Cross-scale force line safety boundary determination method and system for knee joint replacement and readable storage medium

Abstract

The invention provides a cross-scale force line safety boundary determination method and system for knee joint replacement and a readable storage medium, and the method comprises the steps: S1, obtaining the physical property and service data of a knee joint prosthesis, and carrying out the treatment through collecting the individual feature data and multi-modal clinical data of a patient, so as to form an available data set; s2, performing simulation analysis on different alignment angles of the knee joint prosthesis according to the available data set, and simulating mechanical response under various alignment error combinations to determine stress and pressure distribution; s3, acquiring multi-dimensional experimental data of the knee joint prosthesis, and evaluating performance through material fatigue and wear experiments; s4, according to simulation analysis and experimental data, knee joint prosthesis service evaluation based on cross-scale correlation is carried out, the safety force line range is determined, and the service performance is predicted. According to the method, the problems of inaccurate service performance prediction and inconformity with postoperative experience of a patient caused by scale splitting, material dispersity and disjunction of a simulation model and clinical reality are solved.

Description

Technical Field

[0001] This invention relates to the multidisciplinary field of medical devices, biomechanics, materials science and artificial intelligence, and in particular to a method, system and readable storage medium for determining cross-scale force line safety boundaries for knee replacement. Background Technology

[0002] The knee joint mainly comprises ligaments, articular surfaces, muscles, tendons, and other structures, making it one of the most complex joints in the human body, most prone to degeneration and injury. With the rapid development of medical technology, knee joint diseases, especially osteoarthritis, have become a significant factor affecting the quality of life of a large population worldwide. As a common type of bone and joint disease in clinical practice, the incidence of knee osteoarthritis (OA) is showing a year-on-year increasing trend, causing not only great inconvenience to patients' lives but also impacting their quality of life.

[0003] Total knee arthroplasty, as an effective treatment for end-stage osteoarthritis, relies heavily on the design and performance of the knee prosthesis for its success rate and patient satisfaction. Against the backdrop of a global aging population, the demand for knee prostheses is constantly increasing. By relieving pain and improving mobility, they significantly improve patients' quality of life, helping patients, especially the elderly, maintain their ability to live independently, and thus play a vital role in modern medicine.

[0004] Currently, the design and service evaluation of total knee arthroplasty (TKA) mainly rely on the following existing technologies: (1) Traditional methods based on experience-based design and standardized testing: Its core principle relies on the surgeon's clinical experience and the engineer's mechanical design experience for prosthesis design. The evaluation of prosthesis performance depends on standardized mechanical testing and limited clinical follow-up data. This method treats the knee prosthesis as a macroscopic mechanical structure. The design focuses on the prosthesis's geometry and material composition. During testing, major in-body loads are simulated in the laboratory to evaluate its macroscopic mechanical properties such as wear, stiffness, and stability.

[0005] (2) Multi-scale simulation technology based on computer simulation: To gain a deeper understanding of the mechanical behavior of prostheses within the body, researchers have developed various computational models to predict the interaction forces, contact stresses, and strain distribution between the prosthesis and bone and soft tissue. This allows for optimization of the prosthesis shape during the design phase, avoiding stress concentration. All these knee joint models can be broadly categorized into three types: physical models, biomechanical models, and finite element models. A physical model is a virtual model that only simulates a part of the real human knee joint. In most experimental studies, physical models are used to determine the mechanical response characteristics of the knee joint. Biomechanical models, on the other hand, satisfy physical theorems such as equilibrium equations and consist of a set of mathematical relationships. They can be further divided into phenomenological models and anatomical models. Phenomenological models do not focus on the local anatomical structure of the knee joint, while anatomical models consider the anatomical structures that form the various parts of the knee joint. Although biomechanical models of the knee joint can calculate the contact force between the femur and tibia, as well as the tension of ligaments, they assume rigidity of the skeleton and cannot calculate the stress distribution of the knee joint. Therefore, they remain idealized models with limitations. The finite element model is a three-dimensional model based on the finite element method. It allows for the assignment of material properties to each component of the model, the application of boundary conditions, the setting of working conditions, and the performance of related analyses.

[0006] (3) Multimodal data fusion and data-driven methods: In determining the safety boundary of the force line in knee replacement, multimodal data fusion can integrate multidimensional information from different dimensions, including material numerical data, clinical text data, gait data, and images. This fusion helps to more accurately assess the degree of postoperative artificial joint damage, soft tissue condition, and bone geometry, thereby optimizing postoperative planning and ensuring precise alignment of the prosthesis.

[0007] Currently, the mainstream methods for multimodal synthesis of medical images and clinical text are based on deep learning, with representative networks being convolutional neural networks (CNNs) and generative adversarial networks (GANs). CNNs, due to their powerful feature extraction capabilities, are widely used in medical image synthesis. Their principle is to discover complex, nonlinear statistical correlations within massive datasets. For example, by training a neural network to take preoperative image data as input, it can output possible postoperative force alignment results. Its structure relies on the neural network architecture rather than physical laws.

[0008] The existing technology has the following problems and drawbacks: (1) Safety boundary issues of knee replacement (TKA) prosthesis alignment based on functional alignment (FA) concept: Traditional knee prosthesis design and evaluation largely rely on experience-based design, standard biomechanical testing, and limited clinical follow-up data. This approach struggles to fully capture the prosthesis's true performance in complex physiological environments, particularly regarding individual differences, long-term wear and tear, and osseointegration. Furthermore, traditional knee prostheses are not sufficiently precise in mimicking human movement. For example, they often focus only on achieving flexion between the femur and tibia during leg flexion, neglecting the external rotation of the femur relative to the tibia. This makes accurate leg flexion difficult, and may even damage the prosthesis or the patient's femur and tibia due to the inconsistency with human physiology, thus worsening the condition.

[0009] (2) Service evaluation of knee prostheses based on cross-scale correlation: Traditional performance evaluation of artificial joints suffers from a contradiction between postoperative patient experience and clinical experience. Existing assessments rely heavily on limited clinical data and standardized experiments, making it difficult to accurately reflect prosthesis behavior under complex physiological conditions. Mechanical properties and data obtained in the laboratory under idealized conditions often fail to accurately correspond to the degree of functional improvement and satisfaction reported by patients in clinical settings. Furthermore, the lack of a cross-scale (materials-structure) systematic evaluation method, the disconnect between simulation and experimental data, the failure to achieve accurate individualized predictions, and the insufficient predictive power and generalization performance of the models all contribute to the problem.

[0010] Therefore, it is necessary to study a method, system, and readable storage medium for determining cross-scale force line safety boundaries for knee replacement in order to address the shortcomings of existing technologies and solve or mitigate one or more of the aforementioned problems. Summary of the Invention

[0011] In view of this, the present invention provides a method, system, and readable storage medium for determining the cross-scale force line safety boundary of knee replacement. Focusing on the alignment safety boundary problem of knee replacement (TKA) prostheses, it involves artificial knee joints and multimodal data fusion. It can accurately determine the force line safety boundary of knee replacement, solving the problems of incomplete evaluation, poor predictive generalization ability, and high R&D and clinical trial-and-error costs caused by traditional methods relying on experience, limited data, and single models. Simultaneously, it can achieve the correlation fusion and accurate evaluation of the cross-scale performance of artificial knee joint prostheses' "material-structure," solving the problems of inaccurate service performance prediction and discrepancies with patient postoperative experience caused by scale fragmentation, material dispersion, and the disconnect between simulation models and clinical reality.

[0012] On the one hand, a method for determining the cross-scale force line safety boundary for knee arthroplasty includes the following steps: S1: Obtain data on the physical properties and service life of knee prostheses, and manage them by collecting individual patient characteristic data and multimodal clinical data to form a usable dataset; S2: Based on the available dataset, perform simulation analysis of different alignment angles of the knee prosthesis, and use the finite element model to simulate the mechanical response under various combinations of alignment errors to determine the stress and pressure distribution; S3: Obtain multi-dimensional experimental data of knee joint prostheses and evaluate their performance by conducting material fatigue and wear tests as well as in vitro friction tests. S4: Based on simulation analysis and experimental data, conduct service evaluation of knee joint prostheses based on cross-scale correlation. Establish an evaluation model by integrating material scale and structural scale data through equivalent stress registration, and determine the range of safe force lines and predict service performance.

[0013] In addition to the aspects and any possible implementations described above, a further implementation is provided in which obtaining knee joint prosthesis physical properties and service data in S1 includes: S11: Collect patient age, gender, BMI, implantation time, diagnosis, preoperative scores, and alignment angle measurement results, and classify the imaging data to evaluate postoperative condition and wear. S12: Integrate gait loading parameters such as flexion-extension angle, adduction moment, flexion-extension moment, axial force, front-to-back displacement, and rotation angle, and extract spatial features of angle data and temporal features of displacement stress data through CNN algorithm; S13: A multi-layer cross-feature fusion method is used to aggregate features from different modalities into a feature vector, and the BiLSTM algorithm is used for time series modeling to capture temporal correlations; S14: Classify the time series modeling results through a fully connected layer to obtain damage evaluation categories, and divide the dataset into training and test sets to train and evaluate the model to determine performance advantages.

[0014] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the simulation analysis of different alignment angles of the knee prosthesis based on the available dataset in S2 includes: S21: Based on the three-dimensional model of the knee joint prosthesis, multiple finite element models were established for coronal alignment ranging from negative six degrees to positive six degrees, and the mechanical response under different alignment design combinations was studied; S22: Apply standard gait loads to each model to calculate the Mises stress and contact pressure of the liner, and base the results on the experimental results of the stress response butt joint material. S23: A life prediction model is proposed to determine the impact of different alignments on the life of the prosthesis, and the stress response of each unit level is updated in real time to obtain the maximum stress and maximum contact pressure under different alignment errors.

[0015] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the acquisition of multi-dimensional experimental data of the knee prosthesis in S3 includes: S31: Tensile fatigue tests are performed on metallic materials. A high-frequency fatigue testing machine is used to clamp the sample, apply cyclic loading, record the stress-strain curves, and evaluate the fatigue life. S32: For polyethylene materials, perform wear tests on components removed from samples in accordance with ISO standards, apply loads and record mass changes to evaluate wear performance; S33: An in vitro friction and wear test was conducted on a prosthesis implanted on a pseudobone under different loads and angles to analyze the wear under time and load conditions; Statistical methods and visualization tools are used to process experimental data to evaluate the performance characteristics of materials and structures, and optimization design recommendations are determined based on the evaluation results.

[0016] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the service evaluation of the knee prosthesis based on cross-scale correlation, performed in S4 according to simulation analysis and experimental data, includes: S41: Determine the service threshold of the prosthesis by screening the comfort force line data through multivariate statistical analysis of clinical data to optimize the parameter range as the judgment standard, and determine the material performance degradation threshold based on the maximum shear stress limit and von Mises stress. S42: The material SN curve and load-wear curve are equivalently registered with the stress analysis data of the finite element simulation unit of the spur to characterize the service performance of the corresponding stress unit. S43: Utilize biomechanical analysis and patient clinical data to optimize cross-scale fusion results, establish a cross-scale service evaluation model for knee prostheses, and integrate experimental data, clinical data, simulation data, and material scale data to make cross-scale information decisions to obtain service life or service performance and reliability information for each unit.

[0017] As described above and in any possible implementation, a further implementation is provided in which determining the prosthesis service threshold in S41 includes: We screened the comfort line data with high satisfaction rates from the preoperative and postoperative evaluation data collected in clinical practice, and performed multivariate statistical analysis to optimize the parameter range of the comfort line data. If the sum of stresses caused by various failure modes of materials exceeds the maximum shear stress limit of the tibial liner material, it is determined to be a performance degradation, and the maximum shear stress limit and von Mises stress are used as parameters to evaluate the material strength to determine the degradation threshold.

[0018] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the establishment of a cross-scale service evaluation model for knee joint prostheses in S43 includes: Based on the equivalent registration fusion of material-scale fatigue and wear test performance results and finite element simulation analysis of corresponding stress elements; The fusion results were optimized by combining biomechanical analysis with patient clinical data, and different force line alignment angle parameter groups were set according to the patient's individual basic information characteristics to input the model to obtain service performance evaluation time-series evolution data. By comparing the knee prosthesis alignment comfort line judgment criteria, a safe force line range for individual patients was determined, and the service life under different reliability levels was determined by comparing the material performance degradation threshold.

[0019] In addition to the aspects and any possible implementations described above, a further implementation is provided in which determining the safety force line range and predicting service performance in S4 includes: Different sets of force line alignment angle parameters were set according to the patient's individual basic information characteristics, and short-term service performance evolution data were obtained through a cross-scale service evaluation model; The range of safe force lines is determined by comparing with the criteria for judging the comfort force lines, and the long-term service performance time-series evolution data is obtained through the model; By comparing the material performance degradation threshold, the service life of the knee prosthesis under different reliability levels can be predicted, or the reliability results under different service life conditions can be obtained.

[0020] In addition to the aspects and any possible implementations described above, a further implementation is provided: a cross-scale force line safety boundary determination system for knee replacement, characterized in that the cross-scale force line safety boundary determination system for knee replacement includes: The dataset module is used to acquire data on the physical properties and service life of knee prostheses. It is used to collect individual patient characteristic data and multimodal clinical data to form a usable dataset. The pressure distribution module is used to perform simulation analysis of different alignment angles of the knee joint prosthesis based on the available dataset. It uses a finite element model to simulate the mechanical response under various combinations of alignment errors to determine the stress and pressure distribution. The performance evaluation module is used to acquire multi-dimensional experimental data of knee joint prostheses and to evaluate their performance by conducting material fatigue and wear tests as well as in vitro friction tests. The modeling and evaluation module is used to conduct service evaluation of knee joint prostheses based on simulation analysis and experimental data, and to establish an evaluation model by integrating material scale and structural scale data through equivalent stress registration, and to determine the range of safe force lines and predict service performance.

[0021] In addition to the aspects described above and any possible implementation thereof, a readable storage medium is further provided, the readable storage medium including a processor and a memory connected to the processor, the processor containing a program that executes the described method for determining cross-scale force line safety boundaries for knee replacement.

[0022] Compared with the prior art, the present invention can achieve the following technical effects: 1. This invention combines the advantages of mechanistic models and data-driven models, maximizing the use of clinical and simulation data, and integrating biomechanical models and a large amount of data on different loads and alignment stresses. In determining the safety boundary of the knee replacement force line, it improves the model's predictive ability and generalization performance while ensuring the model's physical rationality, thereby reducing the cost of experimental and clinical trial-and-error and greatly enhancing the scientific rigor and precision of knee replacement surgery. Using digital human models and intelligent evaluation systems, hospitals can conduct personalized preoperative simulations to accurately determine the surgical force line and reduce errors caused by subjective judgment. This data-driven preoperative planning not only improves the precision of surgical procedures but also reduces surgical time and radiation exposure, aligning with the principle of high efficiency and low damage pursued in modern medicine. Through the application of this scientific method, the standardization and normalization of clinical practice are significantly improved, providing patients with safer and more effective treatment options.

[0023] 2. This invention uses equivalent stress to register and correlate finite elements at the structural scale of the prosthesis with those at the material scale. It then uses fuzzy decision-making to link the relationships between performance parameters monitored at key structural locations with the corresponding performance parameter relationships at the structural scale simulation results, achieving cross-scale correlation and fusion of the structure. Furthermore, it utilizes biomechanical analysis and individual patient clinical data to optimize and form a service evaluation model. Essentially, this invention considers the dispersion and rationality of materials, using sample information to correct simulation results and improve the accuracy of service performance evaluation of artificial joint prostheses. In addition, with the help of digital postoperative follow-up assessment, doctors can adjust treatment plans in a timely manner, ensuring that each patient receives personalized and continuously optimized medical services, further consolidating the hospital's trust and reputation among patients.

[0024] Of course, any product implementing this invention does not necessarily need to achieve all of the technical effects described above at the same time. Attached Figure Description

[0025] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a step diagram of a method for determining cross-scale force line safety boundaries for knee replacement provided in an embodiment of the present invention; Figure 2 This is a flowchart of a method for determining cross-scale force line safety boundaries for knee replacement provided in an embodiment of the present invention. Detailed Implementation

[0027] To better understand the technical solution of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0028] It should be understood that the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0029] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0030] like Figure 1 As shown, this invention provides a method for determining the cross-scale force line safety boundary for knee replacement. This method achieves accurate evaluation of the service performance of knee prostheses and determination of the safe force line range through multi-dimensional data fusion and cross-scale correlation analysis.

[0031] Step S1 involves acquiring knee prosthesis physical properties and service data, and then processing this data by collecting individual patient characteristic data and multimodal clinical data to form a usable dataset. This step is the foundational data preparation stage of the entire method, involving the collection, cleaning, and integration of multi-source heterogeneous data.

[0032] Step S11 involves collecting patient information including age, gender, BMI, implantation time, diagnosis, preoperative scores, and alignment angle measurements. The imaging data is then categorized to evaluate postoperative condition and wear. In one embodiment, the collection of individual patient characteristic data covers key factors influencing the service performance of the knee prosthesis. Age data reflects the patient's physiological state and bone density trends; patients are typically divided into three age groups: youth, middle-aged, and elderly, corresponding to different bone metabolism levels and activity levels. Gender influences the patient's skeletal structure and muscle strength distribution; male patients generally have larger bone size and higher load-bearing capacity. BMI, as the ratio of weight to height, is directly related to the load on the knee joint; patients with a BMI over 30 are classified as obese, and their knee prostheses bear significantly higher stress than those of patients with normal weight.

[0033] Implantation time data recorded the actual service life of the prosthesis, providing a time baseline for subsequent wear assessment and lifespan prediction. Diagnoses included various etiologies such as primary osteoarthritis, rheumatoid arthritis, and traumatic arthritis. Different etiologies lead to variations in bone condition and soft tissue environment, affecting the prosthesis's fixation effectiveness and long-term stability. Preoperative scoring employed a knee joint function scoring system, including four dimensions: pain intensity, range of motion, stability, and functional status. Each dimension was quantified from 0 to 100 points.

[0034] Alignment angle measurements were obtained through X-ray imaging, including data in three directions: coronal, sagittal, and axial. The coronal angle reflects the prosthesis's alignment in the inversion / exversion direction; the normal range is -3 degrees to +3 degrees, and values ​​outside this range are considered malalignment. Image data classification employed deep learning to automatically analyze postoperative X-ray images, categorizing and labeling them according to prosthesis position, osseointegration, and degree of wear.

[0035] Step S12 involves fusing gait loading parameters such as flexion-extension angle, adduction moment, flexion-extension moment, axial force, anterior-posterior displacement, and rotation angle. A CNN algorithm is used to extract spatial features of the angle data and temporal features of the displacement-stress data. Gait analysis is a crucial method for assessing the functional status of knee prostheses. A three-dimensional gait analysis system is used to collect kinematic and dynamic parameters during the patient's walking process. The flexion-extension angle reflects the range of motion of the knee joint in the sagittal plane. In a normal gait cycle, the flexion-extension angle ranges from 0 to 60 degrees. After prosthesis implantation, the flexion-extension angle may be limited. The adduction moment represents the bending moment load borne by the knee joint in the coronal plane. Excessive adduction moment can lead to excessive stress on the medial compartment of the prosthesis.

[0036] CNN algorithms play a crucial role in gait data processing, extracting spatial features of gait parameters by constructing multi-layer convolutional neural networks. Convolutional layers use one-dimensional convolutional kernels to extract features from time-series data, pooling layers reduce the dimensionality of these features, and the ReLU activation function enhances the network's non-linear expressive power. Spatial feature extraction of angle data focuses on the amplitude, frequency, and phase characteristics of joint angle changes, reflecting the patient's gait pattern and joint functional state. Temporal feature extraction of displacement stress data focuses on the temporal patterns of load changes, including the time of load peak occurrence, the gradient of load changes, and the duration of load.

[0037] Step S13 involves aggregating features from different modalities into a feature vector using a multi-layer cross-feature fusion method, and then employing the BiLSTM algorithm for time series modeling to capture temporal correlations. Multi-layer cross-feature fusion is a core technology for processing multimodal data, achieving effective fusion of features from different modalities through a cross-attention mechanism. The first layer of fusion processes basic patient information and image features, generating a fused feature vector through feature alignment and weight allocation. The second layer integrates gait parameters and load data, using a gating mechanism to control the contribution of different features. The third layer integrates the fusion results from the first two layers, generating a comprehensive feature vector containing individual patient features, image features, and gait features.

[0038] The BiLSTM algorithm captures the forward and backward dependencies of time-series data through a bidirectional long short-term memory network. The forward LSTM processes information from the beginning of the time series to the current time step, while the backward LSTM processes information from the end of the time series to the current time step. The hidden states in both directions are fused through concatenation or addition. The BiLSTM's forget gate controls the degree of retention of historical information, the input gate determines the degree of acceptance of new information, and the output gate adjusts the amount of information output at the current time step. Capturing temporal dependencies is crucial for understanding the dynamic service process of knee joint prostheses, as wear and loosening of the prosthesis is a gradual time process.

[0039] Step S14: The time series modeling results are classified using a fully connected layer to obtain damage evaluation categories. The dataset is then divided into training and testing sets for model training and evaluation to determine performance advantages. The fully connected layer, as the last layer of the classification network, maps the feature vectors output by the BiLSTM to the damage evaluation category space. Damage evaluation categories are divided into five levels based on the prosthesis's service status: normal, mild wear, moderate wear, severe wear, and failure. A normal state indicates that the prosthesis is fully functional and all indicators are within normal ranges. Mild wear indicates slight surface wear on the prosthesis, but it does not affect overall function. Moderate wear indicates that the prosthesis wear has worsened and has begun to affect joint function. Severe wear indicates that the prosthesis is severely worn, and joint function is significantly limited. A failure state indicates that the prosthesis can no longer function properly and requires revision surgery.

[0040] The dataset was partitioned using stratified sampling to ensure a consistent proportion of samples at each impairment level across the training and test sets. The training set comprised 70% of the total data and was used for learning and optimizing model parameters. The validation set comprised 15% of the total data and was used for tuning model hyperparameters and implementing early stopping strategies. The test set comprised 15% of the total data and was used for the final evaluation of model performance. Model training employed the cross-entropy loss function, with the Adam algorithm chosen as the optimizer, a learning rate of 0.001, and a batch size of 32.

[0041] Step S2 involves performing simulation analysis of different alignment angles of the knee prosthesis based on the available dataset. A finite element model is used to simulate the mechanical response under various combinations of alignment errors to determine the stress and pressure distribution. This step uses numerical simulation to study the influence of the prosthesis alignment angle on mechanical properties, providing a theoretical basis for subsequent safety boundary determination.

[0042] Step S21 involves establishing multiple finite element models based on the three-dimensional model of the knee prosthesis, selecting alignment conditions ranging from -6 degrees to +6 degrees in the coronal plane, and studying the mechanical response under different alignment design combinations. Finite element modeling is the core step in simulation analysis; firstly, an accurate three-dimensional geometric model of the knee prosthesis needs to be constructed. The prosthesis model consists of three main parts: the femoral component, the tibial component, and the polyethylene liner. The femoral component is made of cobalt-chromium-molybdenum alloy, the tibial component is made of titanium alloy, and the polyethylene liner is made of ultra-high molecular weight polyethylene. The geometric model is constructed based on the CAD data of the actual prosthesis product to ensure the geometric accuracy and dimensional accuracy of the model.

[0043] The coronal alignment angles range from -6 degrees to +6 degrees, with 13 different alignment models established at one-degree intervals. Negative angles indicate inward displacement of the prosthesis, while positive angles indicate outward displacement. Each alignment angle corresponds to an independent finite element model, and the relative positions of the prosthesis components in the model are adjusted accordingly based on the alignment angle. Tetrahedral elements are used for mesh generation, with mesh refinement applied to the contact areas to ensure the accuracy of contact stress calculations. The femoral and tibial components use coarser meshes with element sizes of approximately 2 mm. The polyethylene liner uses a finer mesh with element sizes of approximately 0.5 mm, especially in the surface areas in contact with the metal components.

[0044] The definition of material properties is crucial to the accuracy of simulation results. The elastic modulus of the cobalt-chromium-molybdenum alloy was set to 210 GPa, Poisson's ratio to 0.3, and density to 8.3 g / cm³. 3 The elastic modulus of the titanium alloy was set to 110 GPa, Poisson's ratio to be 0.34, and density to be 4.5 g / cm³. 3 The elastic modulus of ultra-high molecular weight polyethylene was set to 1 GPa, Poisson's ratio to be 0.4, and density to be 0.94 g / cm³. 3 The contact is defined as a surface-to-surface contact, and the coefficient of friction is set according to the material pairing relationship. The coefficient of friction between metal and polyethylene is set to 0.1.

[0045] Step S22 involves applying standard gait loads to each model to calculate the Mises stress and contact pressure of the liner, based on stress-response material testing results. The application of standard gait loads simulates the load conditions experienced by the knee joint during normal walking. Load data are derived from gait analysis experiments, including axial pressure, anterior-posterior shear force, lateral-lateral shear force, and bending moment loads in all directions. Axial pressure exhibits a bimodal distribution during the gait cycle, with the first peak occurring during heel strike (approximately 2.5 times body weight) and the second peak occurring during toe-off (approximately 3 times body weight). Anterior-posterior shear force is initially negative and then positive during the gait cycle, with a maximum value approximately 0.5 times body weight.

[0046] The load was applied using a time-history loading method, dividing a complete gait cycle into 100 time steps, each corresponding to a specific load value. The load application point was located at the distal end of the femoral component, and the load was transferred to the prosthesis surface through a rigid connection unit. Boundary conditions included complete fixation constraints on the bottom surface of the tibial component, simulating a firm fixation between the tibia and the prosthesis. The motion constraints of the femoral component were set according to the physiological movement characteristics of the knee joint, allowing rotation in the flexion-extension direction and axial translation, while restricting movement in other directions.

[0047] Mises stress is calculated based on the von Mises yield criterion, which is applicable to the strength assessment of both metallic and polymeric materials. Mises stress converts a three-dimensional stress state into an equivalent one-dimensional stress value, facilitating comparison with the material's yield strength. Contact pressure is calculated through contact elements, reflecting the interaction forces between prosthetic components. The stress distribution and contact pressure distribution of the polyethylene liner, as the load-bearing structure, are key indicators for evaluating prosthetic performance.

[0048] Step S23 proposes a life prediction model to determine the impact of different alignments on the prosthesis's lifespan and updates the stress response at each unit level in real time to obtain the maximum stress and maximum contact pressure under different alignment errors. The life prediction model is based on the fatigue damage accumulation theory and uses Miner's linear cumulative damage rule to calculate the prosthesis's fatigue life. This model considers the material's SN curve characteristics and the influence of the actual load spectrum, converting complex variable amplitude loads into equivalent constant amplitude loads for life calculation. The fatigue life calculation formula is based on the relationship between stress amplitude and cycle number; different stress levels correspond to different fatigue lives.

[0049] Real-time updates of element-level stress response are achieved through iterative calculations using the finite element solver. The calculation results at each time step serve as the initial conditions for the next time step, forming a continuous stress evolution process. The stress response includes the principal stresses, shear stresses, and Mises stresses of each element, and these stress values ​​are dynamically updated as the load changes. The identification of the maximum stress is achieved by traversing the stress values ​​of all elements, recording the global maximum stress and its corresponding element location at each time step.

[0050] The stress distribution varies significantly under different alignment errors. With inversion alignment, the inner side of the prosthesis experiences greater stress, while the outer side experiences relatively less stress. With eversion alignment, the stress distribution is reversed, with increased stress on the outer side and decreased stress on the inner side. With neutral alignment, the stress distribution is relatively uniform, and the overall stress level is lowest. This difference in stress distribution directly affects the fatigue life and wear performance of the prosthesis, providing an important reference for alignment selection during clinical surgery.

[0051] Step S3 involves acquiring multi-dimensional experimental data of the knee joint prosthesis, and evaluating its performance through material fatigue and wear tests, as well as in vitro friction tests. Acquiring experimental data is crucial for validating simulation results and establishing cross-scale correlations. Multiple experimental methods are used to comprehensively evaluate the performance characteristics of the prosthesis materials and structure.

[0052] Step S31 involves conducting tensile fatigue tests on metallic materials. A high-frequency fatigue testing machine is used to clamp the sample, apply cyclic loading, and record the stress-strain curves to assess fatigue life. The fatigue performance of metallic materials is a key factor affecting the long-term service life of prostheses. Tensile fatigue testing assesses the fatigue characteristics of materials by simulating cyclic loading under actual service conditions. Standard fatigue test bars are used as test samples, with dimensions conforming to ASTM standards. Sample materials include two types: cobalt-chromium-molybdenum alloy and titanium alloy, corresponding to the femoral and tibial components, respectively.

[0053] The high-frequency fatigue testing machine operates at a frequency of 20Hz, ensuring testing efficiency while avoiding sample temperature rise issues caused by excessively high frequencies. The clamping system employs hydraulic clamps to ensure reliable sample fixation and accurate load transfer. Cyclic loading utilizes a sinusoidal waveform with a stress ratio of 0.1. The maximum stress level is determined based on simulation analysis results, covering low, medium, and high stress levels. At least five parallel tests are conducted for each stress level to ensure data reliability and reproducibility.

[0054] Stress-strain curves were recorded using high-precision sensors and a data acquisition system. The load cell had an accuracy of 0.1% of full scale, and the displacement cell had an accuracy of 0.001 mm. The data acquisition frequency was set to 1000 Hz to ensure the capture of dynamic changes in load and displacement. Fatigue life was assessed based on the number of cycles at specimen fracture, while simultaneously recording the initiation and propagation of fatigue cracks. The relationship between stress level and fatigue life was established by plotting SN curves, providing material parameters for subsequent life prediction.

[0055] Step S32 involves removing the polyethylene material from the sample and conducting an abrasion test conforming to ISO standards. Loads are applied and mass changes are recorded to assess abrasion performance. The abrasion performance of the polyethylene liner directly affects the prosthesis's lifespan and clinical outcomes. The abrasion test is conducted according to ISO 14243, simulating the physiological movements and load conditions of the knee joint. The test sample is cut from the actual polyethylene liner, maintaining its original material properties and surface condition. The sample is cylindrical, with a diameter of 25 mm and a height of 10 mm.

[0056] The wear testing machine adopts a pin-disc structure, with polyethylene samples serving as pins and metal mating parts as discs. The mating parts are made of cobalt-chromium-molybdenum alloy, with a surface roughness controlled within Ra 0.05 micrometers to simulate the surface condition of an actual prosthesis. The test load is set based on gait analysis data, using a variable load mode to simulate actual load changes. The load range is from 200N to 2000N, with a frequency of 1Hz, corresponding to the normal walking frequency.

[0057] The lubricating medium used was simulated synovial fluid, containing bioactive components such as proteins, hyaluronic acid, and phospholipids. The pH was controlled at 7.4, and the temperature was maintained at 37 degrees Celsius to simulate the physiological environment of the human joint cavity. The lubricating medium was replaced periodically during the experiment to prevent the accumulation of wear particles from affecting the results. Mass change was measured using a precision balance with an accuracy of 0.01 mg. The sample mass was measured at regular intervals to calculate the cumulative wear.

[0058] Wear performance evaluation includes three aspects: wear rate, wear mechanism, and wear particle characteristics. Wear rate is calculated as the mass loss per unit time or unit load, reflecting the material's wear resistance. Wear mechanism is determined by observing the morphology of the worn surface using scanning electron microscopy, including types such as adhesive wear, abrasive wear, and fatigue wear. The collection and analysis of wear particles are achieved by filtering the lubricating medium; the particle size distribution and morphology are important for assessing biocompatibility.

[0059] Step S33: An in vitro friction and wear test is conducted on the prosthesis under different loads and angles after implantation into the artificial bone. The wear under different time and load conditions is analyzed. The in vitro friction and wear test constructs a test environment closer to real-world applications to evaluate the friction and wear performance of the prosthesis under complex load conditions. The artificial bone is made of polyurethane foam material, with density and mechanical properties similar to human cancellous bone. The implantation of the prosthesis is performed according to standard surgical procedures to ensure good fixation between the prosthesis and the artificial bone.

[0060] The experimental setup includes a load application system, an angle control system, and a friction force measurement system. The load application system is driven by a servo motor, enabling precise control of the load magnitude and its variation. The angle control system uses a stepper motor to achieve multi-axis motion of the prosthesis, simulating flexion, extension, inversion / exversion, and internal / external rotation of the knee joint. The friction force measurement system employs multi-dimensional force sensors to monitor friction force changes in all directions in real time.

[0061] The different load conditions are set based on the patient's activity level, including three levels: daily activity load, exercise load, and ultimate load. Daily activity load corresponds to the load level of normal walking and climbing stairs, exercise load corresponds to the load level of running and jumping, and ultimate load corresponds to the load level of accidental impact. Different angle conditions include three situations: normal alignment, slight misalignment, and severe misalignment, each corresponding to a different prosthesis installation angle.

[0062] Time-domain analysis was achieved through long-term continuous testing, with a test cycle set at 5 million cycles, equivalent to 5 years of normal use by a patient. Changes in the friction coefficient, wear amount, and surface morphology were recorded periodically during the test to analyze the temporal evolution of wear. Load condition analysis established a quantitative relationship between load and wear by comparing wear results under different load levels.

[0063] Step S34 involves using statistical methods and visualization tools to process experimental data, evaluate the performance characteristics of materials and structures, and determine optimization design recommendations based on the evaluation results. Various statistical methods are employed in processing the experimental data, including descriptive statistics, correlation analysis, and regression analysis. Descriptive statistics calculate the mean, standard deviation, maximum, and minimum values ​​of each performance index to assess the distribution characteristics and dispersion of the data. Correlation analysis studies the relationships between different performance indices and identifies key factors affecting prosthesis performance.

[0064] Regression analysis establishes a mathematical model between performance indicators and influencing factors to predict performance under different conditions. Multiple linear regression models are used to analyze the combined impact of multiple factors on a single performance indicator, while nonlinear regression models are used to handle complex nonlinear relationships. Statistical significance testing ensures the reliability of the analysis results, with a confidence level set at 95%.

[0065] Visualization tools include various chart types such as scatter plots, bar charts, box plots, and heatmaps. Scatter plots are used to show the relationship between two variables, bar charts are used to compare performance differences under different conditions, box plots are used to show the distribution characteristics of data, and heatmaps are used to show the correlation patterns of multidimensional data. Three-dimensional visualization technology is used to display complex spatial distribution data, such as stress distribution and wear distribution.

[0066] Step S4 involves conducting a service evaluation of the knee prosthesis based on simulation analysis and experimental data, using cross-scale correlation. An evaluation model is established by fusing material-scale and structural-scale data through equivalent stress registration, and the safe force line range and predicted service performance are determined. This step is the core of the entire method, achieving a comprehensive evaluation of the knee prosthesis's service performance by establishing the correlation between material-scale experimental data and structural-scale simulation data.

[0067] Step S41 involves determining the prosthesis service threshold. This is achieved by using multivariate statistical analysis of clinical data to screen comfort line data and optimize parameter ranges as the criterion. The material performance degradation threshold is then determined based on the maximum shear stress limit and von Mises stress. Determining the service threshold is fundamental to establishing evaluation criteria and requires comprehensive consideration of both clinical efficacy and material performance requirements.

[0068] Step S411 involves screening preoperative and postoperative evaluation data collected clinically for comfort line data with high satisfaction levels, and then performing multivariate statistical analysis to optimize the parameter range of the comfort line data. The screening of clinical data is based on patient satisfaction surveys and functional assessments. The knee joint function scoring scale is used to quantitatively assess the patient's postoperative status. The scoring scale includes four dimensions: pain assessment, functional assessment, range of motion assessment, and stability assessment. Each dimension uses a scoring standard from 0 to 25 points, for a total score of 100 points. Patients with high satisfaction are defined as those with a total score exceeding 80 points and individual item scores exceeding 16 points.

[0069] Comfort alignment data were extracted by analyzing imaging data from patients with high satisfaction levels, measuring the prosthesis alignment parameters. Coronal alignment angles were obtained by measuring the angle between the femoral and tibial prosthesis axes, while sagittal alignment angles were obtained by measuring the prosthesis tilt angle on lateral radiographs. Axial alignment angles were obtained by measuring the prosthesis rotation angle through three-dimensional reconstruction of CT scan data. Data collection covered follow-up time points at 6 months, 1 year, 2 years, and 5 years postoperatively to ensure the long-term stability of the alignment data.

[0070] Multivariate statistical analysis employed principal component analysis to identify the main force line parameters influencing patient satisfaction. By calculating the correlation coefficients between each force line parameter and satisfaction scores, key force line indicators were determined. Cluster analysis grouped patients with high satisfaction levels according to force line characteristics, identifying different comfort force line patterns. Discriminant analysis established a classification model between force line parameters and satisfaction levels, determining the parameter boundaries of comfort force lines.

[0071] Parameter range optimization was achieved through confidence interval estimation, calculating the mean and standard deviation of each force line parameter in the patient population with high satisfaction. The parameter range of the comfort force line was defined as the interval between the mean and 1.96 times the standard deviation, corresponding to a 95% confidence level. The comfort range for the coronal force line angle was determined to be -2.5 degrees to +2.5 degrees, the comfort range for the sagittal force line angle was determined to be -3 degrees to +3 degrees, and the comfort range for the axial force line angle was determined to be -5 degrees to +5 degrees.

[0072] In step S412, if the sum of stresses caused by various material failure modes exceeds the maximum shear stress limit of the tibial liner material, it is determined to be a performance degradation, and the maximum shear stress limit and von Mises stress are used as parameters to evaluate material strength and determine the degradation threshold. The determination of the material performance degradation threshold is based on material mechanics theory and experimental data, comprehensively considering the influence of multiple failure modes.

[0073] The failure modes of tibial liner materials include four main types: tensile failure, compressive failure, shear failure, and fatigue failure. Tensile failure occurs when the material is subjected to excessive tensile stress, and the failure stress equals the tensile strength. Compressive failure occurs when the material is subjected to excessive compressive stress, and the failure stress equals the compressive strength. Shear failure occurs when the material is subjected to excessive shear stress, and the failure stress equals the shear strength. Fatigue failure occurs when the material is subjected to cyclic loading, and the failure stress is lower than the static strength.

[0074] The maximum shear stress theory posits that material failure depends on the magnitude of the maximum shear stress, occurring when this stress exceeds the material's shear stress limit. The shear stress limit of ultra-high molecular weight polyethylene (UHMWPE) is determined through shear tests, showing a limit of approximately 15 MPa. The von Mises stress theory, on the other hand, states that material failure depends on the magnitude of the distortion energy density, occurring when the von Mises stress exceeds the material's yield strength. The yield strength of UHMWPE is approximately 22 MPa.

[0075] The principle of stress superposition is used to calculate the equivalent stress under combined stress states by vector synthesis of stresses of different directions and types. Tensile stress, compressive stress, and shear stress are synthesized according to their direction and magnitude to obtain the magnitude and direction of the combined stress. The superposition of fatigue stresses considers the influence of stress amplitude and mean stress, and uses the modified Goodman equation for equivalent transformation.

[0076] The degradation threshold is determined using a safety factor method, where the ultimate strength of the material is divided by a safety factor to obtain the allowable stress. The selection of the safety factor considers the dispersion of material properties, the uncertainty of the load, and the complexity of the usage environment. For implantable medical devices, a safety factor between 2 and 3 is typically chosen. Based on a shear stress limit of 15 MPa and a safety factor of 2.5, the degradation threshold for shear stress is determined to be 6 MPa. Based on a yield strength of 22 MPa and a safety factor of 2.5, the degradation threshold for von Mises stress is determined to be 8.8 MPa.

[0077] Step S42 involves equivalent registration of the material's SN curve and load-wear curve with the stress analysis data of the spurious finite element simulation unit to characterize the service performance of the corresponding stress unit. Equivalent registration is a key technology for establishing the correlation between material and structural dimensions. By matching material experimental data with finite element analysis results, effective cross-scale information transfer is achieved.

[0078] The SN curve of a material describes the relationship between stress amplitude and fatigue life. Fatigue life data under different stress levels are obtained through fatigue tests. The SN curve of a cobalt-chromium-molybdenum alloy shows that its fatigue limit is approximately 400 MPa, at which stress level the material has an infinite fatigue life. The SN curve of a titanium alloy shows that its fatigue limit is approximately 500 MPa, and its fatigue performance is superior to that of the cobalt-chromium-molybdenum alloy. The SN curve of ultra-high molecular weight polyethylene shows that its fatigue performance is closely related to the stress level, exhibiting a longer fatigue life under low stress.

[0079] Load-wear curves describe the relationship between load magnitude and wear rate, obtaining wear data under different load conditions through wear tests. The load-wear curves for polyethylene materials show a non-linear relationship between wear rate and load; the wear rate is low under low loads, but increases rapidly with increasing load. The coefficient of friction also has a significant impact on the wear rate; an increase in the coefficient of friction leads to a significant increase in the wear rate.

[0080] Finite element simulation stress analysis provides stress distribution data for the sculptor under different loading conditions, with each element corresponding to a specific stress state. The stress data includes various stress components such as principal stresses, shear stresses, and von Mises stresses, providing rich information for equivalent registration. The time history of element stress reflects the influence of load variations on stress distribution, providing load spectrum data for fatigue analysis.

[0081] Equivalent registration is achieved through a stress mapping algorithm, matching the stress values ​​of finite element elements with the stress conditions of material tests. For fatigue analysis, the von Mises stress amplitude of the element is mapped to the stress axis of the SN curve to find the corresponding fatigue life. For wear analysis, the contact pressure of the element is mapped to the load axis of the load-wear curve to find the corresponding wear rate. The registration process considers the multiaxiality of the stress state and the randomness of the load, employing an equivalent stress conversion method to handle complex stress states.

[0082] Service performance is characterized by establishing a unit-level performance database, with each finite element element corresponding to a set of service performance parameters. These parameters include fatigue life, wear life, reliability, and failure probability. Fatigue life is calculated based on the SN curve and load spectrum, while wear life is calculated based on the wear rate and wear limit. Reliability and failure probability take into account the randomness of material properties and loads, and are calculated using probabilistic methods.

[0083] Step S43 involves optimizing the cross-scale fusion results using biomechanical analysis and patient clinical data to establish a cross-scale service evaluation model for knee prostheses. This model integrates experimental data, clinical data, simulation data, and material-scale data to make cross-scale information decisions, thereby obtaining the service life or service performance and reliability information for each unit. The establishment of the cross-scale service evaluation model is the core achievement of the entire method, enabling a comprehensive assessment of prosthesis service performance through the fusion of multi-source data.

[0084] Biomechanical analysis provides fundamental data on knee joint motion and loads, including kinematic and dynamic parameters. Kinematic parameters describe the knee joint's trajectory in three-dimensional space, including variations in flexion / extension, valgus / varus, and internal / external rotation angles. Dynamic parameters describe the magnitude and direction of loads on the knee joint, including the time histories of joint reactions and moments. These parameters provide boundary conditions and load inputs for finite element analysis.

[0085] Patient clinical data includes information from three stages: preoperative assessment, intraoperative documentation, and postoperative follow-up. Preoperative assessment data includes basic patient information, medical history, imaging results, and functional assessment results. Intraoperative documentation data includes the surgical approach, prosthesis type, alignment angle, and soft tissue balance. Postoperative follow-up data includes functional recovery, imaging changes, and the occurrence of complications. This data provides the basis for personalized adjustments to the model.

[0086] Cross-scale fusion employs a multi-layered data fusion architecture, comprising three layers: data layer fusion, feature layer fusion, and decision layer fusion. Data layer fusion standardizes and synchronizes raw data from different sources, ensuring data consistency and comparability. Feature layer fusion correlates and maps feature parameters at different scales, establishing cross-scale feature relationships. Decision layer fusion synthesizes and weighs the evaluation results from different layers to arrive at the final evaluation conclusion.

[0087] The knee joint prosthesis cross-scale service evaluation model employs a multi-input multi-output neural network structure. Inputs include multi-dimensional information such as patient characteristics, load parameters, material parameters, and geometric parameters. The hidden layers utilize a deep neural network structure, extracting complex correlations from the data through multi-layer nonlinear transformations. Outputs include service performance indicators such as fatigue life, wear life, reliability, and failure probability.

[0088] The model was trained using supervised learning, with training data derived from real-world clinical follow-up cases. The training objective was to minimize the error between predicted and observed results, using mean squared error as the loss function. The optimization algorithm employed gradient descent with an adaptive learning rate, updating network parameters via backpropagation. Model validation utilized cross-validation to ensure the model's generalization ability and prediction accuracy.

[0089] Cross-scale information decision-making employs multi-attribute decision theory to comprehensively evaluate information of different scales and types. Decision attributes include multiple dimensions such as material properties, structural properties, biocompatibility, and clinical efficacy. Attribute weights are determined through expert review and statistical analysis, reflecting the importance of different attributes to overall performance. The decision-making method combines the analytic hierarchy process (AHP) and fuzzy comprehensive evaluation to address uncertainties and fuzziness in the decision-making process.

[0090] The service life calculation for each element is based on the fatigue damage accumulation theory, transforming the complex load history into equivalent fatigue damage. Damage accumulation adopts the Palmgern-Miner rule, and the element is considered to have experienced fatigue failure when the accumulated damage reaches 1. Wear life calculation is based on the comparison between wear volume and wear limit; the element is considered to have experienced wear failure when the wear volume exceeds the limit. The combined life is the smaller value between fatigue life and wear life, reflecting the actual service capability of the element.

[0091] Service performance evaluation employs a multi-index comprehensive scoring method, which involves standardizing different performance indices and then weighting and summing them. Performance indices include strength, stiffness, durability, and stability. Weighting is determined based on clinical importance and engineering experience, with strength and durability indices having higher weights. Reliability information is obtained through probabilistic analysis, considering the randomness of material properties, load conditions, and environmental factors.

[0092] Step S44 involves setting different sets of force line alignment angle parameters based on the patient's individual basic information characteristics, and obtaining short-term service performance evolution data through a cross-scale service evaluation model. Personalized assessment is an important requirement for clinical applications, achieving accurate service performance prediction by considering individual patient differences.

[0093] Patient baseline information includes multiple aspects such as age, sex, weight, height, bone mineral density, activity level, and disease type. Age affects the rate of bone metabolism and repair capacity; older patients generally have weaker bone integration capabilities. Sex affects bone geometry and mechanical properties; female patients typically have lower bone mineral density than male patients. Weight and height determine a patient's body mass index, directly impacting the load borne by the knee joint.

[0094] Bone mineral density (BMD) is measured using dual-energy X-ray absorptiometry (DXA) and reflects the patient's bone condition. Patients with lower BMD tend to have less stable prostheses and are more prone to loosening. Activity level is assessed through a combination of questionnaires and objective measurements, including factors such as the intensity of daily activities, frequency of exercise, and occupational characteristics. Patients with higher activity levels have prostheses that bear greater loads, requiring higher durability from the prosthesis.

[0095] The force alignment angle parameter group is set based on the actual situation of clinical surgery, taking into account the influence of surgical precision and individual differences. The parameter group includes four levels: ideal alignment, mild deviation, moderate deviation, and severe deviation. Ideal alignment corresponds to the standard angle recommended by the surgical guidelines, with deviations within 1 degree in all directions. Mild deviation corresponds to the clinically acceptable deviation range, with deviations between 1 and 3 degrees in all directions. Moderate deviation corresponds to the deviation range requiring attention, with deviations between 3 and 5 degrees in all directions. Severe deviation corresponds to the deviation range requiring intervention, with deviations exceeding 5 degrees in all directions.

[0096] Each parameter set corresponds to a specific set of angle values, and a multi-dimensional parameter space is formed by combining angle deviations in different directions. The combination of coronal, sagittal, and axial angles produces a variety of possible alignment states, each corresponding to different mechanical environments and service conditions. The parameter sets are selected using orthogonal experimental design methods to reduce computational load while ensuring representativeness.

[0097] The input parameters of the cross-scale service evaluation model include two parts: patient characteristic parameters and force line angle parameters. Patient characteristic parameters are standardized before being input into the model to ensure that parameters of different dimensions have equal weights. Force line angle parameters are directly input into the model as geometric boundary conditions, influencing stress distribution and load transfer. The model's calculation process includes two stages: forward propagation and result output. Forward propagation uses a neural network to calculate intermediate features, and the result output converts these intermediate features into service performance indicators.

[0098] Short-term service performance evolution data reflects the performance changes of the prosthesis in the early stages of implantation, typically within two years post-surgery. Evolutionary data includes changes in stress distribution, adjustments in contact conditions, and adaptation of material properties. Changes in stress distribution are primarily caused by the osseointegration process; as the integration between bone tissue and the prosthesis gradually stabilizes, the stress distribution tends to become more uniform. Adjustments in contact conditions reflect the break-in process of the articular surfaces; initially, there may be localized high-stress areas, which gradually improve with use.

[0099] Evolutionary data were acquired through time series analysis, using service time as an input variable to predict performance status at different time points. The selection of time points was based on the clinical follow-up schedule, including key time points such as 1 month, 3 months, 6 months, 1 year, and 2 years post-surgery. Each time point corresponds to a complete set of performance data, including indicators such as stress level, wear degree, fatigue damage, and overall score.

[0100] Step S45: Determine the safe alignment range by comparing with the alignment comfort line judgment criteria, and obtain long-term service performance time-series evolution data through the model. Determining the safe alignment range is an important basis for clinical decision-making; the acceptable alignment deviation range is determined through comparative analysis.

[0101] The criteria for determining alignment comfort lines are based on statistical analysis of large-sample clinical data, reflecting the overall satisfaction level of the patient population. The criteria include both objective and subjective indicators. Objective indicators are based on imaging measurements and functional tests, while subjective indicators are based on patient satisfaction surveys and quality of life assessments. Thresholds for objective indicators were determined through statistical analysis, while thresholds for subjective indicators were determined through questionnaires.

[0102] The safe force line range was determined using a comparative analysis method, comparing the predicted results for individual patients with comfort force line standards. When the predicted service performance indicators meet the comfort force line standards, the corresponding force line angle is considered safe. The boundaries of the safe range were determined by progressively adjusting the force line angle parameters to find the maximum deviation range that meets the standards. The safe range may vary among different patients, reflecting the need for individualized treatment.

[0103] The safe range of force lines is represented in the form of multi-dimensional intervals, with each dimension corresponding to a safe range in an angular direction. The coronal safe range is typically between -3 degrees and +3 degrees, the sagittal safe range is typically between -4 degrees and +4 degrees, and the axial safe range is typically between -6 degrees and +6 degrees. Specific values ​​are adjusted according to the individual characteristics of the patient, with higher-risk patients having relatively narrower safe ranges.

[0104] Long-term service performance time-series evolution data reflects the performance change trend of the prosthesis throughout its entire service life, typically spanning 15 to 20 years. The evolution data is acquired based on extrapolation prediction methods, using short-term data to predict long-term trends. The prediction model considers the influence of various factors such as material aging, fatigue accumulation, wear progression, and biological changes.

[0105] The temporal evolution of the prosthesis can be divided into three stages: an initial stable period, a mid-term slow degradation period, and a late-term rapid degradation period. The initial stable period corresponds to 2 to 5 years post-surgery, during which the prosthesis performance is basically stable with minimal changes in various indicators. The mid-term slow degradation period corresponds to 5 to 15 years post-surgery, during which the prosthesis performance begins to decline slowly, and wear and fatigue damage gradually accumulate. The late-term rapid degradation period corresponds to after 15 years post-surgery, during which the prosthesis performance declines rapidly, potentially requiring revision surgery.

[0106] The accuracy of the evolutionary data was validated by comparison with long-term follow-up data to ensure the reliability of the prediction results. The validation data came from multi-center long-term follow-up studies, including real-world data from different types of prostheses and different patient populations. Analysis of the prediction error showed that the model has high accuracy in short- to medium-term predictions, but there is some uncertainty in long-term predictions.

[0107] Step S46 involves comparing the material performance degradation threshold to predict the service life of the knee prosthesis under different reliability levels or obtaining reliability results under different service life conditions. Predicting service life is a core issue in prosthesis design and clinical application; reliability analysis provides quantitative life assessment results.

[0108] The comparative analysis of material performance degradation thresholds is achieved by comparing predicted stress levels with degradation thresholds. When the predicted stress exceeds the degradation threshold, the material is considered to have begun to degrade, affecting the normal function of the prosthesis. Degradation thresholds include multiple aspects such as fatigue threshold, wear threshold, strength threshold, and stability threshold, each corresponding to a different failure mode.

[0109] The fatigue threshold is determined based on the material's fatigue limit; fatigue failure occurs when the stress amplitude exceeds the fatigue limit. The wear threshold is determined based on a critical wear rate; the wear rate accelerates significantly when the wear rate exceeds this critical value. The strength threshold is determined based on the material's yield strength or ultimate strength; plastic deformation or fracture occurs when the stress exceeds this strength threshold. The stability threshold is determined based on the bonding strength between the prosthesis and bone tissue; loosening occurs when the interfacial stress exceeds this bonding strength.

[0110] Reliability is defined based on probability theory, representing the probability that a prosthesis will function normally within a specified time. Reliability calculations consider the influence of various random factors, such as material properties, load conditions, environmental factors, and manufacturing quality. The distribution characteristics of these random factors are determined through statistical analysis; commonly used distribution types include normal distribution, log-normal distribution, and Weibull distribution.

[0111] Service life at different reliability levels is calculated using the inverse function of the reliability function. The reliability function describes the relationship between reliability and time, typically exhibiting a decreasing trend. Given a reliability level, the corresponding service life is obtained by solving the inverse function of the reliability function. Commonly used reliability levels include 90%, 95%, and 99%, corresponding to different risk tolerance levels.

[0112] The predicted service life is represented by a probability distribution, including statistical parameters such as expected life, median life, and characteristic life. Expected life represents the average service time, median life represents the service time corresponding to 50% reliability, and characteristic life represents the scaling parameter of the Weibull distribution. The shape parameter of the life distribution reflects the variation of the failure rate: when the shape parameter is less than 1, the failure rate decreases; when it is equal to 1, the failure rate remains constant; and when it is greater than 1, the failure rate increases.

[0113] Reliability results under different service life conditions are obtained through forward calculation of the reliability function. Given a service life, the corresponding reliability level is obtained by calculating the value of the reliability function. This analytical method is applicable to situations where the service life is known, and assesses the reliability level of the prosthesis within that service life. In clinical applications, the reliability level at key time points such as 10 years, 15 years, and 20 years is often the focus.

[0114] The present invention also provides a cross-scale force line safety boundary determination system for knee replacement, the cross-scale force line safety boundary determination system for knee replacement comprising: The dataset module is used to acquire data on the physical properties and service life of knee prostheses. It is used to collect individual patient characteristic data and multimodal clinical data to form a usable dataset. The pressure distribution module is used to perform simulation analysis of different alignment angles of the knee joint prosthesis based on the available dataset. It uses a finite element model to simulate the mechanical response under various combinations of alignment errors to determine the stress and pressure distribution. The performance evaluation module is used to acquire multi-dimensional experimental data of knee joint prostheses and to evaluate their performance by conducting material fatigue and wear tests as well as in vitro friction tests. The modeling and evaluation module is used to conduct service evaluation of knee joint prostheses based on simulation analysis and experimental data, and to establish an evaluation model by integrating material scale and structural scale data through equivalent stress registration, and to determine the range of safe force lines and predict service performance.

[0115] The present invention also provides a readable storage medium comprising a processor and a memory, the memory being connected to the processor, the processor containing a program that executes the described method for determining cross-scale force line safety boundaries for knee replacement. Example 1: The working principle and overall process of this invention are as follows: Figure 2 As shown, it includes: (1) Knee joint prosthesis physical properties and service data management We collect basic physical property data of knee joint prostheses, and perform data cleaning and consistency analysis on structural material properties and pre- and post-operative patient characteristics to form a complete and usable object-oriented dataset. Simultaneously, we preprocess and analyze comparative data such as imaging data and gait data.

[0116] a. Basic data collection and data governance Patient information: age, gender, BMI, implantation time, whether it is a condylar knee implant, diagnosis (instability, infection, loosening, surrounding fracture, etc.), preoperative Oxford knee score, preoperative range of motion, etc. The results of the alignment angle measurements (coronal plane: preoperative and postoperative hip-knee-ankle angle HKA, distal medial angle of femoral prosthesis MDFA, proximal medial angle of tibial prosthesis MPTA). Postoperative follow-up and clinical assessment results (HSS score): including pain, function, range of motion, muscle strength, flexion deformity and stability, and postoperative imaging analysis.

[0117] b. Postoperative status evaluation based on imaging data Postoperative images of the patient's joints were classified to evaluate the postoperative condition as good or bad. Postoperative joint prosthesis failure (wear) analysis based on imaging data was conducted to compare differences and observe the wear at the joint site.

[0118] c. Simulation analysis of loads and constraints for knee joint prostheses Obtain the maximum mises stress and maximum contact pressure under different alignment errors. Gait loading parameters, such as gait, flexion-extension angle (KFA), knee adduction moment (KAM), knee flexion-extension moment (KFM), maximum axial reaction force, anterior-posterior displacement, tibial rotation angle, etc.

[0119] (2) Simulation analysis of different alignment angles of knee joint prosthesis This study investigates the mechanical response of knee prostheses under different coronal alignment (6° varus to 6° valgus) design combinations using finite element method (FEM) simulation. Based on common alignment errors in total knee arthroplasty, various alignment scenarios ranging from -6° to 6° were selected in the coronal plane, creating hundreds of models. Gait loads were applied to each model, and the Mises stress and contact pressure of the liner under the interaction of different alignment errors were calculated. Based on the stress response and combined with a life prediction model, the impact of different alignment angles on prosthesis lifespan was determined.

[0120] (3) Multidimensional experimental and service performance analysis of knee joint prostheses The experiments included fatigue tests on the femoral and tibial components of the artificial knee joint (cobalt-chromium-molybdenum alloy), wear tests on the tibial liner material (ultra-high molecular weight polyethylene), and in vitro friction and wear tests.

[0121] a. Metal Fatigue Testing: Fatigue testing is used to evaluate the durability of artificial knee joint materials during long-term use. Tensile fatigue tests are performed on samples of the target material using a high-frequency fatigue testing machine. During the test, the sample is subjected to cyclic loading to simulate the stress conditions of a human knee joint during daily activities. By recording the number of loading cycles and the stress-strain curves of the sample, the fatigue life and durability of the material can be evaluated.

[0122] b. Polyethylene material wear test: Following the wear test method of ISO 14243, samples of femoral and tibial components and tibial padding material were extracted from the artificial knee joint prototype and tested on a friction and wear testing machine. During the test, a certain load was applied between the sample and the friction pair, and reciprocating motion was performed to simulate the movement of the human knee joint. By recording the initial mass and the mass change after a certain period, the wear performance and mass wear of the material can be evaluated.

[0123] c. In vitro friction and wear test: The prosthesis is implanted on the Sawbones, different loads are applied, different force line alignments are set, and in conjunction with the simulation experiment design, representative experimental parameters are selected to conduct in vitro friction and wear tests to obtain prosthesis wear data under different loads.

[0124] (4) Service evaluation of knee prostheses based on cross-scale correlation Artificial joint materials, surgical procedures, and individual patient factors all affect the implantation alignment and service life of artificial joints. This invention addresses this by conducting prosthesis failure threshold analysis, a cross-scale service evaluation model, and analysis of the safe force line range and service performance prediction of artificial knee joints.

[0125] a. Prosthesis service threshold: There are relatively many types of knee prosthesis failures, common types include polyethylene liner wear, aseptic loosening, tibial plateau fatigue fracture, etc. It is necessary to determine the knee prosthesis alignment comfort line and the knee material performance degradation threshold based on statistical data.

[0126] b. Cross-scale service evaluation model: The performance results of material-scale fatigue and wear tests are correlated with equivalent stress to characterize the service performance of the corresponding stress elements in the finite element simulation analysis. Furthermore, biomechanical analysis and patient clinical data are used to optimize the cross-scale fusion results and establish a cross-scale service evaluation model for knee joint prostheses.

[0127] c. Analysis of the safe force line range of artificial knee joints: Based on the patient's individual basic information characteristics, different groups of force line alignment angle parameters are set. The service performance evaluation time-series evolution data are obtained through a cross-scale service evaluation model. By comparing with the knee prosthesis alignment comfort force line judgment standard, the safe force line range of artificial knee joints for individual patients is determined.

[0128] d. Prediction of service performance of artificial knee joints: Based on the patient's basic information characteristics, long-term service performance time-series evolution data are obtained through a cross-scale service evaluation model. The data are compared with the knee joint material performance degradation threshold to determine the service life of the knee joint prosthesis under different reliability levels.

[0129] The above methods were demonstrated through actual experimental operations: (1) Knee joint prosthesis physical properties and service data management First, it is necessary to systematically collect patient demographic characteristics (such as age, sex, BMI), surgical details (implantation time, prosthesis type), diagnostic status (including joint instability, infection, prosthesis loosening, and surrounding bone damage), and preoperative functional status assessment (Oxford knee score, range of motion) to ensure the comprehensiveness and accuracy of the data. Then, medical imaging technology should be used to accurately measure and record key anatomical angles before and after surgery, including the coronal hip-knee-ankle angle (HKA) and the medial angles of the femoral and tibial prostheses (MDFA, MPTA), laying the foundation for in-depth analysis of prosthesis alignment. Finally, a standardized postoperative follow-up procedure should be established, HSS score data collected, and the patient's rehabilitation progress comprehensively assessed, including pain relief, joint function recovery, increased range of motion, muscle strength enhancement, deformity correction, and improved joint stability. This data should be combined with detailed analysis of imaging data to objectively evaluate the surgical outcome and long-term prosthesis performance.

[0130] Secondly, angle, displacement, and stress data of the prosthesis were collected during different gait loading processes. These data reflect the stress state of the prosthesis under different gait loading conditions. To address the issues of temporal correlation and feature confusion, a multimodal data fusion method was adopted. In the feature extraction stage, a CNN algorithm was used to extract features from different data modalities. For angle data, it can be treated as image data, and spatial features were extracted using convolutional and pooling layers. For displacement and stress data, it can be treated as time-series data, and time-related features were extracted using convolutional and pooling layers. In the feature aggregation stage, a multi-layer cross-feature fusion method was used to aggregate features from different data modalities. By cross-referencing the features of angle, displacement, and stress data, a richer feature representation can be obtained. In this way, features from different data modalities can be aggregated into a single feature vector. In the time-series modeling stage, a BiLSTM algorithm was used to model the feature vectors over time. Through a bidirectional long short-term memory network structure, the temporal correlation in the feature vectors can be captured, thereby better understanding the state changes of the prosthesis at different time points. Finally, in the classification stage, the results of the time-series modeling were classified using fully connected layers. Fully connected layers can map the results of time series modeling to specific damage assessment categories, enabling accurate prosthesis damage assessment.

[0131] To validate and evaluate the proposed method, experiments were conducted using real gait loading data. The dataset was divided into training and test sets. The CNN and BiLSTM models were trained on the training set and then evaluated on the test set. The performance and advantages of the proposed method were assessed by comparing it with existing methods. Furthermore, based on the experimental results, the network structure and parameter settings can be adjusted to improve the model's accuracy and robustness. This research approach aims to overcome the limitations of existing time-series classification models in prosthetic injury evaluation due to temporal correlation and feature confusion. Through multi-layer cross-feature fusion of CNN and BiLSTM, accurate feature extraction, aggregation, and interpretation of different data modalities can be achieved, thereby enabling accurate prosthetic injury evaluation.

[0132] To conduct simulation analysis of knee joint prostheses, it is necessary to collect knee joint prosthesis loads and constraints, gait loading parameters, including gait, flexion-extension angle (KFA), knee adduction moment (KAM), knee flexion-extension moment (KFM), maximum axial reaction force, anterior-posterior displacement, tibial rotation angle, etc.

[0133] (2) Simulation analysis of different alignment angles of knee joint prosthesis To determine the stress on various parts of the knee prosthesis under different alignment angles, this project proposes to use finite element method (FEM) simulation analysis to perform structural mechanics analysis under different angle combinations based on a three-dimensional model of the knee prosthesis. Combining material experimental results from physical experiments, a life prediction model for the knee prosthesis under different alignment angles will be established.

[0134] a. Select simulation design schemes with different alignment angle combinations. Using finite element simulation technology, study the mechanical response of the knee prosthesis under different combinations of coronal alignment, sagittal alignment, and rotational alignment designs. Based on the alignment errors frequently occurring in total knee arthroplasty, establish multiple models with various alignment conditions ranging from -6° to 6° in the coronal plane.

[0135] b. Perform gait loading finite element analysis on each combined alignment angle model. By integrating current load loading methods, standard gait loads are sequentially applied to the selected simulation models, and the Mises stress and contact pressure of the liner under the interaction of different alignment errors are calculated.

[0136] c. The stress response at each unit level of different simulation models is updated in real time. Based on the stress response and in conjunction with the results of material-level physical experiments, a life prediction model is proposed to determine the impact of different alignment angles on the prosthesis life.

[0137] (3) Multidimensional experimental and service performance analysis of knee joint prostheses This study aims to evaluate the performance of the femoral and tibial components (cobalt-chromium-molybdenum alloy) and the tibial liner material (ultra-high molecular weight polyethylene) of artificial knee joints in terms of fatigue and wear. At the same time, in vitro friction and wear experiments were carried out at the structural level to provide a scientific basis for improving and optimizing the design and material selection of artificial knee joint prostheses.

[0138] Fatigue test of metallic materials: a. Take a sample of the target material and conduct a tensile fatigue test.

[0139] b. Use a GPS350 high-frequency fatigue testing machine to clamp the sample at both ends of the top plate and apply fatigue load.

[0140] c. Record the number of loading cycles and the stress-strain curves of the samples to evaluate the fatigue life and durability of the material.

[0141] Wear test of polyethylene material: a. Samples of femoral and tibial components and tibial liner material were extracted from the artificial knee joint prototype.

[0142] b. The wear test shall be conducted on a friction and wear testing machine using a wear test method conforming to ISO 14243.

[0143] c. Record the initial mass, and after in vitro wear tests at 0.5, 1, 2, 3, 4, and 5 MC, record the mass changes and mass wear conditions.

[0144] External friction and wear test: a. Take a sample of the target material and implant the prosthesis into the artificial bone Sawbones.

[0145] b. Referencing the simulation input parameters, apply different loads and angles to conduct friction and wear tests on the structural components.

[0146] c. Analyze knee joint prosthesis wear under time and load conditions.

[0147] Data Analysis: The performance characteristics of artificial knee joint materials and structures are evaluated by analyzing the results of fatigue, wear, and in vitro friction and wear tests. Statistical methods and visualization tools are used to process and interpret the test data. Based on the evaluation results, suggestions for improving and optimizing the design and material selection of artificial knee joint prostheses can be proposed to enhance their performance and durability, thereby improving patients' quality of life and rehabilitation outcomes.

[0148] (4) Service evaluation of knee prostheses based on cross-scale correlation This study aims to obtain stress cloud maps of prosthesis units using fatigue SN curves of metallic knee prostheses, load-wear results of polyethylene materials, and finite element simulation technology. Based on equivalent stress, registration and fusion are then performed to achieve cross-scale information fusion between material and structural dimensions. Furthermore, by integrating experimental, clinical, simulation, material-scale data, and biomechanical methods, cross-scale information decision-making is conducted to establish a service evaluation model for knee prostheses. This model enables alignment evaluation and failure assessment of knee prostheses from both cross-scale and clinical perspectives, yielding service life, performance, and reliability information for each unit.

[0149] a. Service life threshold of prostheses Determination of knee prosthesis alignment comfort line: Based on the preoperative and postoperative evaluation data collected clinically, the comfort line data with higher satisfaction were selected. Multivariate statistical analysis was performed on the above data to optimize the range of comfort line data parameters, which were then used as the standard for determining the comfort line.

[0150] And the threshold for knee joint material performance degradation: When the polyethylene liner is worn, the sum of stresses caused by various material failure modes should not exceed the maximum shear stress limit of the tibial liner material; generally, the shear stress reaches its maximum value at a thickness of approximately 1 mm below the joint contact surface. The maximum shear stress limit and von Mises stress of the material can be used as parameters to assess material strength.

[0151] b. Cross-scale service evaluation model The SN curve and load-wear curve of the material are equivalently registered with the stress analysis data of the finite element simulation unit of the spur.

[0152] Characterizes the service performance of the corresponding stress element in finite element simulation analysis.

[0153] The cross-scale fusion results were optimized using biomechanical analysis and patient clinical data.

[0154] Establish a cross-scale service evaluation model for knee joint prostheses.

[0155] c. Analysis of the safe force line range of artificial knee joints Experimental design for safety force line analysis: Based on the patient's individual basic information characteristics, different sets of force line alignment angle parameters are set.

[0156] Obtain the evolution results of force line analysis features: Input personal basic information data and alignment angle information, and obtain the time series evolution data of service performance evaluation through the cross-scale service evaluation model.

[0157] Determining the safe force line range: Based on short-term (e.g., 1 year) service performance evolution data and the knee prosthesis alignment comfort force line judgment criteria, determine the safe force line range of the artificial knee joint for each individual patient.

[0158] d. Prediction of service performance of artificial knee joints Obtain service performance time-series evolution data: Based on the patient's basic information characteristics and alignment characteristics, long-term service performance time-series evolution data are obtained through a cross-scale service evaluation model.

[0159] Long-term service performance prediction: By comparing the degradation threshold of knee joint materials, the service life of knee joint prostheses under different reliability levels can be predicted, or the reliability results under different lifespan conditions can be obtained.

[0160] The foregoing has provided a detailed description of a method, system, and readable storage medium for determining cross-scale force line safety boundaries for knee replacement, as provided in the embodiments of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas; furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

[0161] Certain terms are used in the specification and claims to refer to specific components. Those skilled in the art will understand that hardware manufacturers may use different names to refer to the same component. This specification and claims do not distinguish components based on differences in name, but rather on differences in function. The terms "comprising" and "including" used throughout the specification and claims are open-ended and should be interpreted as "comprising / including but not limited to". "Approximately" means that within an acceptable margin of error, those skilled in the art can solve the technical problem and substantially achieve the technical effect within a certain margin of error. The following descriptions in the specification are preferred embodiments for carrying out this application; however, these descriptions are for the purpose of illustrating the general principles of this application and are not intended to limit the scope of this application. The scope of protection of this application shall be determined by the appended claims.

[0162] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a product or system comprising a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a product or system. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the product or system that includes said element.

[0163] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0164] The foregoing description illustrates and describes several preferred embodiments of this application. However, as previously stated, it should be understood that this application is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the application concept described herein through the foregoing teachings or techniques or knowledge in related fields. Any modifications and variations made by those skilled in the art that do not depart from the spirit and scope of this application should be within the protection scope of the appended claims.

Claims

1. A cross-scale force line safety boundary determination method for knee replacement, characterized in that, The knee replacement-oriented cross-scale force line safety boundary determination method comprises the following steps: S1: Obtain knee prosthesis physical properties and service data, and manage through collection of patient individual characteristic data and multi-modal clinical data to form a usable data set; S2: According to the simulation analysis of the different alignment angles of the knee prosthesis in the available data set, the mechanical response under a plurality of alignment error combinations is simulated through a finite element model to determine the stress and pressure distribution; S3: Obtain multi-dimensional experimental data of the knee prosthesis, and evaluate the performance by carrying out material fatigue and wear experiments and in-vitro friction tests; S4: According to the simulation analysis and experimental data, the service evaluation of the knee prosthesis based on cross-scale correlation is carried out, the evaluation model is established by fusing the material scale and structure scale data through equivalent stress registration, and the safe force line range and the predicted service performance are determined.

2. The cross-scale force line safety boundary determination method for knee replacement according to claim 1, wherein, The knee prosthesis physical property and service data obtained in S1 comprises: S11: Collect patient age, gender, BMI, implantation time, diagnosis, preoperative score and alignment angle measurement results, and classify the image data to evaluate the postoperative state and wear amount; S12: Fuse gait loading parameters such as flexion angle, adduction torque, flexion torque, axial force, anteroposterior displacement and rotation angle, and extract the spatial features of the angle data and the time features of the displacement stress data through the CNN algorithm; S13: Use a multi-layer cross-feature fusion method to aggregate different modal features into a feature vector, and use a BiLSTM algorithm for time series modeling to capture time correlation; S14: Classify the time series modeling results through a fully connected layer to obtain damage evaluation categories, and divide the data set into a training set and a test set to train and evaluate the model to determine the performance advantage.

3. The cross-scale alignment safety margin determination method for knee replacement of claim 1, wherein, The simulation analysis of different alignment angles of the knee prosthesis in the available data set in S2 comprises: S21: Based on the three-dimensional model of the knee prosthesis, a plurality of finite element models are established for the alignment from negative six degrees to positive six degrees, and the mechanical response under different alignment design combinations is studied; S22: Apply standard gait load to each model to calculate the liner Mises stress and contact pressure, and based on the stress response, the material experiment results are obtained; S23: A life prediction model is proposed to determine the influence of different alignment on the service life of the prosthesis, and the stress response of each unit is updated in real time to obtain the maximum stress and maximum contact pressure under different alignment errors.

4. The cross-scale alignment safety margin determination method for knee replacement of claim 1, wherein, The multi-dimensional experimental data of the knee prosthesis obtained in S3 comprises: S31: Tensile fatigue test is carried out on metal materials, a high-frequency fatigue testing machine is used to clamp the sample, cyclic loading is applied, stress-strain curve is recorded, and fatigue life is evaluated; S32: For polyethylene materials, the components are taken out from the sample for wear test in accordance with ISO standard, load is applied, and mass change is recorded to evaluate the wear performance; S33: Implant the prosthesis on the artificial bone and apply different loads and angles for in-vitro friction and wear test, and analyze the wear under time domain and load conditions; Use statistical methods and visualization tools to process test data to evaluate material and structural performance characteristics, and determine optimization design suggestions according to the evaluation results.

5. The cross-scale alignment safety margin determination method for knee replacement of claim 1, wherein, The S4 in the simulation analysis and experimental data based on cross-scale correlation knee prosthesis service evaluation, comprising: S41: determine the prosthesis service threshold, through the multivariate statistical analysis of clinical data to filter the comfortable force line data optimization parameter range as the determination standard, and determine the material performance degradation threshold based on the maximum shear stress limit and von Mises stress; S42: the material S-N curve and the load wear curve are equivalent to the stress analysis data of the finite element simulation unit of the prosthesis, and the service performance of the corresponding stress unit is characterized; S43: using biomechanical analysis and patient clinical data to optimize the cross-scale fusion results, establish a cross-scale service evaluation model of knee prosthesis, and make cross-scale information decision based on test data, clinical data, simulation data and material scale data to obtain the service life or service performance and reliability information of each unit.

6. The cross-scale alignment safety margin determination method for knee replacement of claim 5, wherein, The S41 in determining the prosthesis service threshold, comprising: The preoperative and postoperative evaluation data collected in the clinic are filtered to obtain the comfortable force line data with high satisfaction, and the comfortable force line data parameter range is optimized by multivariate statistical analysis; If the sum of stress caused by various material failure forms exceeds the maximum shear stress limit of tibial pad material, it is determined that the performance is degraded, and the maximum shear stress limit and von Mises stress are used as the parameter index to evaluate the material strength to determine the degradation threshold.

7. The cross-scale alignment safety margin determination method for knee replacement of claim 5, wherein, The S43 in establishing the cross-scale service evaluation model of knee prosthesis, comprising: According to the equivalent registration and fusion of material scale fatigue and wear test performance results and finite element simulation analysis of corresponding stress unit; Through biomechanical analysis and patient clinical data, the fusion results are optimized, and different force line angle parameter groups are set according to the personal basic information characteristics of the patient to input the model to obtain the service performance evolution data; Compare the comfortable force line determination standard of knee prosthesis alignment to determine the safe force line range for the patient, and compare the material performance degradation threshold to determine the service life under different reliability.

8. The trans-scale force line safety boundary determination method for knee replacement of claim 1, wherein, The S4 in determining the safe force line range and predicting the service performance, comprising: Different force line alignment angle parameter groups are set according to the personal basic information characteristics of the patient, and short-term service performance evolution data is obtained through the cross-scale service evaluation model; Compare the comfortable force line determination standard to determine the safe force line range, and obtain the long-term service performance time evolution data through the model; Compare the material performance degradation threshold to determine the service life of the knee prosthesis under different reliability or obtain the reliability results under different service life.

9. A cross-scale force line safety boundary determination system for knee replacement, characterized in that, The cross-scale force line safety boundary determination system for knee replacement, comprising: A data set module for obtaining the physical properties and service data of knee prosthesis, collecting patient individual characteristic data and multi-modal clinical data to form a usable data set; A pressure distribution module for simulating the mechanical response of knee prosthesis under various alignment error combinations by finite element model to determine the stress and pressure distribution according to the usable data set; A performance evaluation module for obtaining multi-dimensional experimental data of knee prosthesis by carrying out material fatigue and wear experiments and in vitro friction tests to evaluate the performance. The modeling evaluation module is used for cross-scale correlation-based knee prosthesis service evaluation according to simulation analysis and experimental data, an evaluation model is established by fusing material scale and structure scale data through equivalent stress registration, and a safe force line range and a service performance are predicted.

10. A readable storage medium, characterized by, The readable storage medium comprises a processor and a memory, the memory is connected with the processor, the processor comprises a program, and the program performs the cross-scale force line safety boundary determination method for knee replacement in any one of claims 1-8.

Images