Knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis

Abstract

The invention relates to the technical field of medical data processing, and discloses a knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis, which comprises a data processing module for acquiring knee joint three-dimensional movement original data by adopting Magic software, performing coordinate correction and model processing, and generating a model group. And the model analysis module imports the model group into a C3D clinical window to realize real-time visualization of a three-dimensional moving track, biomechanical parameter calculation and clinical pathology correlation evaluation. The man-machine interaction control module provides customized interaction logic and a graphical interface, and supports parameter input, analysis process control, real-time monitoring and automatic generation of a diagnosis report. The cloud publishing module converts a diagnosis report and an interface into a PHP dynamic webpage, and supports SSL encryption, multi-user concurrent access and cross-platform remote calling through HTTP deployment. Real-time visualization of the knee joint three-dimensional movement track is supported, and the data reliability is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical data processing, in particular to a knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis. BACKGROUND

[0002] Knee joint diseases have a high incidence in clinical practice, especially osteoarthritis, ligament injury and meniscus lesions, which seriously affect the quality of life of patients. Traditional diagnostic methods mainly rely on static imaging examinations (such as X-ray, CT, MRI) and physical examinations, but these methods only provide anatomical structure information of the knee joint, and cannot fully reflect the spatial movement characteristics and biomechanical state of the knee joint in the dynamic movement process, limiting the accuracy of clinical diagnosis and the feasibility of early intervention.

[0003] The existing knee joint movement analysis technology mainly has the problems of limited data precision, two-dimensional image or single sensor data cannot completely capture the three-dimensional movement information of the knee joint with multiple degrees of freedom, is easily affected by noise and sensor errors, leading to movement parameter deviation, and is difficult to support accurate clinical judgment; insufficient model analysis capability, the existing methods mainly rely on manual annotation or simplified models for key point recognition and movement trajectory analysis, lack of automatic and accurate three-dimensional knee joint modeling and key anatomical point recognition technology, and it is difficult to realize high-precision biomechanical parameter calculation and disease correlation evaluation; limited interaction and visualization, traditional systems lack flexible human-computer interaction interface and customized analysis process, cannot meet the needs of different clinical applications, lack of real-time visualization and automatic generation of diagnosis report capability, affecting the intuitive understanding and judgment of the dynamic state of the patient by the clinician.

[0004] Therefore, it is necessary to design a knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis to solve the problems in the prior art. SUMMARY

[0005] In view of this, the present application provides a knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis, aiming to solve the problems of low data precision, insufficient model analysis and limited interaction and visualization.

[0006] The present application provides a knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis, comprising: a data processing module, a model analysis module, a human-computer interaction control module and a cloud publishing module; The data processing module is used to capture original movement data of the knee joint by using Magic software, and to perform coordinate correction and model processing operations; the coordinate correction includes noise filtering and coordinate systemization of the original movement data, and the model processing operation includes simplification, key anatomical point labeling and precision optimization of the knee joint three-dimensional model, to generate a standardized model group; The model analysis module is used for seamlessly importing the model group into C3D clinical window software, and the C3D clinical window provides functions of real-time visualization of three-dimensional movement trajectory of the knee joint, calculation of biomechanical parameters, and evaluation of clinical pathology correlation; The human-computer interaction control module is used for writing customized interaction logic in combination with application requirements of clinical diagnosis, integrating a clinical protocol template, and providing a graphical user interface to realize user parameter input, analysis process control, real-time result monitoring, and automatic generation of a diagnosis report. The cloud publishing module is used for converting the diagnosis report and the interaction interface into a PHP dynamic web page format, deploying and publishing on a cloud server through an HTTP protocol, and supporting SSL encrypted transmission, multi-user concurrent access, and cross-platform remote calling.

[0007] Further, when the data processing module performs noise filtering and coordinate systematization on the original movement data, the following steps are included: The noise filtering uses an extended Kalman filter to define a state vector containing joint angles, angular velocities and angular accelerations, to adaptively adjust process noise covariance parameters according to a kinematic model, to increase a tolerance range when joint acceleration changes exceed an acceleration change threshold, and to fuse optical and inertial sensor data streams in a measurement update stage to perform a residual threshold monitoring process to eliminate abnormal measurement values. The coordinate systematization initializes a calibration process based on a preoperative bone CT scan model through a quaternion rotation matrix algorithm, extracts geometric center points of a distal femoral condyle and a tibial plateau to construct a local coordinate system, uses a quaternion to represent a rotation operation and solves optimal rotation parameters by minimizing point cloud registration errors, and converts data to a unified anatomical coordinate system.

[0008] Further, when the data processing module simplifies a three-dimensional model of the knee joint, marks key anatomical points, and optimizes precision, the following steps are included: The three-dimensional model of the knee joint is simplified by using a vertex clustering polygon reduction algorithm to divide a triangular mesh into an octree structure, analyzing vertex normal vector consistency in a region node, merging vertices when curvature changes are lower than a curvature change threshold, and preserving topological details of a condyle key anatomical region; key anatomical point marking is realized through a pre-trained U-Net convolutional neural network.

[0009] Further, when the data processing module realizes key anatomical point marking through a pre-trained U-Net convolutional neural network, the following steps are included: The pre-trained U-Net convolutional neural network comprises a down-sampling layer and an up-sampling layer, inputs three-dimensional voxel grid data, and outputs a probability distribution heat map of the lower pole of the patella and the anterior horn of the meniscus; the precision optimization adopts an iterative closest point algorithm to perform model registration operation, sets a convergence condition as a root mean square error being lower than an error threshold or a maximum iteration number being reached, and combines a Laplace smoothing filter to eliminate surface noise interference.

[0010] Further, when the C3D clinical window provides real-time visualization of knee joint three-dimensional movement trajectory, biomechanical parameter calculation, and clinical pathology correlation evaluation functions, it includes: The biomechanical parameter calculation adopts an inverse dynamics model to calculate joint torque and contact force in real time in combination with a musculoskeletal simulation toolbox; the clinical pathology correlation evaluation is realized based on a support vector machine classifier, a feature vector integrates torque peak value, contact force time integral, and trajectory asymmetry index in a gait cycle, a radial basis kernel function is used for pattern recognition; the classifier executes a training process on an osteoarthritis case database, and outputs a quantitative risk score.

[0011] Further, when the biomechanical parameter calculation adopts an inverse dynamics model to calculate joint torque and contact force in real time in combination with a musculoskeletal simulation toolbox, it includes: Based on kinematic data and ground reaction force input, a motion equation containing a mass matrix, a Coriolis force term, and a gravity term is solved; the contact force calculation adopts a Hertz contact theory model, inputs meniscus curvature radius and material elastic modulus parameters, and generates a tibiofemoral joint pressure distribution visualization cloud chart.

[0012] Further, the human-computer interaction control module combines clinical diagnosis application requirements to write customized interaction logic; integrates a clinical protocol template, and provides a graphical user interface to realize user parameter input, analysis process control, real-time result monitoring, and automatic generation of a diagnosis report, including: The human-computer interaction control module integrates a visual drag-and-drop process editor in the customized interaction logic, supports defining an analysis step sequence through a graphical node interface; data flow between nodes is executed by an event-driven mechanism to perform parameter validity verification operation; the clinical protocol template contains an osteoarthritis grading evaluation template, predefines an anatomical index threshold and loads patient historical data; the graphical user interface provides real-time 3D model interaction function, realizes touch zooming operation, motion trajectory frame-by-frame playback, and diagnosis report generation process based on a WebGL rendering engine.

[0013] Further, when the human-computer interaction control module performs a diagnosis report generation process based on a WebGL rendering engine, it includes: Integrate interface visual elements, embed dynamic biomechanical parameter charts; the interaction logic performs user parameter input, analysis process control, and real-time result monitoring operations.

[0014] Further, the cloud publishing module converts the diagnostic report and the interactive interface into a PHP dynamic web page format, including: The cloud publishing module uses a model-view-controller architecture of a Laravel framework to implement data processing logic; a controller responds to an HTTP request and calls a database model to extract biomechanical parameters; a view layer uses a template engine to dynamically render interface content, integrates an AJAX asynchronous technology to perform a data non-refresh loading process, and updates a result through a lightweight request; an interactive chart is generated using a visualization library, supports joint torque polar coordinate chart and pathological risk trend line display; and a web page performance is optimized through resource compilation and compression technology to perform a page loading process.

[0015] Further, the cloud publishing module is deployed and published on a cloud server through an HTTP protocol, supports SSL encrypted transmission, multi-user concurrent access, and cross-platform remote calling, including: In the HTTP protocol deployment, the SSL encrypted transmission uses a TLS protocol and a high-strength key to implement end-to-end data protection; a server sends a digital certificate chain when a client is connected, and generates a secure session key through a key exchange mechanism; the multi-user concurrent access implements traffic distribution operation through a load balancer, configures a weighted round robin algorithm to dynamically adjust a connection strategy according to server load, and sets a session retention mechanism; and the cross-platform remote calling supports RESTful API interface design, defines a data access endpoint, and transmits structured information in a JSON data format.

[0016] Compared with the prior art, the knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis provided by the embodiment of the application has the beneficial effects that: three-dimensional movement data of a knee joint are collected by Magic software, and a noise filtering and coordinate systematization are performed by combining an extended Kalman filter and a quaternion rotation matrix algorithm, high-precision capture and standardized processing of multi-degree-of-freedom knee joint movement are realized, and data reliability and analysis accuracy are improved. A vertex clustering polygon reduction algorithm is used to simplify a three-dimensional model, key anatomical points are automatically labeled by a pre-trained U-Net convolutional neural network, and precision optimization is realized by combining an iterative closest point algorithm and a Laplace smoothing filter, so that the three-dimensional model of the knee joint can be used for high-precision biomechanical analysis. The C3D clinical window supports real-time visualization of the three-dimensional movement trajectory of the knee joint, inverse dynamics calculation of joint torque and contact force, and clinical pathology correlation evaluation based on a support vector machine, realizes gait cycle analysis, mechanical parameter peak value statistics, and risk score output. The man-machine interaction control module realizes interactive parameter input, analysis process control, three-dimensional model real-time operation, and automatic generation of a diagnosis report by using a graphical drag-and-drop process editor and a WebGL rendering engine, and improves the convenience of clinical use. The cloud publishing module converts the diagnosis report and the interactive interface into a PHP dynamic web page format by using a Laravel MVC architecture, is deployed through an HTTP protocol, supports SSL encryption, load balancing, multi-user concurrent access, and cross-platform RESTful API calling, realizes data security sharing and remote collaboration. BRIEF DESCRIPTION OF DRAWINGS

[0017] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of preferred embodiments, and are not meant to limit the present application. Furthermore, the same reference numerals in different drawings are intended to represent the same components throughout the different drawings. In the drawings: Figure 1 A functional block diagram of a knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis provided by an embodiment of the application is shown. DETAILED DESCRIPTION

[0018] Exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. Rather, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be accurately conveyed to those skilled in the art. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0019] The current prior art mainly adopts a combination of medical imaging and computer modeling to perform three-dimensional reconstruction and kinematic analysis of the knee joint. Image data of the knee joint is obtained through MRI or CT scanning, and the software for medical image processing (such as Mimics, Magic, etc.) is used to segment the bones and soft tissues, and then a three-dimensional anatomical model is established. Subsequently, the kinematic data of the patient's knee joint in a static or dynamic state is collected by using a motion capture system, and the kinematic data is fused with the three-dimensional model through a registration algorithm. On this basis, the doctor can realize the visualization of the kinematic trajectory of the knee joint and the calculation of part of the biomechanical parameters in the C3D clinical analysis software or similar platform, which provides a reference for surgical planning, rehabilitation training and disease diagnosis. However, the data from different image sources in the prior art often has errors in coordinate registration, especially at key anatomical marker points (such as the lateral epicondyle, medial epicondyle and tibial plateau edge), which affects the reliability of the three-dimensional model. The existing images are mostly static scans, and it is difficult to realize high-resolution and continuous kinematic state modeling, resulting in a gap between the kinematic analysis results and the actual physiological conditions. Most systems can only provide kinematic trajectory display and basic parameter calculation, and lack intelligent analysis, pathological prediction and personalized diagnosis and treatment support.

[0020] For example, in a preoperative evaluation of an anterior cruciate ligament (ACL) reconstruction surgery, the doctor obtains the ligament and soft tissue data of the knee joint by using MRI, and supplements the bone structure by CT scanning. Due to the slight change in the posture of the knee joint of the patient during the two imaging processes, the coordinate offset occurs after the fusion of the MRI and CT data, and the registration error of the femur and tibia reaches 2-3 mm. This error directly leads to inaccurate measurement results of the ACL insertion point, and the doctor deviates when drilling the graft tunnel during the operation. After the operation, the patient still feels instability of the knee joint during walking and weight-bearing activities, and has to undergo a second correction surgery, which prolongs the rehabilitation period and increases the medical cost.

[0021] In view of the above problems, it is particularly necessary to design a knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis.

[0022] To this end, referring to Figure 1 , a knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis includes: a data processing module, a model analysis module, a human-computer interaction control module and a cloud publishing module; The data processing module is used to capture the original movement data of the knee joint by using the Magic software, and to perform coordinate correction and model processing operations. The coordinate correction includes noise filtering and coordinate systemization of the original movement data, and the model processing operation includes simplification, key anatomical point marking and precision optimization of the three-dimensional model of the knee joint, to generate a standardized model group; A model analysis module for seamlessly importing the model group into C3D clinical window software, which provides real-time visualization of three-dimensional movement trajectory of the knee joint, biomechanical parameter calculation and clinical pathology correlation evaluation function; A human-computer interaction control module for writing customized interaction logic combined with clinical diagnosis application requirements; integrating clinical protocol templates to provide a graphical user interface to achieve user parameter input, analysis process control, real-time result monitoring and automatic generation of diagnosis report; A cloud publishing module for converting the diagnosis report and the interaction interface into PHP dynamic web page format, deploying and publishing on the cloud server through HTTP protocol, supporting SSL encryption transmission, multi-user concurrent access and cross-platform remote calling.

[0023] Specifically, the knee joint three-dimensional movement analysis platform is composed of a data processing module, a model analysis module, a human-computer interaction control module and a cloud publishing module, forming a complete clinical auxiliary diagnosis and information sharing system. The data processing module processes the original data obtained by the knee joint three-dimensional action capture based on Magic software, realizes coordinate correction through noise filtering and coordinate systemization, and then performs simplification, key anatomical point marking and precision optimization on the three-dimensional model, thereby generating a standardized model group that can be used in clinical practice. The model analysis module seamlessly imports the standardized model group into C3D clinical window software, supports real-time three-dimensional visualization of knee joint movement trajectory, can calculate joint mechanics parameters, and conducts pathological correlation analysis combined with clinical requirements, providing quantitative basis for disease assessment and surgical planning. The human-computer interaction control module designs interaction logic combined with clinical diagnosis process, integrates multiple clinical protocol templates, provides a friendly graphical user interface, and supports doctors to complete parameter input, analysis process control, real-time result monitoring and automatic generation of diagnosis report. The cloud publishing module converts the generated diagnosis report and interaction interface into PHP dynamic web page format, realizes remote deployment through HTTP protocol, has the functions of SSL encryption, multi-user concurrent access and cross-platform calling, thereby realizing data sharing and remote diagnosis between different clinical institutions.

[0024] The working principle and process of the present application are that the data processing module is responsible for completing the collection and preprocessing of the original action data. After obtaining the three-dimensional action capture data of the knee joint through the Magic software, the original data is filtered to eliminate errors caused by external interference, and a unified world coordinate system is established through the coordinate system method, thereby ensuring the consistency and accuracy of subsequent model analysis. On this basis, the three-dimensional anatomical model of the knee joint is processed, including model simplification to reduce redundant data, key anatomical point marking to facilitate clinical identification, and precision optimization to improve the reliability of the model in action and biomechanical analysis, and finally a standardized model group is generated. The model analysis module seamlessly imports the standardized model group into the C3D clinical window software. This module presents the trajectory changes of the knee joint in different action states through three-dimensional visualization function, realizes the intuitive display of joint dynamic action, and can calculate the key action parameters (such as joint angle, displacement, load and stress distribution) in real time, and realize disease correlation analysis combined with pathological information database, thereby providing doctors with scientific quantitative diagnostic basis. The man-machine interaction control module provides an intuitive operation interface for clinicians. The module is embedded with customized interactive logic and clinical protocol templates, and users can input parameters, control analysis process and monitor results through the interface. In the background, the calculation and data calling are automatically executed according to the user's set diagnosis and treatment requirements, and the standardized diagnosis report is automatically generated after the analysis is completed, shortening the operation process of clinical diagnosis. The cloud publishing module realizes the remote sharing of diagnosis results. This module converts the diagnosis report and interactive interface into PHP dynamic web pages, deploys and publishes them on the cloud server through HTTP protocol, and uses SSL encryption to ensure the security of data transmission. At the same time, it supports multi-user concurrent access and cross-platform remote calling, and doctors can quickly retrieve the results on different terminal devices, which is convenient for multi-center consultation and remote medical application. Through the closed-loop working mechanism of "data processing-model analysis-man-machine interaction-cloud publishing", the platform realizes the accurate acquisition, intelligent analysis, visual display and remote sharing of three-dimensional action data of the knee joint.

[0025] As a preferred embodiment, the scheme of the present application is implemented as follows: the data processing module uses Magic software to process the original three-dimensional movement data of the patient's knee joint. The noise in the original data is filtered out by the extended Kalman filter method, and the anatomical coordinate system is established based on the bone structure of the patient's preoperative CT scan, realizing the coordinate systematization of the movement data; the polygon simplification algorithm is used to simplify the structure of the three-dimensional model of the knee joint, the pre-trained U-Net convolutional neural network is used to mark the key anatomical points such as the lateral epicondyle of femur, the medial epicondyle of femur, the tibial plateau and the inferior pole of patella, and the iterative closest point algorithm (ICP) is used to perform model precision optimization, generating a standardized model group. The model analysis module seamlessly imports the above-mentioned model group into the C3D clinical window software, realizes the real-time visualization of the three-dimensional movement trajectory of the knee joint, and calculates the joint torque and contact force parameters based on the inverse dynamics model; the gait cycle movement and biomechanical characteristics are identified by the support vector machine classifier, and the osteoarthritis related risk score is output, so as to realize the clinical pathology correlation evaluation. The man-machine interaction control module presets the osteoarthritis grading protocol template according to the clinical diagnosis requirements, and realizes the parameter input, analysis process control, real-time result monitoring and automatic generation of diagnosis report through the graphical user interface. The doctor can select the diagnosis template in the interface and input the patient's basic information, and automatically output the diagnosis report containing the movement curve, mechanical parameter chart and pathological risk result. The cloud publishing module converts the diagnosis report and the interactive interface into PHP dynamic web page format, and deploys it on the cloud server through HTTP protocol, supports SSL encryption transmission, multi-user concurrent access and cross-platform remote calling, so as to realize the remote sharing and joint diagnosis between patients and multi-center medical institutions.

[0026] Through the above technical scheme, the present application uses Magic software combined with extended Kalman filter and coordinate systematization method to reduce noise interference, and realizes high-precision standardized model group construction through three-dimensional model simplification and key anatomical point marking, thereby improving the accuracy of movement and biomechanical analysis. By seamlessly importing the standardized model into the C3D clinical window, the real-time visualization of the three-dimensional movement trajectory of the knee joint is realized, and combined with the inverse dynamics calculation and pathological correlation evaluation model, the quantitative biomechanical parameters and disease risk score can be quickly output. The man-machine interaction control module supports customized diagnosis logic and protocol template, provides an intuitive graphical user interface, and can realize parameter input, process control, real-time monitoring and report automatic generation, reducing the doctor's manual operation and improving the clinical work efficiency. Through the cloud publishing module, the diagnosis report and the interactive interface are converted into PHP dynamic web page and deployed to the cloud, supporting SSL encryption, multi-user concurrent access and cross-platform remote calling, realizing the data sharing between patients and multi-center hospitals.

[0027] The application further proposes that when the data processing module performs noise filtering and coordinate systemization on the original movement data, the following are included: The noise filtering uses an extended Kalman filter to define a state vector containing joint angles, angular velocities, and angular accelerations, adaptively adjusts process noise covariance parameters according to a kinematics model, and increases the tolerance range when the joint acceleration changes exceed an acceleration change threshold; in the measurement update stage, optical and inertial sensor data streams are fused, and a residual threshold monitoring process is performed to eliminate abnormal measurement values; The coordinate systemization initializes the calibration process based on the preoperative bone CT scan model through a quaternion rotation matrix algorithm, extracts the geometric center points of the distal femoral intercondylar fossa and the tibial platform to construct a local coordinate system, uses quaternions to represent rotation operations, and solves the optimal rotation parameters by minimizing the point cloud registration error to convert the data to a unified anatomical coordinate system.

[0028] Specifically, the data processing module performs noise filtering and coordinate systemization processing on the original data captured by the knee joint three-dimensional movement to ensure the accuracy and reliability of subsequent analysis. The noise filtering uses an extended Kalman filter (EKF) to model and predict the state vector containing joint angles, angular velocities, and angular accelerations. Through the kinematics model, the process noise covariance parameters are adaptively adjusted to cope with data fluctuations under different motion states. When the amount of change in joint acceleration exceeds the preset threshold, the process noise tolerance range is automatically increased to avoid the influence of abnormal deviations caused by sudden actions or sensor jitter on the filtering results. The preset rules of the threshold can be determined based on clinical statistical data, for example: the range of normal acceleration change in the gait cycle is 0-15 rad / s², and when the instantaneous acceleration change exceeds 15 rad / s², the tolerance increase process is triggered; for jumping or rapid turning actions, the threshold can be dynamically adjusted to 30 rad / s² to ensure that the filter can still output stably under intense action. In the measurement update stage, optical capture data and inertial measurement unit (IMU) data streams are fused, and an abnormal measurement value is eliminated through a residual threshold monitoring mechanism. The residual threshold can be set to mark an abnormal value when the root mean square error (RMSE) of the difference between the state prediction and the actual measurement value exceeds 0.5°, and the data point is removed or corrected through interpolation of adjacent time steps. Coordinate systemization is achieved through a quaternion rotation matrix algorithm. A precise three-dimensional bone model is constructed using preoperative bone CT scan data, and a local coordinate system is initialized at the geometric center points of the distal femoral intercondylar fossa and the tibial platform. The original three-dimensional movement data is rotated through a quaternion representation operation to calculate the optimal rotation parameters from the sensor coordinate system to the anatomical coordinate system. The optimal rotation parameters are obtained by minimizing the point cloud registration error to ensure that the joint movement trajectory is highly matched with the bone model. All original movement data is converted to a unified anatomical coordinate system.

[0029] As a preferred embodiment, the scheme of the application is implemented as follows: a patient's knee joint three-dimensional motion capture is performed, and gait data are synchronously collected by using an optical motion capture system and an inertial measurement unit (IMU). In the original collected data, due to the instantaneous jitter of the patient during the rapid starting and turning actions, the joint acceleration fluctuation reaches 20 rad / s 2, which is obviously higher than the change range of 0-15 rad / s 2 in the normal gait cycle. In the data processing module, the extended Kalman filter automatically detects the abnormal acceleration change, triggers the process noise covariance parameter increasing mechanism, so that the filter can still stably output the joint angle and angular velocity information under high dynamic action, while fusing the IMU data to correct the deviation caused by the loss or shielding of the optical data. The residual threshold is set to be greater than 0.5°, which is marked as abnormal, the abnormal data points at the corresponding time step are automatically removed, and the adjacent time step interpolation smoothing processing is performed. In the coordinate systematization link, the three-dimensional model of the femur and tibia established by the preoperative CT scan is used as the reference coordinate system. The geometric center points of the distal condyle intercondylar fossa of the femur and the tibial plateau are automatically extracted to construct the local anatomical coordinate system. The original captured data is rotated to the unified anatomical coordinate system by using the quaternion rotation matrix algorithm, the optimal rotation parameter is determined by minimizing the point cloud registration error, and the high matching of the patient's knee joint motion trajectory and the bone model is ensured.

[0030] Through the above technical scheme, the extended Kalman filter is used to model the joint angle, angular velocity and acceleration state vector, and the process noise covariance parameter is adaptively adjusted in combination with the motion model, so that the influence of sensor noise and external interference on the motion data is reduced. In the measurement update stage, the optical capture and inertial measurement unit (IMU) data are fused, the abnormal measurement values are removed through the residual threshold monitoring mechanism, the data continuity and authenticity are ensured, and the accuracy of subsequent analysis is improved. Through the local coordinate system initialization based on the preoperative bone CT scan, the quaternion rotation matrix algorithm and the determination of the optimal rotation parameter by minimizing the point cloud registration error, the original movement data is converted to the unified anatomical coordinate system, and the data of different collection environments and different patients have high comparability and consistency.

[0031] The application further proposes that when the data processing module simplifies the knee joint three-dimensional model, marks the key anatomical points and optimizes the precision, it includes: The knee joint three-dimensional model simplification divides the triangular mesh into a spatial octree structure by using the vertex clustering polygon reduction algorithm, analyzes the consistency of the vertex normal vector in the region node, merges the vertices when the curvature change is lower than the curvature change threshold, and retains the topological details of the intercondylar fossa key anatomical area; the key anatomical point marking is realized by using the pre-trained U-Net convolutional neural network.

[0032] Specifically, when simplifying the three-dimensional model of the knee joint, marking key anatomical points, and optimizing precision, the data processing module divides the original triangular mesh into a spatial octree structure through a vertex clustering polygon reduction algorithm, analyzes the consistency of the vertex normal vector in each regional node to determine the local curvature variation of the mesh surface. When the curvature variation is lower than the preset curvature variation threshold (for example, set to 0.02 rad / unit length), the vertices in the region are automatically merged to achieve model simplification, while the topological details in the intercondylar fossa and other key anatomical areas are preserved to ensure the precision of the key structures of the model. In the key anatomical point marking process, a pre-trained U-Net convolutional neural network is used to process the simplified three-dimensional model, inputting three-dimensional voxel grid data, and the network outputs a probability distribution heat map of key anatomical points such as the infrapatellar pole and the anterior horn of the meniscus. The final marking position is determined by combining the probability threshold (for example, 0.7), thereby achieving high-precision and automated key point recognition.

[0033] As a preferred embodiment, the scheme of the present application is implemented as follows: In a clinical gait analysis of the knee joint, a high-resolution three-dimensional mesh model of the patient's right knee joint is obtained, containing about 1.2 million vertices of the original triangular mesh. To facilitate subsequent real-time analysis and storage, the data processing module first divides the triangular mesh into a spatial octree structure, counts the vertex normal vector in each node, and calculates the local curvature variation. When the curvature variation is lower than the preset threshold of 0.02 rad / unit length, the vertices in the region are clustered and merged, thereby reducing the number of model vertices to about 300,000, while preserving the original topological details in the intercondylar fossa and meniscus key anatomical areas to ensure structural accuracy. The simplified three-dimensional model is input into a pre-trained U-Net convolutional neural network, which outputs a probability distribution heat map of the infrapatellar pole and the anterior horn of the meniscus, and sets the probability threshold to 0.7 to determine the final key point position. Through this process, the infrapatellar pole, the anterior horn of the medial and lateral meniscus, and the intercondylar fossa of the femur are successfully and automatically marked as key anatomical points, providing high-precision basic data for subsequent three-dimensional motion trajectory analysis, joint mechanics calculation, and clinical pathological evaluation.

[0034] Through the above technical scheme, the present application uses a vertex clustering polygon reduction algorithm combined with a spatial octree structure to reasonably merge the mesh vertices, significantly reducing the number of vertices of the knee joint three-dimensional model, reducing the storage and computing burden, and ensuring the integrity of the local topological structure. The original topological details are preserved in the intercondylar fossa and meniscus key anatomical areas to avoid losing important anatomical information during model simplification. The pre-trained U-Net convolutional neural network is used to mark the key anatomical points of the simplified three-dimensional model, achieving automated identification of the infrapatellar pole, the anterior horn of the meniscus, and the intercondylar fossa of the femur, reducing manual annotation errors, and improving analysis efficiency.

[0035] The application further proposes that when the data processing module implements key anatomical point labeling through the pre-trained U-Net convolutional neural network, the pre-trained U-Net convolutional neural network comprises the following steps: The pre-trained U-Net convolutional neural network comprises a down-sampling layer and an up-sampling layer, inputs three-dimensional voxel grid data, and outputs a probability distribution heat map of the patellar lower pole and the meniscus anterior horn key points; the precision optimization adopts an iterative closest point algorithm to perform model registration operation, sets a convergence condition as a root mean square error lower than an error threshold value or reaching a maximum iteration number, and combines a Laplace smoothing filter to eliminate surface noise interference.

[0036] Specifically, when labeling key anatomical points of a knee joint three-dimensional model, the three-dimensional model after simplification and preliminary optimization is converted into voxel grid data and input into the pre-trained U-Net convolutional neural network. The network is composed of symmetrical down-sampling layers and up-sampling layers, wherein the down-sampling layers gradually extract multi-scale features through convolution and pooling operations, and the up-sampling layers gradually restore spatial resolution through deconvolution, and output a probability distribution heat map of the patellar lower pole and the meniscus anterior horn key points. The value of each voxel in the probability heat map represents the possibility of the point being the target key point, and the threshold value is set to 0.7, that is, when the probability is higher than 0.7, the point is identified as a key anatomical point, so as to ensure the accuracy of labeling and avoid misjudgment of low confidence points. After the key points are preliminarily labeled, the iterative closest point (ICP) algorithm is further used for model precision optimization, the current model is registered with the reference bone model generated by the preoperative CT scan through point cloud registration, and the root mean square error (RMSE) is gradually minimized. The convergence condition of the ICP algorithm is set as: when the RMSE is lower than 0.5 mm or reaches the maximum iteration number 50 times, the iteration is stopped, so as to balance the calculation efficiency and the registration accuracy. In order to eliminate the noise and abnormal points that may exist on the surface of the model, the Laplace smoothing filter is also used to process the surface of the model, the local convexity and concavity are modified through the weighted average of the neighborhood vertices, and the surface continuity is improved.

[0037] As a preferred embodiment, the scheme of the application is implemented as follows: after a patient receives a knee joint three-dimensional motion capture scan, the generated three-dimensional model is input into the pre-trained U-Net convolutional neural network, and the network outputs a probability distribution heat map of the patellar lower pole and the meniscus anterior horn key points. In the heat map, the probability value of the patellar lower pole is 0.85, and the probability value of the meniscus anterior horn is 0.78, both of which are higher than the set threshold value 0.7, and these positions are labeled as key anatomical points. The three-dimensional model is registered with the standard bone model generated by the preoperative CT scan through the iterative closest point, and the ICP algorithm reaches the root mean square error 0.42 mm at the 35th iteration, which is lower than the error threshold value 0.5 mm, and the registration is stopped. The Laplace smoothing filter is used to eliminate the local convexity and concavity on the surface of the model due to scanning noise, so that the surface of the model is continuous and smooth, and a three-dimensional knee joint model with high precision and applicable to biomechanical analysis and clinical diagnosis is generated.

[0038] Through the technical solution, the pre-trained U-Net convolutional neural network can automatically identify the positions of the lower patellar pole and the anterior horn of the meniscus, and generate a probability distribution heat map, thereby improving the accuracy and reliability of key point identification. In the precision optimization process, the iterative closest point algorithm sets the convergence condition that the root mean square error is lower than the preset threshold or the maximum number of iterations is reached, thereby ensuring accurate registration of the three-dimensional model and the standard bone model, and avoiding deviations caused by scanning noise or action errors. In combination with the Laplace smoothing filter, the model surface noise interference is further eliminated, and the three-dimensional knee joint model surface is continuous and smooth.

[0039] The application further provides that when the C3D clinical window provides real-time visualization of knee joint three-dimensional movement trajectory, biomechanical parameter calculation, and clinical pathology correlation evaluation functions, it includes: The biomechanical parameter calculation uses an inverse dynamics model combined with a musculoskeletal simulation toolbox to calculate joint torque and contact force in real time; the clinical pathology correlation evaluation is realized based on a support vector machine classifier, the feature vector integrates torque peak value, contact force time integral, and trajectory asymmetry index in the gait cycle, and a radial basis kernel function is used for pattern recognition; the classifier performs a training process on an osteoarthritis case database and outputs a quantitative risk score.

[0040] Specifically, in the C3D clinical window, the real-time visualization function of the knee joint three-dimensional movement trajectory can dynamically display the rotation angle, movement trajectory, and position change of each joint of the knee joint during the gait cycle of the patient, so that the clinician can intuitively observe the movement state of the knee joint. The biomechanical parameter calculation uses an inverse dynamics model combined with a musculoskeletal simulation toolbox to calculate joint torque and contact force in real time according to the collected kinematics data and ground reaction force, accurately reflecting the stress condition of the knee joint at different movement stages. The clinical pathology correlation evaluation is realized based on a support vector machine (SVM) classifier, the joint torque peak value, contact force time integral, and movement trajectory asymmetry in the gait cycle are integrated into a multi-dimensional feature vector, and a radial basis kernel function is used for pattern recognition. Through training of the model in the osteoarthritis case database, a quantitative risk score is output.

[0041] As a preferred embodiment, the scheme of the application is implemented as follows: for a 65-year-old patient with early osteoarthritis, the joint displacement and angular velocity data during walking are obtained through three-dimensional motion capture of the knee joint. The C3D clinical window renders the captured data into a three-dimensional motion trajectory in real time, and calculates the knee joint flexion torque peak value of 35 Nm and the internal-external turning contact force of 250 N through an inverse dynamics model combined with a musculoskeletal simulation toolbox. The torque peak value, contact force time integral and trajectory asymmetry index in the gait cycle are integrated into a feature vector, which is input into a pre-trained support vector machine classifier for analysis. The classifier uses a radial basis kernel function to identify the pattern, and generates a quantitative risk score of 0.78 (full score 1.0 indicates high risk) based on the osteoarthritis case database, indicating that the patient's knee joint load is high and there is a trend of early degeneration, thereby providing a diagnostic basis for doctors.

[0042] Through the above technical scheme, the C3D clinical window of the application can visualize the three-dimensional movement trajectory of the knee joint in real time, enabling the clinician to intuitively observe the joint dynamic changes of the patient during walking or movement. The biomechanical parameter calculation combined with the inverse dynamics model and the musculoskeletal simulation toolbox can accurately calculate the torque and contact force of the joint at different movement stages, providing a quantitative basis for knee joint load analysis. The clinical pathology correlation evaluation is based on a support vector machine classifier, which integrates the torque peak value, contact force time integral and trajectory asymmetry index in the gait cycle, uses a radial basis kernel function for pattern recognition, and outputs a quantitative risk score after training on the osteoarthritis case database, providing doctors with disease progression evaluation.

[0043] The application further proposes that when the biomechanical parameter calculation uses an inverse dynamics model combined with a musculoskeletal simulation toolbox to calculate joint torque and contact force in real time, it includes: Based on the motion data and ground reaction force input, the motion equation containing the mass matrix, Coriolis force term and gravity term is solved; the contact force calculation uses the Hertz contact theory model, inputs the meniscus curvature radius and material elastic modulus parameters, and generates a tibiofemoral joint pressure distribution visualization cloud chart.

[0044] Specifically, in the process of biomechanical parameter calculation, the displacement, velocity and acceleration information of the knee joint in walking or flexion and extension movement of the patient are collected through the kinematics data acquisition, and the ground reaction force data are also collected as external constraint input. Based on these inputs, the kinematic equation of the joint is solved by using the inverse dynamics model, which includes the mass matrix, the Coriolis force term and the gravity term, and the joint moment of each movement stage is accurately calculated to realize the quantitative analysis of the internal force of the knee joint. For the calculation of joint contact force, the Hertz contact theory model is used, the meniscus curvature radius and material elastic modulus parameters are input into the model, the pressure distribution on the femur and tibia contact surface is calculated, and the tibiofemoral joint pressure cloud map is generated through three-dimensional visualization.

[0045] As a preferred embodiment, the scheme of the application is implemented as follows: in the walking analysis of a 55-year-old patient with knee osteoarthritis, the three-dimensional kinematic data and ground reaction force of the knee joint are collected by a high-resolution motion capture system. The collected kinematics data are input into the inverse dynamics model, combined with the musculoskeletal simulation toolbox, to calculate the flexion and extension moment, the internal and external rotation moment and the rotation moment of the knee joint in the walking cycle in real time. The curvature radius of the meniscus is set to 12 mm, the elastic modulus is set to 5 MPa, the Hertz contact theory model is used to calculate the femur and tibia contact force distribution, and a three-dimensional pressure cloud map is generated. Through the cloud map, it can be directly displayed that the tibiofemoral joint pressure is concentrated in the lateral condyle area, prompting the patient that the lateral cartilage is under too much stress, assisting the doctor to develop a personalized rehabilitation training program or surgical intervention plan, so as to realize accurate clinical diagnosis and treatment decision.

[0046] Through the above technical scheme, the real-time biomechanical analysis based on the inverse dynamics model can accurately calculate the force moment and contact force distribution of the knee joint in different movement states, and realize the quantitative evaluation of joint load. By inputting the kinematics data and ground reaction force, solving the kinematic equation by combining the mass matrix, the Coriolis force term and the gravity term, the moment change curve of the joint in the flexion, internal and external rotation and rotation directions can be obtained; the Hertz contact theory model is used to calculate the femur and tibia contact force distribution, and a three-dimensional pressure cloud map is generated by combining the meniscus curvature and elastic modulus, which shows the joint stress concentration area.

[0047] The application further proposes a human-computer interaction control module to write customized interaction logic combined with the application requirements of clinical diagnosis; integrate clinical protocol templates, provide a graphical user interface to realize user parameter input, analysis process control, real-time result monitoring and automatic generation of diagnosis report, including: The human-computer interaction control module integrates a visual drag-and-drop process editor, supports defining analysis step sequences through a graphical node interface, and performs parameter validity verification operations on data flow between nodes by an event-driven mechanism. The clinical protocol template includes an osteoarthritis grading evaluation template, presets anatomical index thresholds, and loads patient historical data. The graphical user interface provides real-time 3D model interaction functions, realizes touch zooming operations, action trajectory frame-by-frame playback, and generates a diagnosis report.

[0048] Specifically, in the human-computer interaction control module, customized interaction logic is combined with clinical diagnosis application requirements to provide a comprehensive and flexible operation and analysis environment. The module integrates a visual drag-and-drop process editor, and users can define the order and execution conditions of analysis steps through a graphical node interface. The data flow between each node adopts an event-driven mechanism to perform real-time parameter validity verification to ensure that input data and intermediate calculation results meet preset standards. The clinical protocol template includes an osteoarthritis grading evaluation scheme, sets anatomical index thresholds for key anatomical points such as the distal femoral intercondylar fossa, tibial plateau, and patellar position, for example, the knee joint varus-valgus angle threshold is preset to ±5°, and an abnormal warning is triggered when the joint space width is less than 3 mm, and patient historical data can be loaded for longitudinal comparison. The graphical user interface provides real-time three-dimensional model interaction functions, supports touch zooming, rotating operations, and action trajectory frame-by-frame playback based on the WebGL rendering engine, and users can dynamically observe the joint action state during the entire gait cycle. At the same time, a diagnosis report is automatically generated, integrating key action parameters, abnormal indicators, and three-dimensional visual screenshots into a standardized document format.

[0049] As a preferred embodiment, the scheme of the present application is implemented as follows: when performing knee gait analysis on a male patient, the clinician starts the customized interaction logic process through the human-computer interaction control module, sets the analysis steps in the graphical node interface, including data import, joint angle calculation, three-dimensional trajectory generation, and abnormality detection, and each node performs validity verification on the input data through an event-driven mechanism, such as checking whether the joint varus-valgus angle is within the preset threshold of ±5°. The osteoarthritis grading evaluation template is called, and patient historical images and gait data are automatically loaded for comparative analysis of knee joint varus-valgus angle, patellofemoral joint space width, and joint torque peak value. The graphical user interface displays the three-dimensional knee joint model through WebGL rendering, and the doctor can touch and rotate, zoom, and frame-by-frame playback the patient's action trajectory during the entire gait cycle, while automatically generating a diagnosis report, marking abnormal indicators in red, and attaching three-dimensional visual screenshots and quantitative data, such as a patellofemoral joint space of 2.8 mm, which exceeds the threshold range, indicating an early risk of osteoarthritis.

[0050] Through the technical solution, the human-computer interaction control module integrates a visual drag-type flow editor, so that a clinician can flexibly define an analysis step sequence according to diagnosis needs of different patients, each node has an event-driven data verification mechanism, and whether input parameters are within a preset threshold range can be automatically checked, such as key indexes of a knee joint varus angle, a patellofemoral joint space, and a joint torque peak value, and a warning is issued when an exception occurs. Through an osteoarthritis grading evaluation template, patient historical images and gait data can be automatically loaded, combined with current three-dimensional motion capture results, index comparison and risk scoring are achieved. A graphical user interface is based on a WebGL rendering engine, supports touch rotation, scaling, and gait trajectory frame-by-frame playback, and real-time display of joint three-dimensional models and motion parameter changes, and a report including visual screenshots, quantitative data, and diagnosis conclusions is automatically generated.

[0051] The human-computer interaction control module further generates a diagnosis report based on the WebGL rendering engine, and includes the following steps. The interface integrates visual elements and embeds dynamic biomechanical parameter charts; the interaction logic performs user parameter input, analysis flow control, and real-time result monitoring operations.

[0052] Specifically, in the human-computer interaction control module, the diagnosis report generation process based on the WebGL rendering engine embeds three-dimensional knee joint models, gait trajectories, and biomechanical parameters such as joint torque and contact force into the interface through dynamic charts to achieve visual display. The user can input patient basic information and analysis parameters such as gait cycle, joint load threshold, and key anatomical point selection through the interface, perform real-time analysis process, update three-dimensional model motion trajectory and related biomechanical data, and automatically label and prompt abnormal parameters. The module integrates real-time monitoring functions, dynamically generates diagnosis conclusions, mechanical distribution cloud diagrams, and quantitative statistical tables during the analysis process, and finally automatically generates a comprehensive diagnosis report including interactive charts and three-dimensional model screenshots.

[0053] As a preferred embodiment, the scheme of the present application is implemented as follows: in a knee gait analysis, the doctor inputs the patient ID, gait acquisition data path, and analysis cycle parameters in the WebGL interface through the human-computer interaction control module. The three-dimensional knee joint model is automatically loaded, and the real-time joint torque and contact force distribution charts are embedded in the interface. The user can rotate and scale the three-dimensional model through drag operation, frame-by-frame playback the knee joint motion trajectory, and observe the joint torque peak value and tibial femoral contact force changes. When the torque or contact force exceeds the preset threshold, the interface automatically highlights the abnormal part. After the analysis is completed, the module integrates the motion trajectory, mechanical cloud diagram, and quantitative statistical table to generate an interactive diagnosis report, and the doctor directly views the three-dimensional model, dynamic chart, and key parameters in the report to provide intuitive basis for clinical decision-making.

[0054] By the technical solution, the application can realize real-time visualization and dynamic interaction of knee joint movement data and biomechanical parameters through the man-machine interaction control module based on the WebGL rendering engine to generate a diagnosis report, and users can intuitively observe three-dimensional model movement trajectory, joint torque and contact force distribution in the same interface, while real-time monitoring and analyzing process and parameter changes, thereby improving diagnosis efficiency and accuracy; and the diagnosis report can be automatically generated in a structured and visualized manner through embedding dynamic charts and interactive controls.

[0055] The application further proposes that when the cloud publishing module converts the diagnosis report and the interactive interface into a PHP dynamic web page format, the following are included: The cloud publishing module adopts a model-view-controller architecture of a Laravel framework to realize data processing logic; a controller responds to an HTTP request and calls a database model to extract biomechanical parameters; a view layer uses a template engine to dynamically render interface content, integrates an AJAX asynchronous technology to perform a data non-refresh loading process, and updates results through lightweight requests; interactive charts are generated using a visualization library, and support joint torque polar coordinate chart and pathological risk trend line display; and web page performance is optimized through resource compilation and compression technology to perform a page loading process.

[0056] Specifically, after the cloud publishing module converts the diagnosis report and the interactive interface into a PHP dynamic web page format, the platform can realize efficient separation of data processing logic and front-end display. Specifically, the model-view-controller architecture of the Laravel framework enables the controller to accurately extract knee joint biomechanical parameters and analysis results by calling the database model when receiving an HTTP request; the view layer dynamically renders interface content through the template engine, while embedding the AJAX asynchronous technology to realize data non-refresh loading, so that users can view the latest calculation results in real time without manual refreshing when browsing the web page; and the interactive charts generate joint torque polar coordinate charts and pathological risk trend lines using the visualization library, which can be dynamically updated according to the user-selected time point or analysis parameters.

[0057] As a preferred embodiment, the scheme of the application is implemented as follows: in the knee joint three-dimensional movement analysis platform, when a doctor selects gait analysis results of a patient in the remote diagnosis interface, the cloud publishing module immediately triggers the Laravel controller to respond to the HTTP request, and calls the database model to extract the knee joint torque data, contact force distribution and pathological risk score of the patient. The view layer dynamically generates a web page through the template engine, and embeds the data into the interactive charts, the joint torque is displayed as a polar coordinate chart to show peak value changes in different gait cycles, and the pathological risk is displayed as a trend line with time changes. The AJAX asynchronous technology enables the doctor to view the updated results in real time without refreshing the page when switching different analysis time periods or parameter settings, while the web page is optimized for fast loading and smooth operation through resource compression.

[0058] Through the technical solution, the cloud publishing module realizes dynamic webpage generation of the diagnosis report and the interactive interface by adopting the model view controller architecture of the Laravel framework. The controller responds to the HTTP request and calls the database to extract the biomechanical parameters. The view layer renders the content through the template engine, and the AJAX is integrated to realize the non-refresh data update. The interactive chart supports the joint torque polar coordinate graph and the pathological risk trend line display. The webpage performance is optimized through resource compression, realizing fast loading and smooth operation.

[0059] The cloud publishing module is further proposed to be deployed and published on the cloud server through the HTTP protocol, support SSL encrypted transmission, multi-user concurrent access and cross-platform remote calling, including: In the HTTP protocol deployment, the SSL encrypted transmission adopts the TLS protocol and a high-strength key to realize end-to-end data protection. The server sends a digital certificate chain when the client is connected, and generates a secure session key through a key exchange mechanism. Multi-user concurrent access is realized through a load balancer to distribute traffic. A weighted round-robin algorithm is configured to dynamically adjust the connection strategy according to the server load, and a session retention mechanism is set. Cross-platform remote calling supports RESTful API interface design, defines data access endpoints, and transmits structured information in JSON data format.

[0060] Specifically, when the cloud publishing module is deployed and published on the cloud server through the HTTP protocol, supports SSL encrypted transmission, multi-user concurrent access and cross-platform remote calling, it can ensure data security, access stability and compatibility. The SSL encrypted transmission adopts the TLS protocol and a high-strength key to realize end-to-end protection. The client receives a digital certificate chain when connecting to the server, and generates an independent session key through a key exchange mechanism to prevent data from being intercepted or tampered with during transmission. Multi-user concurrent access is realized through a load balancer to distribute traffic. A weighted round-robin algorithm is configured to dynamically adjust the server request distribution strategy, and a session retention mechanism is configured to ensure that the session state is not lost when switching between different server nodes, improving response speed and stability. In terms of cross-platform remote calling, the cloud publishing module provides RESTful API interfaces, defines clear data access endpoints, and uses JSON format for structured information transmission, so that mobile terminals, desktop terminals and other third-party applications can easily call, realize remote access and integrated management of diagnosis reports and biomechanical data.

[0061] As a preferred embodiment, the scheme of the application is implemented as follows: a medical institution uploads the diagnosis report and interactive interface of the knee joint three-dimensional movement analysis platform to the cloud server, and doctors and rehabilitation therapists can access patient data through the web or mobile terminal. SSL encrypted transmission uses TLS1.3 protocol and 256-bit high-strength key to realize end-to-end data protection. When the client initiates a connection request, the server returns a digital certificate chain, and the client verifies the certificate and generates an independent session key with the server through key exchange to ensure that all sensitive information (such as patient personal information, joint torque and contact force data) during transmission is not intercepted or tampered with. For multi-user concurrent access, the cloud deployment module uses a load balancer to distribute traffic to multiple server nodes, uses a weighted round-robin algorithm to dynamically allocate requests according to the current load of each node, and sets a session retention mechanism to ensure that users do not lose session state when switching between different nodes. For example, a doctor can continuously view different angles and action trajectories while accessing the patient's three-dimensional model without interruption. In terms of cross-platform remote calling, the platform provides RESTful API interfaces to transmit structured data in JSON format, allowing the hospital's electronic medical record system, mobile rehabilitation application or third-party analysis tool to call the interface to obtain real-time biomechanical parameters and diagnosis reports, realizing data sharing and remote analysis. For example, a rehabilitation therapist can call the interface on a tablet computer to obtain the patient's knee joint load distribution diagram to directly guide the rehabilitation training program.

[0062] Through the above technical scheme, the cloud publishing module of the application is deployed on the cloud server through the HTTP protocol, supports SSL / TLS encryption to realize end-to-end data protection, and the client establishes a secure session through digital certificates and key exchange; multi-user concurrent access distributes requests through a load balancer and a weighted round-robin algorithm, and maintains session state; cross-platform remote calling provides RESTful API interfaces to transmit structured data in JSON format, realizing efficient and secure data access and sharing.

[0063] In summary, the knee joint three-dimensional motion data was collected by Magic software, and combined with the extended Kalman filter and quaternion rotation matrix algorithm for noise filtering and coordinate system, the high-precision capture and standardized processing of multi-degree of freedom knee joint movement were realized, and the data reliability and analysis accuracy were improved. The vertex clustering polygon reduction algorithm was used to simplify the three-dimensional model, and the key anatomical points were automatically labeled by pre-training U-Net convolutional neural network, combined with the iterative closest point algorithm and Laplace smoothing filter to realize the precision optimization, so that the knee joint three-dimensional model can be used for high-precision biomechanical analysis. C3D clinical window supports real-time visualization of knee joint three-dimensional movement trajectory, inverse dynamics calculation of joint torque and contact force, and clinical pathology correlation evaluation based on support vector machine, realizes gait cycle analysis, mechanical parameter peak value statistics and risk score output. The human-computer interaction control module realizes interactive parameter input, analysis process control, three-dimensional model real-time operation and diagnostic report automatic generation through graphical drag-and-drop process editor and WebGL rendering engine, which improves the convenience of clinical use. The cloud publishing module adopts Laravel MVC architecture, converts the diagnostic report and interactive interface into PHP dynamic web page format, deploys through HTTP protocol, supports SSL encryption, load balancing, multi-user concurrent access and cross-platform RESTful API calling, realizes data security sharing and remote collaboration.

[0064] Those skilled in the art will appreciate that embodiments of the application can be provided as methods, systems or computer program products. Accordingly, the application can be embodied in the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the application can be embodied in the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk memory, CD-ROMs, optical storage media, etc.) having computer usable program code embodied therein.

[0065] The application is described with reference to the flowcharts and / or block diagrams according to the embodiments of the application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a machine that implements the flowcharts and / or block diagrams. Figure 1 The device that implements the function specified in one flow or multiple flows and / or blocks. Figure 1 The device that implements the function specified in one flow or multiple flows and / or blocks.

[0066] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks or multiple blocks.

[0067] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks or multiple blocks.

[0068] Finally, it should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, rather than limiting the same. Even though the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced equivalently, and any modification or replacement without departing from the spirit and scope of the present application should be covered within the protection scope of the claims of the present application.

Claims

1. A three-dimensional movement analysis platform for knee joints used in clinical auxiliary diagnosis, characterized in that, include: Data processing module, model analysis module, human-computer interaction control module, and cloud publishing module; The data processing module is used to capture raw motion data of the knee joint in three dimensions using Magic software, and to perform coordinate correction and model processing operations. The coordinate correction includes noise filtering and coordinate systematization of the raw motion data. The model processing operations include simplification of the knee joint three-dimensional model, marking of key anatomical points, and accuracy optimization to generate a standardized model group. The model analysis module is used to seamlessly import the model group into the C3D Clinical Window software. The C3D Clinical Window provides real-time visualization of the three-dimensional movement trajectory of the knee joint, calculation of biomechanical parameters, and clinical pathology correlation assessment functions. The human-computer interaction control module is used to write customized interaction logic based on the needs of clinical diagnostic applications; it integrates clinical protocol templates and provides a graphical user interface to realize user parameter input, analysis process control, real-time result monitoring, and automatic generation of diagnostic reports. The cloud publishing module is used to convert diagnostic reports and interactive interfaces into PHP dynamic web page format, and deploy and publish them on the cloud server via HTTP protocol, supporting SSL encrypted transmission, multi-user concurrent access, and cross-platform remote calls.

2. The knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis according to claim 1, characterized in that, When the data processing module performs noise filtering and coordinate systematization on the raw motion data, it includes: The noise filtering uses an extended Kalman filter to define a state vector containing joint angle, angular velocity, and angular acceleration. The process noise covariance parameter is adaptively adjusted according to the kinematic model, and the tolerance range is increased when the joint acceleration change exceeds the acceleration change threshold. In the measurement update stage, the data streams of optical and inertial sensors are fused, and a residual threshold monitoring process is executed to remove abnormal measurement values. The coordinate system is initialized and calibrated based on the preoperative skeletal CT scan model using a quaternion rotation matrix algorithm. The geometric center points of the distal femoral intercondylar fossa and the tibial plateau are extracted to construct a local coordinate system. The rotation operation is represented by quaternions, and the optimal rotation parameters are solved by minimizing the point cloud registration error, thus transforming the data into a unified anatomical coordinate system.

3. The knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis according to claim 2, characterized in that, When the data processing module simplifies the 3D model of the knee joint, marks key anatomical points, and optimizes accuracy, it includes: The 3D model of the knee joint is simplified by using a vertex clustering polygon reduction algorithm to divide the triangular mesh into a spatial octree structure. The consistency of vertex normal vectors is analyzed within the region nodes. Vertices are merged when the curvature change is lower than the curvature change threshold, preserving the topological details of the key anatomical region of the intercondylar fossa. Key anatomical points are marked by a pre-trained U-Net convolutional neural network.

4. The knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis according to claim 3, characterized in that, When the data processing module implements key anatomical point labeling using a pre-trained U-Net convolutional neural network, it includes: The pre-trained U-Net convolutional neural network includes downsampling and upsampling layers. It takes three-dimensional voxel grid data as input and outputs a probability distribution heatmap of key points at the inferior pole of the patella and the anterior horn of the meniscus. Accuracy optimization uses the iterative nearest point algorithm for model registration. The convergence condition is set as the root mean square error being lower than the error threshold or reaching the maximum number of iterations. It is combined with Laplacian smoothing filtering to eliminate surface noise interference.

5. The knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis according to claim 4, characterized in that, When the C3D clinical window provides real-time visualization of the three-dimensional movement trajectory of the knee joint, calculation of biomechanical parameters, and assessment of clinical pathology relevance, it includes: The biomechanical parameters are calculated in real time using an inverse dynamics model combined with a musculoskeletal simulation toolkit to calculate joint torque and contact force. The clinical pathology correlation assessment is based on a support vector machine classifier. The feature vector integrates the peak torque, contact force time integral, and trajectory asymmetry index in the gait cycle, and uses a radial basis function kernel function for pattern recognition. The classifier is trained on an osteoarthritis case database and outputs a quantitative risk score.

6. The knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis according to claim 5, characterized in that, The calculation of biomechanical parameters, using an inverse dynamics model combined with a musculoskeletal simulation toolbox to calculate joint torques and contact forces in real time, includes: Based on kinematic data and ground reaction force input, the equations of motion, including mass matrix, Coriolis force term and gravity term, are solved; the contact force calculation adopts Hertz contact theory model, inputting meniscus curvature radius and material elastic modulus parameters, and generating a visualization cloud map of tibiofemoral joint pressure distribution.

7. The three-dimensional knee joint movement analysis platform for clinical auxiliary diagnosis according to claim 6, characterized in that, The human-computer interaction control module incorporates customized interaction logic based on clinical diagnostic application needs; it integrates clinical protocol templates and provides a graphical user interface to enable user parameter input, analysis process control, real-time result monitoring, and automatic generation of diagnostic reports, including: The human-computer interaction control module integrates customized interaction logic with a visual drag-and-drop process editor, supporting the definition of analysis step sequences through a graphical node interface; data flow between nodes is driven by an event-driven mechanism to perform parameter validity verification operations; the clinical protocol template includes an osteoarthritis grading assessment template, preset anatomical index thresholds, and loads patient historical data; the graphical user interface provides real-time 3D model interaction functions, and implements touch zoom operations, frame-by-frame playback of motion trajectories, and diagnostic report generation processes based on the WebGL rendering engine.

8. The knee joint three-dimensional movement analysis platform for clinical auxiliary diagnosis according to claim 7, characterized in that, When the human-computer interaction control module performs the diagnostic report generation process based on the WebGL rendering engine, it includes: It integrates visual elements into the interface and embeds dynamic biomechanical parameter charts; the interactive logic executes user parameter input, analysis process control, and real-time result monitoring operations.

9. The three-dimensional movement analysis platform for knee joint used in clinical auxiliary diagnosis according to claim 8, characterized in that, When the cloud-based publishing module converts the diagnostic report and interactive interface into a PHP dynamic webpage format, it includes: The cloud-based publishing module uses a model-view-controller architecture based on the Laravel framework to implement data processing logic. The controller responds to HTTP requests and calls the database model to extract biomechanical parameters. The view layer uses a template engine to dynamically render interface content, integrates AJAX asynchronous technology to execute a data loading process without page refresh, and updates results through lightweight requests. Interactive charts are generated using a visualization library configuration, supporting the display of joint torque polar coordinate graphs and pathological risk trend lines. Webpage performance is optimized through resource compilation and compression technology to execute the page loading process.

10. The three-dimensional knee joint movement analysis platform for clinical auxiliary diagnosis according to claim 9, characterized in that, The cloud-based publishing module is deployed and published on a cloud server via the HTTP protocol, supporting SSL encrypted transmission, concurrent access by multiple users, and cross-platform remote calls, including: In the HTTP protocol deployment, the SSL encrypted transmission uses the TLS protocol and a high-strength key to achieve end-to-end data protection; when the client connects, the server sends a digital certificate chain and generates a secure session key through a key exchange mechanism; the multi-user concurrent access is implemented through a load balancer to distribute traffic, and a weighted round-robin algorithm is configured to dynamically adjust the connection strategy according to the server load, and a session persistence mechanism is set; the cross-platform remote call supports RESTful API interface design, defines data access endpoints, and uses JSON data format to transmit structured information.

Images