting and path smoothing processing are performed on actions performed by the user to obtain action trajectory curves of joints and a torso; According to the key point time series data, angle changes of shoulder joints, hip joints, knee joints, and the like during actions are calculated to obtain joint angle curves; According to the action trajectory curve and the joint angle curve, stability and action coordination of a user posture are analyzed to obtain dynamic posture evaluation parameters; According to the dynamic posture evaluation parameter, the posture state of the spine of the user in the action process is comprehensively judged, and a dynamic posture analysis result is obtained.

[0009] Preferably, the secondary screening module is further configured to: According to the primary screening result, a user with a medium-high risk in the primary screening result is screened out as a secondary screening object; According to the second original data, posture reconstruction and surface fitting are performed on a three-dimensional image of the secondary screening object to obtain a three-dimensional body posture model of the spine and related bones; According to the three-dimensional body posture model, the curvature of the spine, the inclination angle of the pelvis, and the asymmetry of the shoulder and waist are calculated to obtain three-dimensional structural feature data; According to the dynamic posture image and the three-dimensional structural feature data, the stability of the spine, the coordination of the joints, and the balance of the muscles of the secondary screening object in the action process are jointly analyzed to obtain dynamic functional feature data; According to the three-dimensional structural feature data and the dynamic functional feature data, a preset evaluation algorithm is used to quantitatively calculate the severity of the posture abnormality of the spine of the secondary screening object to obtain a quantitative score result; According to the quantitative score result, the posture abnormality of the spine of the secondary screening object is classified to form a secondary screening result.

[0010] Preferably, the secondary screening module according to the three-dimensional structural feature data and the dynamic functional feature data, adopts a preset evaluation algorithm to quantitatively calculate the severity of the posture abnormality of the spine of the secondary screening object to obtain a quantitative score result, and the specific process includes: The three-dimensional structural feature data and the dynamic functional feature data are normalized and noise-filtered to obtain a standardized feature data set meeting the calculation requirements; From the standardized feature data set, posture abnormality representation indexes including the curvature of the spine, the rotation angle of the vertebrae, the range of joint motion, and the action coordination are extracted; The posture abnormality representation indexes are assigned corresponding weight coefficients to obtain an index set with weight coefficients; Based on the index set with weight coefficients, an intermediate score of the severity of the posture abnormality of the spine of the secondary screening object is calculated; The intermediate score result is interval-mapped and normalized to obtain a standardized quantitative score result.

[0011] Preferably, the tertiary screening module is further configured to: Based on the aforementioned third raw data, the collected medical image data undergoes format conversion, noise reduction, and enhancement processing, and the user's historical medical information is structurally organized to obtain a standardized medical analysis dataset. Multi-dimensional feature extraction is performed on the standardized medical image data to obtain image feature parameters, which include spinal alignment features, vertebral body morphology features, intervertebral space features, and soft tissue compression features. The historical disease information is correlated with the image feature parameters to extract the medical history feature indicators related to spinal posture abnormalities, and a comprehensive feature indicator set is obtained. Based on the comprehensive feature index set, pathological risk calculation and abnormal pattern identification are performed on medium- and high-risk users in the secondary screening results to obtain intermediate evaluation results; Based on the intermediate assessment results, the severity of spinal posture abnormalities is graded and determined, and the three-level screening results are output in conjunction with preset medical clinical reference standards.

[0012] Preferably, the comprehensive evaluation module is further used for: According to the instructions of the control module, the first-level screening results, second-level screening results and third-level screening results are received, and the data of different screening levels are classified, stored and structured to obtain the dataset to be comprehensively analyzed. The screening indicators at different levels in the dataset to be comprehensively analyzed are uniformly converted and standardized to ensure that the results at different levels are comparable, thus obtaining a standardized screening result set. The body shape feature data, static posture analysis results, dynamic posture analysis results, three-dimensional structural feature data, and medical imaging feature data in the standardized screening result set are fused to obtain a multi-dimensional comprehensive feature vector. The multidimensional comprehensive feature vector is weighted and weighted to obtain a comprehensive risk score for the user's spinal posture abnormality. Based on the comprehensive risk score and the preset risk grading standards, the degree of spinal posture abnormality of the user is classified and judged to obtain the comprehensive assessment result.

[0013] Preferably, the correction management module is further configured to: Based on the comprehensive assessment results, the user's spinal posture abnormality level, abnormal location, and risk factors are analyzed to obtain personalized correction needs information; Based on the user's basic information in the first raw data, the user's basic physical conditions are matched to obtain individual adaptability parameters; Based on the personalized correction needs information and individual adaptability parameters, the user's correction goals are determined, including the correction area, correction range, and correction period; According to the correction target, a correction method matched with the target is called from a preset correction strategy library, including a posture training scheme, a motion intervention scheme and an auxiliary appliance use scheme, and combination optimization is performed to obtain the personalized correction strategy.

[0014] In a second aspect, the application provides a spine posture abnormality three-stage screening and correction method, which is realized based on the spine degenerative disease grading early warning and rehabilitation management system as described in the first aspect.

[0015] In summary, the beneficial effects of the application are as follows: The application provides a spine posture abnormality three-stage screening and correction system and method, which comprises a control module and a data acquisition module, a first-stage screening module, a second-stage screening module, a third-stage screening module, a comprehensive evaluation module and a correction management module connected with the control module; the data acquisition module is used for collecting first original data for first-stage screening, second original data for second-stage screening and third original data for third-stage screening related to the spine posture of a user; the first-stage screening module is used for performing preliminary analysis on the first original data, identifying the spine posture abnormality risk of the user and obtaining a first-stage screening result; the second-stage screening module is used for performing further analysis and quantitative evaluation on the medium and high risk users in the first-stage screening result according to the second original data and obtaining a second-stage screening result; the third-stage screening module is used for performing professional evaluation on the medium and high risk users in the second-stage screening result according to the third original data and obtaining a third-stage screening result; the comprehensive evaluation module is used for performing comprehensive analysis on the first-stage screening result, the second-stage screening result and the third-stage screening result and obtaining a comprehensive evaluation result; the correction management module is used for generating a personalized correction strategy corresponding to the user according to the comprehensive evaluation result and the first original data; and the control module is used for managing the work flow of each module, coordinating the data acquisition, screening, evaluation and correction management process, monitoring the system state and optimizing the screening and correction strategy. The system effectively solves the technical problems of single screening mode, lack of hierarchical progression and non-personalized correction strategy in the prior art by constructing a three-stage screening and correction system including data acquisition, hierarchical screening, comprehensive evaluation and correction management. Specifically, the data acquisition module can obtain the basic information, posture image, three-dimensional image and medical image of the user according to different precision requirements, thereby forming multi-level original data support; by setting the first-stage, second-stage and third-stage screening modules, the risk identification and quantitative evaluation are realized from shallow to deep and gradually progressive, and the problems of insufficient screening dimension and difficulty in accurately judging the severity of spine posture abnormality in the traditional method are overcome; the comprehensive evaluation module can fuse the screening results at all stages to form a comprehensive evaluation conclusion, improving the scientificity and reliability of diagnosis; and the correction management module automatically generates a personalized correction strategy for different users based on the comprehensive evaluation result and individual characteristic data, thereby breaking through the limitations of the existing correction method and realizing personalized intervention and precise management. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed to be used in the embodiments of the application will be briefly introduced as follows, and other drawings can also be obtained by those of ordinary skill in the art without any creative effort on the premise of not paying any creative effort, and these are within the protection scope of the application.

[0017] Figure 1A structural schematic diagram of a spine posture abnormality three-stage screening and correction system in embodiment 1 of the present application; Figure 2 A flowchart of a process in which the comprehensive evaluation module in embodiment 1 of the present application comprehensively analyzes the first-stage screening result, the second-stage screening result and the three-stage screening result to obtain a comprehensive evaluation result. DETAILED DESCRIPTION

[0018] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. It should be noted that, in this document, relationship terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. In the description of the present application, it should be understood that the orientations or positional relationships indicated by terms such as center, upper, lower, front, rear, left, right, vertical, horizontal, top, bottom, inner and outer are based on the orientations or positional relationships shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. Moreover, the terms “include”, “contain” or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such a process, method, article or device. Without more limitations, the elements defined by the phrase “include” do not exclude the presence of additional identical elements in the process, method, article or device that includes the elements. If there is no conflict, the embodiments of the present application and the various features in the embodiments can be combined with each other, and are all within the protection scope of the present application.

[0019] Embodiment 1 Please refer to Figure 1 Embodiment 1 of the present application discloses a spine posture abnormality three-stage screening and correction system, which comprises: a control module and data acquisition module, first-stage screening module, second-stage screening module, three-stage screening module, comprehensive evaluation module and correction management module connected thereto; Specifically, the control module is the coordination center of the system (e.g., workflow engine, interface manager, scheduler) for managing and connecting various functional modules and providing a unified communication and security interface; the data acquisition, multi-level screening, comprehensive evaluation and correction management modules are functional units connected to it. The purpose is to realize the orderly development of data and control flow between modules, ensure that the screening process automatically advances according to the strategy and realizes closed-loop feedback, so that the system can not only perform batch and low-cost primary screening, but also trigger high-precision follow-up evaluation and medical-level intervention. To achieve this, the control module can use a layered software architecture (including API gateway, message queue or event bus, task scheduler and policy rule engine) to trigger data acquisition, start the corresponding screening algorithm, record logs and perform alarm and rollback when an exception or failure is detected according to the preset strategy (e.g., priority, accuracy requirement, privacy policy); at the same time, it provides permission control, audit logs, remote updates and interfaces for connecting with hospital information systems (HIS / EMR). The benefits of this design include: ensuring traceability of the process, improving system availability and scalability, facilitating on-demand upgrade of screening strategies and meeting the compliance and privacy requirements of different scenarios (school physical examination, family self-examination, outpatient review), thereby improving overall screening efficiency and reliability.

[0020] The data acquisition module is configured to acquire first raw data for primary screening, second raw data for secondary screening, and third raw data for tertiary screening related to the user's spine posture; Specifically, the purpose of the data acquisition module is to selectively obtain data of different granularities and modalities according to the accuracy requirements of the screening levels, providing a reliable data foundation for subsequent coarse-to-fine analysis and decision-making. The implementation process includes: standardized acquisition (including camera calibration, posture standardization guidance, sampling frame rate / resolution setting) at the terminal (mobile App, tablet, depth camera or physical examination platform), timestamping and metadata labeling of multi-modal data, and implementing real-time quality control (such as automatic detection of lighting, occlusion, human key point confidence, signal loss) and encrypted transmission; for two / three-level data, it also includes DICOM interface with image equipment, wearable IMU synchronous sampling, and structured entry of medical history data. By ensuring that different screening levels obtain data matching their accuracy, the error rate and false positive / negative probability of downstream analysis are reduced, and the data reproducibility and clinical usability are improved through quality control and metadata management.

[0021] The primary screening module is configured to perform preliminary analysis on the first raw data, identify the risk of spine posture abnormalities of the user, and obtain a primary screening result. Specifically, the purpose of the first screening module is to preliminarily screen a large population in a low-cost and rapid manner, to screen out most normal or low-risk individuals, and to only send medium and high-risk individuals to the second screening with higher accuracy, so as to save medical resources and realize the early intervention path. The implementation process can include: first, pre-processing the first original data (image denoising, scale normalization), then extracting static features (shoulder height difference, pelvic line, spine projection contour, etc.) and dynamic features (motion trajectory, joint angle fluctuation), and inputting these features into a rule-based or machine learning-based classification / score module to generate confidence and risk level; at the same time, output a brief report and provide automated suggestions (such as retesting, second screening appointment or lifestyle tips). By quickly identifying potential problems, improving screening coverage, reducing the burden on professional places, and providing initial clues and data context for subsequent hierarchical screening, the rate of missed diagnosis is reduced and the efficiency of screening is improved.

[0022] The second screening module is configured to further analyze and quantitatively evaluate the medium and high-risk users in the first screening result according to the second original data, to obtain a second screening result. Specifically, the second screening module aims to more accurately evaluate the structure and function of individuals marked as medium and high risk, in order to distinguish between patients who need clinical intervention and those who only need rehabilitation training, and to provide quantitative basis for clinical decision-making. The second screening module is configured to call the second original data to reconstruct a three-dimensional posture (such as multi-view photogrammetry, structured light / depth sensor or body surface point cloud reconstruction), register and surface fit the three-dimensional model to obtain the spine curvature contour, vertebral inclination and pelvic parameters; at the same time, fuse dynamic data (video or IMU) to calculate the joint angle change curve, posture stability index and muscle group coordination index; normalize and weight the extracted structure and function features according to the pre-set weight rule or the trained evaluation model, to obtain a quantitative score and output the second screening result in the form of grade for clinical or third screening reference, thereby significantly improving the specificity and accuracy of screening, providing quantifiable progress and efficacy evaluation indicators, helping clinicians make more evidence-based referral or intervention decisions, and reducing unnecessary medical imaging examinations.

[0023] The third screening module is configured to perform professional evaluation on the medium and high-risk users in the second screening result according to the third original data, to obtain a third screening result. Specifically, the tertiary screening module aims to clinically confirm individuals determined as high risk or having structural abnormal signs in the secondary screening, identify pathological changes requiring surgery or medical intervention, and form a diagnosis basis with clinical evidence. The tertiary screening module is used for denoising, registration and segmentation of medical images such as DICOM (X-ray plain film for Cobb angle measurement, MRI for intervertebral disc degeneration and soft tissue evaluation), automatically extracting radiological indicators (vertebral rotation, intervertebral space height, bone changes, etc.) and making correlation analysis with clinical history, using clinically verified classification standards or knowledge-driven / machine learning models for lesion pattern recognition and risk estimation, and finally outputting tertiary screening results (including image annotation, classification conclusion and recommended follow-up examination / treatment plan). By providing high-confidence diagnostic evidence and clinically classified judgments, it supports treatment decisions (e.g., conservative rehabilitation vs. surgical intervention), and provides standardized baseline data for disease tracking, reducing misdiagnosis and improving treatment safety and efficacy.

[0024] The comprehensive evaluation module is used for comprehensive analysis of the primary screening result, the secondary screening result and the tertiary screening result to obtain a comprehensive evaluation result. Specifically, the comprehensive evaluation module is used for cross-validation and result fusion of screening information at different levels, such as the preliminary risk prompt provided by the primary screening, the quantitative score obtained by the secondary screening, and the medical imaging conclusion provided by the tertiary screening. The purpose is to avoid errors caused by a single data source and ensure that the final result is both universal and medically reliable. In the implementation process, the comprehensive evaluation module can use multi-source data fusion algorithms such as weighted decision models, Bayesian inference or rule-based analytic hierarchy process to combine low-level broad coverage results with high-level high-precision results to form the final risk level and diagnostic opinion; At the same time, it can generate an auxiliary report to prompt the type and severity of the abnormality, thereby significantly improving the accuracy and robustness of the overall screening conclusion, reducing missed diagnosis and misdiagnosis, and providing scientific and traceable basis for individualized correction and clinical intervention.

[0025] The correction management module is used for generating an individualized correction strategy corresponding to the user according to the comprehensive evaluation result and the first original data. Specifically, the personalized correction strategy generated by the correction management module refers to an intervention plan customized for the posture abnormality type, severity, and body type characteristics of each user, such as recommending specific rehabilitation training actions, scheduling the wearing of orthotics, or giving posture correction suggestions in combination with lifestyle habits. The purpose of the correction management module is to convert the screening and evaluation results into executable intervention measures, thereby realizing a closed loop of prevention and treatment. In the implementation process, the correction management module determines the abnormal position and severity based on the comprehensive evaluation results, and combines the body type characteristics (such as height, weight, age, and gender) in the first raw data to perform personalized matching of intervention intensity and content, and finally generates a plan and can be pushed and tracked in real time through a mobile application or wearable device to improve the pertinence and compliance of intervention measures, while supporting dynamic adjustment to improve the correction effect and long-term health benefits of users.

[0026] The control module is configured to manage the workflow of each module, coordinate the data collection, screening, evaluation, and correction management processes, monitor the system state, and optimize the screening and correction strategies.

[0027] Specifically, the workflow management in the control module refers to unified scheduling and state monitoring of the operation of each module in the entire system, such as determining the data collection sequence, triggering the corresponding level of screening module, collecting the evaluation results and passing them to the correction management module. The purpose of the control module is to ensure the integrity and efficiency of system operation, so that the data flow and logical flow remain orderly connected, while achieving dynamic process adjustment under different individuals and different risk levels. The implementation process can be completed through a distributed control architecture or a centralized scheduling platform, including a workflow engine, a task queue, a state monitoring and exception handling mechanism; in the monitoring process, the control module can also optimize the strategy through performance indicators and error logs, such as shortening the screening delay or dynamically adjusting the screening algorithm parameters. By improving the overall automation level and stability of the system, reducing the cost of manual intervention, and ensuring that the system remains efficient, reliable, and scalable in large-scale use or long-term operation.

[0028] Preferably, the data collection module is further configured to: obtain a preset first-level screening accuracy requirement, a second-level screening accuracy requirement, and a third-level screening accuracy requirement; Specifically, the screening accuracy requirements refer to the required data fineness and detection accuracy for different levels of screening. For example, the first-level screening focuses more on rapid coverage and has relatively low accuracy requirements, while the second-level and third-level screening require higher resolution and medical reliability. The purpose of this step is to provide clear standards for subsequent data collection, so that different levels of screening can achieve a reasonable division between efficiency and accuracy. In the implementation process, multiple sets of accuracy requirement configurations can be pre-stored, such as camera resolution, sampling frequency, and medical image clarity level, and corresponding parameters can be called according to different screening scenarios to ensure that data collection meets the expected standards and avoids resource waste, making the entire screening process hierarchical and efficient, quickly identifying potential risks in a large population, and ensuring more accurate evaluation of high-risk groups in subsequent stages.

[0029] According to the first-level screening accuracy requirement, the basic information, static posture image, and dynamic posture image of the user are collected to obtain the first original data; Specifically, the basic information includes the user's height, weight, age, gender, and other statistical data, the static posture image is the front, rear, and side view photos of the user in standing or sitting posture, and the dynamic posture image is the continuous image sequence of the user in walking, bending, or stretching actions. By quickly obtaining initial data that can reflect the user's posture and action characteristics, preliminary screening can be performed on a large scale at low cost. In the implementation process, the user's multi-angle photos and short videos are taken by a camera device or a mobile terminal, and the basic information form is filled out at the same time, realizing the unified collection of structured data and unstructured image data, which can complete the data preparation work for preliminary risk identification in a short time, provide sufficient input conditions for the first-level screening, reduce the user's cooperation difficulty, and improve the screening popularity.

[0030] According to the second-level screening accuracy requirement, the three-dimensional image of the user is collected, and the second original data is determined according to the three-dimensional image and the dynamic posture image; Specifically, the three-dimensional image is the three-dimensional surface data of the user obtained by a depth camera, structured light scanning, or laser scanning, which can reconstruct the spatial form of the spine and torso. Based on the first-level screening, more detailed three-dimensional analysis is performed on medium and high-risk groups to obtain joint features of posture structure and motion process. In the implementation process, the skeleton shape and surface curvature information of the user are collected by a three-dimensional imaging device, and are registered with the joint motion trajectory in the dynamic posture image, thereby generating the second original data containing the spine curvature, pelvic tilt, and motion coordination, which can reflect both static structure and dynamic function, providing more accurate data support for quantitative evaluation of abnormal risks and improving the scientificity and reliability of medium and high-risk group screening.

[0031] According to the accuracy requirement of the three-level screening, medical image data and historical disease information of the user are collected to obtain the third original data.

[0032] Specifically, the medical image data generally refers to clinical imaging results such as X-ray, CT or MRI, and the historical disease information includes the user's previous spine disease, orthopedic history and related treatment records. By confirming the final medical level of the high-risk population, it is ensured that the screening conclusion has clinical reference value. In the implementation process, by connecting with the hospital database or user uploading mode, medical images meeting the resolution and format requirements are obtained, and the user's medical history data is structured and input to form a medical data set corresponding to the image data, which provides professional medical basis for the three-level screening. Not only can the specific pathological characteristics of the abnormality be identified, but also the comprehensive judgment can be made combined with the previous medical history, thereby providing reliable support for clinical intervention or long-term correction scheme.

[0033] Preferably, the first-level screening module is further used for: According to the basic information, the height, weight, age and gender of the user are counted and classified to obtain the user's body shape feature data; According to the static posture image, the front view static image, the rear view static image and the side view static image of the user are respectively subjected to image feature extraction, and according to the extracted image feature information, a static posture analysis result is obtained; According to the dynamic posture image, motion trajectory analysis and joint angle calculation are performed to obtain a dynamic posture analysis result; According to the static posture analysis result, the dynamic posture analysis result and the user's body shape feature data, the user's spine posture abnormality risk is identified and preliminarily scored to obtain the first-level screening result.

[0034] Specifically, first of all, the collection and processing of basic information are defined and explained (such as height, weight, age, gender, and further calculated BMI, height and body shape ratio, etc.), the purpose of which is to provide individualized benchmarks for subsequent scale normalization of image features and risk discrimination; to achieve this, first of all, the collected basic information is checked and missing value processing is performed, and according to the preset rules or statistical model, the individual is divided into several body shape categories (such as small, normal, obese or children / adults / elderly groups), and the user body shape feature data is obtained. This processing can be completed in the terminal or cloud and saved in standardized numerical values and category identifiers, thereby improving the comparability and scoring fairness of subsequent comparison and reducing misjudgment caused by individual differences. Subsequently, for static posture images (including front, rear and side view photos or key frames), this step proposes to extract image features for each view respectively: for the front view, extract shoulder width, shoulder height difference, hip line offset, left-right symmetry, etc.; for the rear view, extract spine axis projection offset, scapula protrusion, waist line asymmetry, etc.; for the side view, perform spine curvature fitting (such as thoracic and lumbar segment kyphosis / lordosis angle), head and neck forward inclination measurement, etc.; the implementation method can use human key point detection (such as based on convolutional network or key point estimator), contour segmentation, curve fitting (polynomial or spline fitting) and angle / distance calculation, and scale normalization and confidence evaluation are performed on the measured parameters. The obtained static posture analysis results can be directly used to identify structural posture abnormalities and as quantitative input into the scoring module, thereby improving the detection sensitivity and positioning accuracy of static abnormalities. Next, for the processing of dynamic posture images, it is defined as analyzing the motion trajectory and calculating the joint angle of continuous action sequence, the purpose of which is to depict the posture instability, compensation mode or muscle group incoordination revealed in motion or functional action; the implementation details include extracting human key points from dynamic images in time sequence, filtering and trajectory smoothing for key point sequence, dividing action period (such as gait cycle or forward bending-returning upright cycle), calculating joint instantaneous angle and its time sequence change (such as shoulder, hip, knee angle curve), and further extracting stability indicators (angle variance, phase difference, symmetry index), speed / acceleration features, etc. If necessary, wearable IMU data can be fused to improve the time resolution, and the obtained dynamic posture analysis results are helpful for identifying functional abnormalities and motion compensation in dynamic state, complementing the blind area of static detection.Finally, the static posture analysis results, dynamic posture analysis results and user body shape feature data are normalized and fused (feature standardization, principal component analysis, weighted fusion or trained machine learning model can be used) to realize the identification and preliminary scoring of spinal posture abnormalities and obtain the final primary screening results; the purpose of this fusion and scoring is to reduce the noise influence of a single indicator through multi-source and multi-modal evidence to realize more robust risk determination, which can include threshold rules, weighted linear scoring or probability classifiers, and output low / medium / high risk stratification and necessary recheck or upgrade screening recommendations to improve the accuracy and coverage of early screening, reduce missed / incorrect diagnoses, and provide standardized, quantifiable and traceable data input for secondary and tertiary screening.

[0035] Preferably, the primary screening module extracts image features from the front, back and side static images of the user according to the static posture images, and obtains the static posture analysis results according to the extracted image feature information, including the following specific processes: Perform human key point recognition on the front static image to extract shoulder, neck and pelvic position features to obtain front view image feature information; Perform spine midline extraction and left-right symmetry analysis on the back static image to extract shoulder blade, spine midline and waist position features to obtain back view image feature information; Perform spine curvature fitting and head and neck position recognition on the side static image to extract thoracic kyphosis angle, lumbar lordosis angle and head forward inclination degree to obtain side view image feature information; Normalize and fuse the posture parameters under each viewing angle according to the front view image feature information, back view image feature information and side view image feature information to obtain multi-view comprehensive feature information; According to the multi-view comprehensive feature information, the static posture of the user is comprehensively evaluated to identify shoulder height difference, pelvic tilt angle and spinal curvature abnormalities to obtain the static posture analysis results; Specifically, the working goal of the first screening module is to realize comprehensive identification and evaluation of posture abnormalities by multi-view feature extraction and fusion on the static posture image of the user. Specifically, in the front-view static image processing, the system automatically identifies the key position features of the shoulder, neck and pelvis through a human key point detection algorithm (such as a human posture estimation model based on a deep convolutional neural network), and calculates parameters such as shoulder height difference, neck midline deviation and pelvis levelness, thereby obtaining front-view image feature information; in the back-view static image processing, the system focuses on extracting the spinal axis, identifying the overall trend of the spine through image segmentation and skeleton fitting methods, and extracting the shoulder blade protrusion degree, spinal deviation and waist line position features through a symmetry analysis algorithm, thereby obtaining back-view image feature information; in the side-view static image processing, a curve fitting algorithm is used to model the spinal curvature, measure the thoracic kyphosis and lumbar lordosis angles, and combine with head and neck position recognition to quantify the degree of head forward leaning, thereby forming side-view image feature information. Subsequently, the feature parameters under the three views are normalized to eliminate the influence of individual differences such as height and body type, and multi-view comprehensive feature information is generated through a feature fusion strategy (such as weighted average or principal component fusion). Finally, based on the comprehensive feature information, the first screening module systematically evaluates the static posture of the user, can accurately identify common posture abnormalities including shoulder height difference, pelvis tilt angle and spinal curvature abnormalities, and outputs complete static posture analysis results. Through multi-view joint and feature fusion, information loss and misjudgment caused by a single view are avoided, high-precision and comprehensive identification of spinal and posture abnormalities is realized, and reliable basic data support is provided for subsequent dynamic posture analysis and risk classification.

[0036] The specific process of the first screening module performing action trajectory analysis and joint angle calculation according to the dynamic posture image includes: According to the collected dynamic posture image, the continuous dynamic posture image frames are time-sequentially arranged to obtain a dynamic image sequence of the user's action; According to the dynamic image sequence, human key point recognition is performed to extract head, shoulder joint, hip joint, knee joint and ankle joint key point coordinates to obtain key point time sequence data; According to the key point time sequence data, trajectory fitting and path smoothing processing are performed on the action performed by the user to obtain joint and torso action trajectory curves; According to the key point time sequence data, the angle changes of the shoulder joint, hip joint, knee joint and other joints during the action are calculated to obtain joint angle curves; According to the action trajectory curve and the joint angle curve, the stability and action coordination of the user's posture are analyzed to obtain dynamic posture evaluation parameters; According to the dynamic posture evaluation parameters, the posture state of the user in the action process is comprehensively judged, and a dynamic posture analysis result is obtained.

[0037] Specifically, by continuously analyzing the dynamic posture images, the changes of the spine posture of the user in the movement process are comprehensively revealed. First, the system arranges the collected dynamic posture images in time sequence, and constructs each frame of image into a dynamic image sequence in time sequence, to ensure the continuity and traceability of the action analysis; then, through a human key point recognition algorithm (such as OpenPose or HRNet, etc.), the spatial coordinates of the head, shoulder joint, hip joint, knee joint and ankle joint are extracted frame by frame in the sequence, and a key point time sequence data changing with time is generated. On this basis, the trajectory fitting and path smoothing algorithm (such as Kalman filtering or B-spline fitting) is used to model the motion curve of these key points, and continuous curve data reflecting the motion trajectory of the trunk and limbs are obtained, while the interference caused by jitter or noise is reduced. Further, the system calculates the angle change of each main joint in the action process according to the key point time sequence, and forms the angle curve of the shoulder joint, hip joint, knee joint, etc., to quantify the action amplitude and posture change. Then, through comprehensive analysis of the action trajectory curve and the joint angle curve, the system can extract dynamic posture evaluation parameters, such as action stability (trajectory smoothness, repeatability) and action coordination (synchronization of upper and lower limb and trunk angle changes). Finally, based on these evaluation parameters, the system comprehensively judges the posture state of the user's spine in the whole action process, and generates a dynamic posture analysis result. Not only can it identify dynamic abnormalities that static posture cannot find, such as joint compensation or spine deviation during movement, but also can reflect the potential risks of user's spine health and movement function through stability and coordination analysis, providing a scientific basis for subsequent grading screening and personalized intervention.

[0038] Preferably, the secondary screening module is further used for: According to the primary screening result, screening out the medium and high risk users in the primary screening result as the secondary screening object; According to the second original data, posture reconstruction and surface fitting are performed on the three-dimensional image of the secondary screening object to obtain a three-dimensional body posture model of the spine and related bones; According to the three-dimensional body posture model, the curvature of the spine, the inclination angle of the pelvis, and the asymmetry of the shoulder and waist are calculated to obtain three-dimensional structure feature data; According to the dynamic posture image and the three-dimensional structure feature data, the stability of the spine, the coordination of the joints and the balance of the muscles of the secondary screening object in the action process are jointly analyzed to obtain dynamic function feature data; According to the three-dimensional structure feature data and the dynamic function feature data, a preset evaluation algorithm is used to quantitatively calculate the severity of the spinal posture abnormality of the secondary screening object, and a quantitative score result is obtained. According to the quantitative score result, the spinal posture abnormality of the secondary screening object is classified into different levels, and a secondary screening result is formed.

[0039] Specifically, the secondary screening module is further used to realize more refined three-dimensional evaluation on the medium-high risk users identified in the primary screening. First, the system automatically selects users with higher risk levels as secondary screening objects according to the primary screening result, so as to ensure that the screening resources are concentratedly applied to the key population. Subsequently, the system uses the three-dimensional images in the second original data to perform posture reconstruction and surface fitting on the users, so as to construct a three-dimensional body model containing the spine, pelvis and shoulder-waist structure. This model can truly restore the skeletal appearance features of the users. On this basis, the system further calculates parameters such as spinal curvature (such as thoracic kyphosis and lumbar lordosis angle), pelvic tilt angle and shoulder-waist asymmetry, and generates three-dimensional structure feature data reflecting the geometric state of the skeleton. Then, the system combines the dynamic posture images and the three-dimensional structure feature data to comprehensively analyze the spinal stability (such as dynamic central axis offset) involved in the action process, joint coordination (synchronization of upper and lower limbs and spine movement) and muscle balance (left and right torso stress difference), and obtains dynamic function feature data. Further, the system uses a preset evaluation algorithm (such as a weighted scoring model or a regression algorithm based on machine learning) based on the three-dimensional structure feature data and the dynamic function feature data to quantitatively calculate the severity of the spinal posture abnormality, and obtains a quantitative score result which can be directly compared. Finally, the system classifies the abnormality into different levels according to the score interval, such as mild, moderate and severe, and forms a secondary screening result. Through this step, not only is the transition from two-dimensional static analysis to joint evaluation of three-dimensional structure and dynamic function realized, but also more scientific quantitative grading results are provided, which provides accurate basis for subsequent tertiary screening or correction intervention, and significantly improves the accuracy and reliability of the screening.

[0040] Preferably, the specific process of the secondary screening module according to the three-dimensional structure feature data and the dynamic function feature data, using a preset evaluation algorithm to quantitatively calculate the severity of the spinal posture abnormality of the secondary screening object, to obtain a quantitative score result includes: The three-dimensional structure feature data and the dynamic function feature data are normalized and noise-filtered to obtain a standardized feature data set meeting the calculation requirements; From the standardized feature data set, posture abnormality representation indexes including spinal curvature, vertebral rotation angle, joint range of motion and motion coordination are extracted; assign a corresponding weight coefficient to the posture abnormality characterization indicator to obtain an indicator set with a weight coefficient; Based on the indicator set with a weight coefficient, the intermediate score of the spinal posture abnormality severity of the secondary screening object is calculated. The interval mapping and normalization processing are performed on the intermediate score result to obtain the standardized quantitative score result.

[0041] Specifically, the complex three-dimensional structural feature data and dynamic functional feature data are converted into quantifiable and comparable spinal posture abnormality severity scores. First, the system performs normalization processing and noise filtering on the three-dimensional structural feature data and dynamic functional feature data, such as through Z-score standardization or wavelet filtering, to remove noise caused by posture jitter, light interference or sensor error during the collection process, thereby obtaining a standardized feature data set that meets the calculation requirements. Subsequently, the system extracts core indicators that can represent abnormalities from the data set, including spinal curvature (such as thoracic kyphosis angle), vertebral rotation angle (reflecting the degree of spinal torsion), joint range of motion (such as hip or shoulder joint flexibility), and motion coordination (such as the synchronicity of upper and lower limb and spinal movement), to construct a posture abnormality characterization indicator set. On this basis, the system assigns weight coefficients to each indicator based on medical experience or data-driven methods, such as assigning a higher weight to the spinal curvature indicator to highlight key abnormal features, to obtain an indicator set with weights. Then, the system calculates an intermediate score based on the set, which can use weighted summation, fuzzy comprehensive evaluation or machine learning model-based prediction methods to integrate each indicator into a score reflecting the severity. Finally, to ensure the comparability of results between different individuals, the system performs interval mapping and normalization processing on the intermediate score, such as linearly mapping the result to the 0-100 interval to obtain a standardized quantitative score result. Through this process, complex posture and functional data are converted into intuitive score indicators, which not only facilitate clinicians to quickly judge the severity of abnormalities, but also provide users with understandable health risk levels, thereby improving the scientificity and application value of secondary screening.

[0042] Preferably, the tertiary screening module is further configured to: According to the third original data, the collected medical image data is subjected to format conversion, denoising and enhancement processing, and the user's historical disease information is subjected to structured arrangement to obtain a standardized medical analysis data set; Multi-dimensional feature extraction is performed on the standardized medical image data to obtain image feature parameters, wherein the image feature parameters include spinal arrangement features, vertebral morphology features, intervertebral space features and soft tissue compression features; correlate the historical disease information with the image feature parameters, extract a disease history feature index related to the spinal posture abnormality, and obtain a comprehensive feature index set; According to the comprehensive feature index set, pathological risk calculation and abnormal pattern recognition are performed on the medium and high risk users in the secondary screening result, and an evaluation intermediate result is obtained. Based on the evaluation intermediate result, severity grading determination is performed on the spinal posture abnormality, and the tertiary screening result is output in combination with a preset medical clinical reference standard.

[0043] Specifically, the tertiary screening module is also used for more professional and clinical level evaluation for the medium and high risk users identified in the secondary screening. First, the system processes the third original data, that is, converts the collected medical image data (such as X-ray film, CT or MRI image) into a format to ensure its compatibility with the unified standard in subsequent analysis, and improves the image clarity through denoising and image enhancement technology (such as histogram equalization, convolution filtering), while the historical disease information of the user (such as scoliosis history, osteoporosis or trauma history) is structured and arranged to form a standardized medical analysis data set that can be directly called. Subsequently, the system performs multi-dimensional feature extraction on the standardized medical image data, and the extracted image feature parameters include spinal arrangement features (such as straightness or curvature between vertebral bodies), vertebral body shape features (such as wedge deformation, collapse), intervertebral space features (such as stenosis or uneven distribution), and soft tissue compression features (such as muscle or intervertebral disc compression), so as to accurately depict the medical state of the user's spine. Then, the system performs correlation analysis on the image feature parameters and the historical disease information, for example, in combination with the previous spinal trauma history and the abnormal vertebral body condition in the current image, the most relevant disease history feature index related to the spinal posture abnormality is extracted, thereby obtaining a comprehensive feature index set. Based on the comprehensive index set, the system performs pathological risk calculation and abnormal pattern recognition on the secondary screening objects, for example, through pattern matching or machine learning algorithm, identifies typical abnormal patterns such as scoliosis, intervertebral disc herniation or bone degeneration, and obtains an evaluation intermediate result. Finally, the system performs severity grading determination on the spinal posture abnormality based on the evaluation intermediate result according to the medical clinical reference standard (such as Cobb angle grading, WHO bone density standard), and outputs the standardized tertiary screening result. This process ensures the medical reliability and clinical interpretability of the screening result, not only improves the accuracy of the screening, but also provides a scientific basis for subsequent personalized correction and medical intervention.

[0044] Preferably, referring to Figure 2 , the comprehensive evaluation module is also used for: According to the instruction of the control module, the primary screening result, the secondary screening result and the tertiary screening result are received, and the data of different screening levels are classified and stored and structured, and a to-be-comprehensively-analyzed data set is obtained. uniformly converting and standardizing dimensions of different screening indicators in the to-be-comprehensively-analyzed data set, ensuring comparability of results at different levels, and obtaining a standardized screening result set; performing feature fusion on body shape feature data, static posture analysis results, dynamic posture analysis results, three-dimensional structure feature data, and medical image feature data in the standardized screening result set, and obtaining a multi-dimensional comprehensive feature vector; performing weight distribution and weighted calculation on the multi-dimensional comprehensive feature vector, and obtaining a comprehensive risk score of the user's spinal posture abnormality; According to the comprehensive risk score, combined with the preset risk classification standard, the degree of spinal posture abnormality of the user is graded and determined, and the comprehensive evaluation result is obtained.

[0045] Specifically, the comprehensive evaluation module is used to uniformly integrate and comprehensively determine data from different screening levels to form the final spinal posture abnormality evaluation result. First, the comprehensive evaluation module receives the first-level screening result, the second-level screening result, and the third-level screening result under the scheduling of the control module, and stores them according to the data source and type, and uniformly organizes the data of different structures to obtain a to-be-comprehensively-analyzed data set convenient for calling and operation. Then, the system performs dimension conversion and standardization processing on various screening indicators (such as body shape features, angle measurement values, image parameters, etc.) in the data set, solves the problem of inconsistent units and dimensional differences under different data sources, and thus obtains a standardized screening result set with comparability. Then, the system performs feature fusion on the key data in the result set, including body shape feature data, static posture analysis results, dynamic posture analysis results, three-dimensional structure feature data, and medical image feature data, to generate a multi-dimensional comprehensive feature vector that can comprehensively reflect the user's spinal state. On this basis, the system performs weight distribution on each feature in the multi-dimensional comprehensive feature vector according to its importance, and obtains a comprehensive risk score of the user's spinal posture abnormality through weighted calculation. Finally, the system grades and determines the degree of spinal posture abnormality of the user according to the comprehensive risk score, combined with the preset risk classification standard (such as low-risk, medium-risk, and high-risk level thresholds), and outputs the final comprehensive evaluation result. This process realizes the unified processing and fusion analysis of cross-level, multi-dimensional data, can avoid the limitations of single screening level results, improves the scientificity and accuracy of the evaluation, and provides more solid data support for the subsequent development of personalized correction strategies.

[0046] Preferably, the correction management module is further used for: According to the comprehensive evaluation result, the level of spinal posture abnormality, the abnormal position, and the risk factors of the user are analyzed to obtain personalized correction requirement information; According to the basic information of the user in the first original data, the basic body condition of the user is matched to obtain an individual adaptability parameter; According to the individual adaptability parameter and the individualized correction requirement information, a correction target of the user is determined, and the correction target includes a correction part, a correction amplitude, and a correction period. According to the correction target, a correction method matched with the target is called from a preset correction strategy library, including a posture training scheme, a motion intervention scheme, and an auxiliary appliance use scheme, and combination optimization is performed to obtain the individualized correction strategy.

[0047] Specifically, the working mechanism of the correction management module is an individualized intervention design system for posture problems of the user's spine. It will first combine the comprehensive evaluation results to carefully analyze the abnormal level, abnormal part and potential risk factors of the user, so as to determine the specific correction needs of the user. Then, the system will calculate the body adaptability parameter of the user based on the basic information of the user in the first original data (such as age, gender, height, weight and health status), to ensure the safety and feasibility of the subsequent intervention scheme. On this basis, the module will match the requirement information with the individual parameter to determine the specific correction target, including the specific part of the correction, the amplitude that needs to be adjusted and the period required. Finally, the system will screen out the intervention methods matched with these targets from the preset correction strategy library, such as posture training, motion rehabilitation and the use of auxiliary appliances, and through the combination optimization algorithm, the various methods are organically integrated to finally form a scientific, individualized and dynamically adjustable correction strategy. In this way, not only the pertinence of the correction effect is improved, but also the user's compliance and safety are enhanced.

[0048] Embodiment 2 The embodiment 2 of the present application also provides a three-level screening and correction method for spinal posture abnormalities, which is realized based on the spinal degenerative disease grading early warning and rehabilitation management system as described in embodiment 1.

[0049] Specifically, by adopting the spine posture abnormality three-stage screening and correction method provided in Embodiment 2 of the present application, a three-stage screening and correction system including data acquisition, hierarchical screening, comprehensive evaluation and correction management is constructed, effectively solving the technical problems of single screening mode, lack of hierarchical progression and non-personalized correction strategy in the prior art. Specifically, the data acquisition module can acquire basic information, posture images, three-dimensional images and medical images of users according to different precision requirements, thereby forming multi-level original data support; by setting the first-stage, second-stage and third-stage screening modules, the risk identification and quantitative evaluation are realized from shallow to deep and gradually progressive, overcoming the shortcomings of insufficient screening dimensions and difficulty in accurately judging the severity of spine posture abnormalities in the traditional method; the comprehensive evaluation module can integrate the screening results at all stages to form a comprehensive evaluation conclusion, improving the scientificity and reliability of diagnosis; and the correction management module automatically generates personalized correction strategies for different users based on the comprehensive evaluation results and individual characteristic data, thereby breaking through the limitations of the existing one-size-fits-all correction method and realizing personalized intervention and precise management.

[0050] In summary, the present application provides a spine posture abnormality three-stage screening and correction system and method.

[0051] It should be noted that the present application is not limited to the specific configurations and processes described above and shown in the drawings. For the sake of brevity, detailed descriptions of well-known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order of the steps, after understanding the spirit of the present application.

[0052] The functional blocks shown in the structure block diagram described above can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a functional card, etc. When implemented in software, the elements of the present application are program or code segments used to perform the required tasks. The program or code segments can be stored in a machine-readable medium or transmitted through a data signal carried in a carrier wave over a transmission medium or communication link. The "machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segments can be downloaded via a computer network such as the Internet, an intranet, etc.

[0053] The user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are information and data authorized by the user or authorized by all parties, and the collection, use and processing of the relevant data need to comply with the relevant laws, regulations and standards of the place and provide corresponding operation entrances for the user to choose authorization or refusal.

[0054] It should be further noted that the exemplary embodiments mentioned in the present application describe some methods or systems based on a series of steps or devices. However, the present application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiments, or different from the order in the embodiments, or several steps are executed simultaneously.

[0055] The above is only a specific implementation of the present application, and those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system, module and unit described above can refer to the corresponding process in the foregoing method embodiments, which will not be described here. It should be understood that the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of various equivalent modifications or replacements within the technical range disclosed by the present application, and these modifications or replacements should be covered within the protection scope of the present application.

Claims

1. A three-tiered screening and correction system for spinal posture abnormalities, characterized in that, The system includes: The control module and its connected data acquisition module, primary screening module, secondary screening module, tertiary screening module, comprehensive assessment module, and correction management module; The data acquisition module is used to collect first raw data related to the user's spinal posture for primary screening, second raw data for secondary screening, and third raw data for tertiary screening. The first-level screening module is used to perform preliminary analysis on the first raw data, identify the user's risk of spinal posture abnormalities, and obtain the first-level screening results; The secondary screening module is used to further analyze and quantify the medium- and high-risk users in the primary screening results based on the second raw data, and obtain the secondary screening results. The third-level screening module is used to conduct professional assessments of medium- and high-risk users in the second-level screening results based on the third raw data, and to obtain the third-level screening results. The comprehensive evaluation module is used to comprehensively analyze the results of the primary screening, secondary screening, and tertiary screening to obtain a comprehensive evaluation result; The correction management module is used to generate a personalized correction strategy corresponding to the user based on the comprehensive evaluation results and the first raw data. The control module is used to manage the workflow of each module, coordinate the data collection, screening, evaluation and correction management process, monitor the system status and optimize screening and correction strategies.

2. The three-level screening and correction system for spinal posture abnormalities according to claim 1, characterized in that, The data acquisition module is also used for: Obtain the preset primary screening accuracy requirements, secondary screening accuracy requirements, and tertiary screening accuracy requirements; Based on the first-level screening accuracy requirements, the user's basic information, static posture images, and dynamic posture images are collected to obtain the first raw data; Based on the required accuracy of the secondary screening, the user's three-dimensional image is acquired, and the second raw data is determined based on the three-dimensional image and the dynamic posture image. Based on the accuracy requirements of the three-level screening, the user's medical imaging data and historical medical information are collected to obtain the third raw data.

3. The three-level screening and correction system for spinal posture abnormalities according to claim 2, characterized in that, The primary screening module is also used for: Based on the basic information, the user's height, weight, age, and gender are statistically analyzed and categorized to obtain the user's body shape characteristic data; Based on the static posture image, image features are extracted from the user's front-view static image, rear-view static image, and side-view static image, respectively. Based on the extracted image feature information, the static posture analysis result is obtained. Based on the dynamic posture image, motion trajectory analysis and joint angle calculation are performed to obtain the dynamic posture analysis results; Based on the static posture analysis results, dynamic posture analysis results, and user body shape characteristic data, the risk of abnormal spinal posture in users is identified and preliminarily scored, resulting in the first-level screening results.

4. The three-level screening and correction system for spinal posture abnormalities according to claim 3, characterized in that, The primary screening module extracts image features from the user's forward-looking, backward-looking, and side-looking static images based on the static posture images. The specific process of obtaining the static posture analysis results based on the extracted image feature information includes: Human key point recognition is performed on the aforementioned frontal static image to extract the positional features of the shoulder, neck, and pelvis, thereby obtaining the frontal image feature information; The spinal midline and left-right symmetry are extracted from the rear-view static image, and the positional features of the scapula, spinal midline and lumbar region are extracted to obtain the rear-view image feature information; By fitting the spinal curvature and identifying the head and neck position of the side-view static image, the thoracic kyphosis angle, lumbar lordosis angle and the degree of head forward tilt are extracted to obtain the side-view image feature information. Based on the front view image feature information, rear view image feature information and side view image feature information, the pose parameters under each viewpoint are normalized and feature fused to obtain multi-view comprehensive feature information; Based on the multi-view integrated feature information, the user's static posture is comprehensively evaluated to identify shoulder height difference, pelvic tilt angle and abnormal spinal curvature, and the static posture analysis results are obtained. The specific process by which the primary screening module performs motion trajectory analysis and joint angle calculation based on the dynamic posture image to obtain the dynamic posture analysis result includes: Based on the acquired dynamic posture images, the continuous dynamic posture image frames are time-series organized to obtain a dynamic image sequence of user actions. Based on the dynamic image sequence, human key point recognition is performed, and the coordinates of key points such as head, shoulder joint, hip joint, knee joint and ankle joint are extracted to obtain key point time series data. Based on key point time series data, trajectory fitting and path smoothing are performed on the user's actions to obtain the motion trajectory curves of joints and torso. Based on key point time series data, calculate the angle changes of joints such as shoulder, hip, and knee during the movement process to obtain joint angle curves; Based on the motion trajectory curve and joint angle curve, the stability and coordination of the user's posture are analyzed to obtain dynamic posture evaluation parameters. Based on the dynamic posture evaluation parameters, the user's spinal posture during the movement is comprehensively judged to obtain the dynamic posture analysis results.

5. A three-level screening and correction system for spinal posture abnormalities according to claim 2, characterized in that, The secondary screening module is also used for: Based on the results of the primary screening, users with medium to high risk in the primary screening results are selected as subjects for secondary screening. Based on the second raw data, pose reconstruction and surface fitting are performed on the three-dimensional images of the secondary screening objects to obtain a three-dimensional body model of the spine and related bones; Based on the three-dimensional body model, the spinal curvature, pelvic tilt angle, and shoulder-lumbar asymmetry are calculated to obtain three-dimensional structural feature data. Based on the dynamic posture images and the three-dimensional structural feature data, the spinal stability, joint coordination, and muscle balance of the secondary screening object during the movement process are jointly analyzed to obtain dynamic functional feature data; Based on the three-dimensional structural feature data and dynamic functional feature data, a preset evaluation algorithm is used to quantify the severity of spinal posture abnormalities in the secondary screening subjects, and obtain a quantitative scoring result. Based on the quantitative scoring results, the spinal posture abnormalities of the secondary screening subjects are classified into levels to form secondary screening results.

6. The three-level screening and correction system for spinal posture abnormalities according to claim 5, characterized in that, The secondary screening module, based on the three-dimensional structural feature data and dynamic functional feature data, uses a preset evaluation algorithm to quantify the severity of spinal posture abnormalities in the secondary screening subjects, and obtains the quantitative score result through the following specific process: The three-dimensional structural feature data and dynamic functional feature data are normalized and noise filtered to obtain a standardized feature dataset that meets the computational requirements. From the standardized feature dataset, extract postural abnormality characterization indicators including spinal curvature, vertebral rotation angle, joint range of motion, and movement coordination. Assign corresponding weight coefficients to the posture anomaly characterization indicators to obtain a set of indicators with weight coefficients; Based on the set of weighted indicators, an intermediate score of the severity of spinal posture abnormalities in the secondary screening subjects is calculated. The intermediate scoring results are subjected to interval mapping and normalization to obtain the standardized quantitative scoring results.

7. A three-level screening and correction system for spinal posture abnormalities according to claim 2, characterized in that, The three-level screening module is also used for: Based on the aforementioned third raw data, the collected medical image data undergoes format conversion, noise reduction, and enhancement processing, and the user's historical medical information is structurally organized to obtain a standardized medical analysis dataset. Multi-dimensional feature extraction is performed on the standardized medical image data to obtain image feature parameters, which include spinal alignment features, vertebral body morphology features, intervertebral space features, and soft tissue compression features. The historical disease information is correlated with the image feature parameters to extract the medical history feature indicators related to spinal posture abnormalities, and a comprehensive feature indicator set is obtained. Based on the comprehensive feature index set, pathological risk calculation and abnormal pattern identification are performed on medium- and high-risk users in the secondary screening results to obtain intermediate evaluation results; Based on the intermediate assessment results, the severity of spinal posture abnormalities is graded and determined, and the three-level screening results are output in conjunction with preset medical clinical reference standards.

8. A three-level screening and correction system for spinal posture abnormalities according to claim 1, characterized in that, The comprehensive evaluation module is also used for: According to the instructions of the control module, the first-level screening results, second-level screening results and third-level screening results are received, and the data of different screening levels are classified, stored and structured to obtain the dataset to be comprehensively analyzed. The screening indicators at different levels in the dataset to be comprehensively analyzed are uniformly converted and standardized to ensure that the results at different levels are comparable, thus obtaining a standardized screening result set. The body shape feature data, static posture analysis results, dynamic posture analysis results, three-dimensional structural feature data, and medical imaging feature data in the standardized screening result set are fused to obtain a multi-dimensional comprehensive feature vector. The multidimensional comprehensive feature vector is weighted and weighted to obtain a comprehensive risk score for the user's spinal posture abnormality. Based on the comprehensive risk score and the preset risk grading standards, the degree of spinal posture abnormality of the user is classified and judged to obtain the comprehensive assessment result.

9. A three-level screening and correction system for spinal posture abnormalities according to claim 2, characterized in that, The correction management module is also used for: Based on the comprehensive assessment results, the user's spinal posture abnormality level, abnormal location, and risk factors are analyzed to obtain personalized correction needs information; Based on the user's basic information in the first raw data, the user's basic physical conditions are matched to obtain individual adaptability parameters; Based on the personalized correction needs information and individual adaptability parameters, the user's correction goals are determined, including the correction area, correction range, and correction period; Based on the correction goal, a correction method matching the goal is called from a preset correction strategy library, including posture training programs, exercise intervention programs and assistive device usage programs, and combined and optimized to obtain the personalized correction strategy.

10. A three-tiered screening and correction method for spinal posture abnormalities, characterized in that, This is based on the spinal degenerative disease grading, early warning, and rehabilitation management system as described in any one of claims 1-9.

Images