An orthopedic medical instrument digital twin simulation method and system
By constructing a personalized digital twin of the skeleton and a simulation engine that integrates computation and measurement, multibody dynamics calculation and machine learning, we can achieve efficient simulation and prediction of orthopedic medical devices in complex physiological environments. This solves the problem of balancing modeling accuracy and efficiency in traditional R&D, significantly shortening the R&D cycle and reducing costs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- FIRST AFFILIATED HOSPITAL OF DALIAN MEDICAL UNIV
- Filing Date
- 2026-04-01
- Publication Date
- 2026-06-19
Smart Images
Figure CN122242157A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of digital medical device technology, and in particular to a medical device simulation method and system. Background Technology
[0002] Traditional orthopedic medical device development primarily relies on materials science, mechanical engineering, and large-scale manufacturing technologies. Product design and validation processes typically include physical prototype testing, in vitro biomechanical simulation experiments, and animal testing. Taking joint prosthesis development as an example, researchers design standardized prosthetic components based on human anatomical statistics, and during surgery, combine components of different sizes to accommodate individual differences. In the development of spinal internal fixation systems, the fixation effect is simulated using components such as pedicle screws and connecting rods, based on the spinal anatomy and three-dimensional kinematic characteristics. The development of trauma implants requires consideration of the anatomical morphology of different bone sites, the ratio of cortical to cancellous bone, and blood supply characteristics.
[0003] The aforementioned traditional R&D model has the following limitations: First, physical prototype testing and animal experiments are time-consuming and costly, and are limited by sample accessibility and ethical constraints; Second, in vitro simulation conditions can usually only cover limited static or simplified conditions, making it difficult to fully reflect the mechanical response of implants in complex physiological environments; Third, design flaws are often only exposed in later stages of trials or clinical applications, resulting in high R&D risks and low iteration efficiency; Fourth, existing simulation tools face a balance between modeling accuracy and computational efficiency, making it difficult to simultaneously meet the needs of rapid iteration and high-confidence verification. Summary of the Invention
[0004] This application provides a digital twin simulation method and system for orthopedic medical devices, aiming to solve the problem in the prior art that it is difficult to balance modeling accuracy and computational efficiency in medical device R&D simulation, resulting in long R&D cycles and late exposure of design defects.
[0005] Firstly, a digital twin simulation method for orthopedic medical devices is provided, including:
[0006] Constructing a skeletal digital twin: Reconstructing the three-dimensional anatomical geometry of the skeleton based on the patient's medical imaging data, and integrating real-time sensor information and physiological load spectrum, endowing the three-dimensional anatomical geometry with individualized biomechanical properties and dynamic response characteristics, thereby obtaining a skeletal digital twin that integrates form and function;
[0007] Fusion simulation of computation and measurement: Construct a simulation engine that integrates finite element analysis, multibody dynamics calculation and machine learning-driven proxy model to obtain measured data of implant in-situ state, and fuse and dynamically calibrate the physics-based numerical calculation results with the measured data of implant in-situ state; so as to realize real-time simulation and prediction of the mechanical performance of implanted devices under complex physiological environment.
[0008] 3D visualization feedback: Construct a virtual assembly environment containing the medical device to be validated and a personalized skeleton. Driven by the physiological load spectrum, use a 3D visualization engine to dynamically render the spatial position and interaction state of the two, and present key biomechanical indicators intuitively and quantitatively in the form of engineering cloud maps.
[0009] Optionally, in the above scheme, the step of constructing a skeletal digital twin further includes:
[0010] Based on the patient's medical imaging data, the three-dimensional anatomical geometry of the target bone is reconstructed;
[0011] Based on individual bone density distribution, determine the material property gradient between cortical bone and cancellous bone in the target bone;
[0012] By assigning the material property gradient to the three-dimensional anatomical geometry, a skeletal digital twin that conforms to the biomechanical characteristics of the individual is obtained, thus overcoming the shortcomings of traditional homogenized models in representing individual anatomical variations.
[0013] Optionally, in the above scheme, a data preprocessing and fusion step is included before the computational-measurement fusion simulation step, including:
[0014] Collect raw monitoring data generated by various sensors on surgical instruments and transmit it to the computing terminal;
[0015] The original monitoring data is cleaned to remove duplicate, redundant and abnormal data, and the cleaned original monitoring data is sorted and formatted according to the dependency of the numerical calculation process.
[0016] Based on biomechanical algorithms, the processed data is solved and analyzed to deduce the load distribution and mechanical constraints of medical devices.
[0017] The cleaned sensor data is integrated and fused with the load distribution and mechanical constraints to form a unified data set, providing input for subsequent simulation and prediction.
[0018] Optionally, in the above scheme, the step of providing 3D visualization feedback further includes:
[0019] It dynamically displays the stress, strain distribution, and strain energy density inside bones and implants in the form of color cloud maps, displays the displacement field and interface micro-motion direction in the form of vector maps, and supports viewing internal details through interactive cutting, perspective, or cross-sectional viewing methods.
[0020] In the form of digital panels, dynamic curves or dashboards, key biomechanical performance parameters such as contact pressure, micro-motion amplitude, osseointegration potential, long-term fatigue life and stress shielding effect are displayed in real time in a quantitative manner.
[0021] Based on the visualization rendering results and quantitative parameters, a structured simulation verification report containing biomechanical safety and effectiveness assessment conclusions is automatically generated.
[0022] Optionally, in the above scheme, the calculation and measurement fusion simulation step further includes: using contact pressure and strain field data obtained from in vitro biomechanical experiments or limited in vivo monitoring to continuously optimize the boundary conditions, contact friction coefficient, and bone material constitutive model parameters in the simulation online, so as to minimize the deviation between the simulation results and the actual physiological response.
[0023] Optionally, the above scheme may also include a long-cycle performance prediction step:
[0024] Based on the proxy model generated by the computational-measurement fusion simulation engine, at least one parameter among the following in orthopedic medical devices—osseointegration potential, long-term fatigue life, stress shielding effect, and interface fretting wear—can be rapidly predicted and analyzed for sensitivity within a preset short time threshold:
[0025] Optionally, the above scheme may also include a design space exploration and optimization step:
[0026] Based on the surrogate model, a multi-objective optimization algorithm is executed in the design space containing geometric parameters, material parameters and surface feature parameters to automatically find the optimal design scheme of orthopedic medical devices that meets the biomechanical safety constraints.
[0027] The optimization objective of the multi-objective optimization algorithm includes at least one of the following: minimizing stress shielding effect and interface fretting wear, maximizing osseointegration potential and long-term fatigue life.
[0028] The constraints include the physiological load-bearing threshold of bone and / or the strength limit of implant materials.
[0029] Optionally, the above scheme may also include an optimization scheme verification and feedback step:
[0030] The optimized design scheme is re-imported into the virtual assembly environment, and the above-mentioned calculation and measurement fusion simulation and three-dimensional visualization feedback steps are executed.
[0031] The key biomechanical indicators before and after optimization were compared and analyzed, and the optimization effect was presented in a visual form.
[0032] Optionally, in the above scheme, the three-dimensional visualization feedback step further includes:
[0033] In response to user interaction with a specific area in the engineering cloud map, the corresponding biomechanical performance values and their time-varying curves for that area are dynamically displayed.
[0034] The system synchronously displays the physiological load spectrum time history curve corresponding to the current simulation condition and supports user-defined modification of the load spectrum to trigger a resimulation.
[0035] Secondly, a digital twin simulation system for orthopedic medical devices is provided for performing the above-mentioned methods, including:
[0036] A skeletal digital twin module is constructed to reconstruct the three-dimensional anatomical geometry of the skeleton based on the patient's medical imaging data. Real-time sensor information and physiological load spectrum are integrated to endow the three-dimensional anatomical geometry with individualized biomechanical properties and dynamic response characteristics, thereby obtaining a form- and property-integrated skeletal digital twin.
[0037] The computational-measurement fusion simulation module is used to build a computational-measurement fusion simulation engine, which integrates finite element analysis, multibody dynamics calculation and machine learning-driven surrogate models to obtain measured data of the implant in situ state, and fuses and dynamically calibrates the physics-based numerical calculation results with the measured data of the implant in situ state; so as to realize the real-time simulation and prediction of the mechanical performance of implanted devices under complex physiological environments.
[0038] The 3D visualization feedback module is used to construct a virtual assembly environment containing the medical device to be validated and the individualized skeleton. Driven by the physiological load spectrum, it uses a 3D visualization engine to dynamically render the spatial position and interaction state of the two, and presents key biomechanical indicators intuitively and quantitatively in the form of engineering cloud maps.
[0039] Compared with the prior art, this application has at least the following beneficial effects:
[0040] Based on further analysis and research of existing technical problems, this application recognizes that existing technologies struggle to balance modeling accuracy and computational efficiency in medical device R&D simulation, leading to long R&D cycles and delayed exposure of design flaws. This application addresses this issue by constructing a form-and-physical integrated digital twin of the skeleton that incorporates individualized biomechanical properties and dynamic response characteristics. Furthermore, it integrates finite element analysis, multibody dynamics calculations, and machine learning proxy models to build a computational-measurement fusion simulation engine. This engine fuses physical calculation results with measured data and performs dynamic calibration, achieving a unification of high-fidelity modeling and efficient simulation. This enables real-time, high-confidence simulation prediction of the mechanical performance of implantable devices under complex physiological environments in the early stages of R&D, significantly shortening the R&D cycle, allowing for early exposure and rapid correction of design flaws, and reducing R&D risks and costs.
[0041] Specifically:
[0042] (1) This application constructs a morphologically integrated digital twin of the skeleton, reconstructs the three-dimensional anatomical geometry of the skeleton based on the patient's medical imaging data, and integrates real-time sensor information and physiological load spectrum to give the three-dimensional anatomical geometry individualized biomechanical properties and dynamic response characteristics. This overcomes the shortcomings of traditional homogenized models in representing individual anatomical variations, significantly improves the matching degree between the simulation model and the real patient's anatomical structure and biomechanical properties, and lays a solid foundation for high-confidence simulation.
[0043] (2) This application constructs a computational-measurement fusion simulation engine, integrating finite element analysis, multibody dynamics calculation, and machine learning-driven surrogate models. It also fuses and dynamically calibrates the physics-based numerical calculation results with measured data of the implant's in-situ state, achieving an organic unity of high-precision physical modeling and high-efficiency numerical calculation. On the one hand, finite element analysis and multibody dynamics calculation ensure the physical authenticity of the simulation results; on the other hand, the machine learning-driven surrogate model can significantly improve calculation speed while maintaining accuracy, supporting real-time or near-real-time simulation and prediction. Simultaneously, dynamic calibration of simulation parameters using measured data further ensures the consistency between simulation results and actual physiological responses.
[0044] (3) This application constructs a virtual assembly environment containing the medical device to be verified and a personalized skeleton. Driven by the physiological load spectrum, it uses a three-dimensional visualization engine to dynamically render the spatial position and interaction state of the two, and presents key biomechanical indicators in an intuitive and quantitative manner in the form of engineering cloud map. This provides clear and operable decision support for R&D engineers and clinical experts, and significantly shortens the cycle of design scheme evaluation and iteration.
[0045] In summary, this application enables high-confidence and high-efficiency simulation prediction of the mechanical properties of implantable devices under complex physiological environments in the early stages of R&D, allowing design flaws to be exposed and corrected in advance. This significantly shortens the R&D cycle of orthopedic medical devices, reduces reliance on physical prototype testing and animal experiments, and reduces R&D costs and risks, providing strong support for the rapid iteration and innovation of orthopedic medical devices. Attached Figure Description
[0046] Figure 1 This is a flowchart illustrating a digital twin simulation method for orthopedic medical devices, provided as an embodiment of this application.
[0047] Figure 2 This is a flowchart illustrating a digital twin simulation method for orthopedic medical devices, provided as another embodiment of this application. Detailed Implementation
[0048] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.
[0049] In the description of this application, unless otherwise stated, the terms "including", "comprising", "having", etc., also mean "not limited to" (certain units, components, materials, steps, etc.).
[0050] In the technical solution of this application, the user information (including but not limited to user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, and displayed data) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of the relevant data all comply with the relevant laws, regulations, and standards of the relevant countries and regions, and necessary confidentiality measures have been taken. They do not violate public order and good morals, and corresponding operation entry points are provided for users to choose to authorize or refuse.
[0051] The core technologies in traditional orthopedic medical device R&D mainly revolve around materials science, mechanical engineering, and large-scale manufacturing, aiming to develop safe, effective, reliable, and standardized implants and devices. Traditional product design and development primarily rely on physical prototype testing, animal experiments, and limited in vitro simulations. Developing joint prostheses requires designing standardized prostheses of different sizes based on extensive human anatomical data, allowing for the intraoperative combination of components of different sizes to accommodate individual differences and improve surgical flexibility. Developing spinal internal fixation systems requires simulating rigid fixation using pedicle screws, connecting rods, and plates on human spinal skeleton models or experimental animals, based on the unique anatomical structure and three-dimensional kinematics of each segment, as well as the proximity of nerves and blood vessels, to create a stable environment for spinal fusion. Extensive simulation experiments are then conducted before clinical application. Trauma implants, such as bone plates, intramedullary nails, and screws, require understanding the mechanical requirements at different stages, the anatomical morphology of different bones, the ratio of cortical to cancellous bone, and blood supply characteristics. Researchers also need a deep understanding of the pathways of surrounding nerves, blood vessels, and tendons to avoid iatrogenic injuries.
[0052] In current orthopedic medical device R&D practices, traditional methods have long relied on physical prototype testing, animal experiments, and limited in vitro simulated operating conditions, resulting in systemic bottlenecks such as lengthy development cycles, high resource consumption, and low iteration efficiency. Particularly in the biomechanical verification stage of implant-bone interaction, existing simulation tools often fail to accurately reproduce individualized anatomical differences, dynamic physiological loads, and the micromechanical behavior of the bone-implant interface. This means design flaws are often only exposed in later stages of testing or even clinical application, significantly increasing R&D risks and costs. Orthopedic medical devices, especially spinal implants involving complex biomechanical environments, heavily rely on in vitro mechanical testing, animal experiments, and ultimately, human clinical trials for safety and efficacy verification. This model has several fundamental flaws: physical testing, limited by sample accessibility, ethical constraints, and high costs, can only cover extremely limited, pre-defined static or simplified operating conditions. The human physiological environment is inherently dynamic, variable, and highly individualized, encompassing a complex biomechanical load spectrum including daily activities, unexpected loads, and even pathological conditions. Traditional testing methods cannot exhaustively cover diverse physiological conditions such as multiaxial composite loads, long-term cyclic loads, and unique stress distributions under individualized anatomical variations. This results in poor extrapolation of test results, making it difficult to fully reflect the long-term performance and failure risk of implants in real human environments. Furthermore, physical testing cycles are lengthy, with each stage—from prototype fabrication and fixture design to test execution and data analysis—being time-consuming and labor-intensive, significantly hindering overall R&D progress and failing to meet market demands for rapid iterative innovation.
[0053] This invention belongs to the field of digital twin technology and medical device research and development, specifically involving a digital twin research and development platform for orthopedic medical devices based on multi-scale modeling, computational measurement fusion and three-dimensional visualization feedback, which is used to realize rapid virtual iteration, early safety verification and biomechanical performance prediction of orthopedic medical devices.
[0054] The core of the digital twin orthopedic medical device intelligent R&D acceleration platform lies in constructing a high-precision, customizable, and dynamically responsive digital twin model of the skeletal system. Unlike traditional static computer models or simplified simulations, digital twins emphasize real-time, bidirectional dynamic mapping and interaction between the virtual model and the physical entity. In this invention, the skeletal digital twin is not merely a digital reproduction of anatomical geometry, but deeply integrates the biomechanical properties, physiological load spectrum, and time-varying characteristics of individualized bones, forming a virtual organ capable of simulating the structure-function relationship of real bone tissue under dynamic physiological conditions. Its core value lies in constructing a highly realistic virtual experimental field, enabling in-depth insight and prediction of the performance and safety of orthopedic medical devices before the physical prototype manufacturing and biological experiments.
[0055] In one embodiment, combined Figure 1 and Figure 2A digital twin simulation method for orthopedic medical devices is provided, including:
[0056] Constructing a skeletal digital twin: Reconstructing the three-dimensional anatomical geometry of the skeleton based on the patient's medical imaging data, and integrating real-time sensor information and physiological load spectrum, endowing the three-dimensional anatomical geometry with individualized biomechanical properties and dynamic response characteristics, thereby obtaining a skeletal digital twin that integrates form and function;
[0057] Fusion simulation of computation and measurement: Construct a simulation engine that integrates finite element analysis, multibody dynamics calculation and machine learning-driven proxy model to obtain measured data of implant in-situ state, and fuse and dynamically calibrate the physics-based numerical calculation results with the measured data of implant in-situ state; so as to realize real-time simulation and prediction of the mechanical performance of implanted devices under complex physiological environment.
[0058] 3D visualization feedback: Construct a virtual assembly environment containing the medical device to be validated and a personalized skeleton. Driven by the physiological load spectrum, use a 3D visualization engine to dynamically render the spatial position and interaction state of the two, and present key biomechanical indicators intuitively and quantitatively in the form of engineering cloud maps.
[0059] In one embodiment, a digital twin simulation method for orthopedic medical devices is provided. This method achieves rapid and accurate simulation prediction and verification of the mechanical properties of orthopedic medical devices under complex physiological environments by constructing a high-fidelity digital twin of the skeleton, integrating physical calculations and measured data into a computational-measurement fusion simulation engine, and employing a three-dimensional visualization feedback mechanism. Specifically, the method includes the following steps:
[0060] Step S1: Construct a skeletal digital twin.
[0061] This step aims to establish a virtual model that can realistically reflect the individualized skeletal geometry and biomechanical properties, providing a high-fidelity foundation for subsequent simulations.
[0062] Specifically, the patient's medical imaging data, such as computed tomography (CT) or magnetic resonance imaging (MRI), is first acquired. This imaging data is then segmented and reconstructed in three dimensions to obtain the precise three-dimensional anatomical geometry of the target bone. This reconstruction process restores the macroscopic morphology of the bone, including the distribution areas of cortical and cancellous bone.
[0063] Building upon this foundation, real-time sensor information and physiological load spectra are further integrated. Real-time sensor information, derived from wearable devices or intraoperative sensors, acquires an individual's actual mechanical response under specific activity states. The physiological load spectra are pre-constructed, encompassing daily activities (such as walking, running, and climbing stairs), accidental loads (such as fall impacts), and multiaxial composite loads and long-term cyclic loads under pathological conditions. By fusing these data, individualized biomechanical properties (such as material parameter gradients like elastic modulus and Poisson's ratio determined based on bone density distribution) and dynamic response characteristics (such as deformation and stress response under different load conditions) are assigned to the three-dimensional anatomical geometry, thereby constructing a form- and property-integrated digital twin of the skeleton.
[0064] This digital twin not only reproduces the static anatomical morphology of the skeleton, but also has the ability to simulate its structure-function relationship under dynamic physiological conditions, overcoming the shortcomings of traditional homogenized models in representing individual anatomical variations.
[0065] Step S2: Calculation and measurement fusion simulation.
[0066] This step aims to achieve high-precision, high-efficiency biomechanical simulation, supporting real-time prediction of implantable device performance.
[0067] First, a computational-measurement fusion simulation engine was constructed. This engine integrates multiple computational methods: finite element analysis (FEA) is used to accurately calculate the stress and strain distribution inside the bone and implant; multibody dynamics analysis is used to simulate load transfer and relative motion during joint movement; and a machine learning-driven surrogate model acts as an accelerator, learning from a large number of finite element simulation results to achieve near real-time computational response while maintaining accuracy.
[0068] During the simulation, the engine acquires measured data on the implant's in-situ state. This measured data can come from in vitro biomechanical tests (such as contact pressure and strain data measured by a mechanical testing machine) or limited in vivo monitoring (such as real-time data collected by implant sensors in animal experiments or clinical trials).
[0069] Subsequently, the engine integrates and dynamically calibrates the physics-based numerical calculation results (i.e., simulation results obtained from finite element analysis or multibody dynamics calculations) with the aforementioned measured data. Specifically, by comparing the simulation results with the measured data, it inversely optimizes key input parameters in the simulation, such as boundary conditions, contact friction coefficient, and bone material constitutive model parameters, minimizing the deviation between the simulation results and the actual physiological response. This dynamic calibration mechanism ensures the high confidence level of the simulation model.
[0070] Through the aforementioned calculation and measurement fusion mechanism, the engine can perform real-time simulation and prediction of the mechanical properties of implanted devices under complex physiological environments, including key indicators such as contact pressure, micro-displacement, stress shielding effect, stress distribution and fatigue life of the implant itself.
[0071] Step S3: 3D visualization feedback.
[0072] This step aims to present the simulation results to R&D personnel in an intuitive and interactive way, providing a clear engineering basis for design decisions.
[0073] First, a virtual assembly environment is constructed that includes the medical device to be validated and the individualized bone. The 3D model of the orthopedic medical device to be validated (such as a spinal fixation system, joint prosthesis, or trauma implant) is precisely virtually assembled with the individualized bone digital twin constructed in step S1 to simulate its spatial position and fixation method in the body.
[0074] Driven by the physiological load spectrum, a 3D visualization engine (such as a professional rendering engine based on OpenGL or Unity) dynamically renders the spatial position and interaction state of the medical device and the bone. During the rendering process, the system can intuitively and quantitatively present key biomechanical indicators such as stress concentration areas, strain energy density distribution, and micromotion amplitude within the bone and implant in the form of a color engineering cloud map. For example, areas with high stress are highlighted in red, while areas with low stress are highlighted in blue, forming a clear visual contrast.
[0075] Furthermore, this visualization feedback system supports interactive operation. Users can rotate and zoom the model, use the profile tool to view internal details, or click on a specific area to view the numerical values of that area's biomechanical properties and their changes over time. The system also displays key performance parameters in real time in the form of digital panels, dynamic curves, or dashboards, providing users with multi-dimensional decision support.
[0076] Through the synergistic effect of the above three steps, the method provided in this embodiment can perform high-confidence and high-efficiency simulation prediction of the biomechanical properties of orthopedic medical devices in the early stage of research and development, so as to expose and quickly correct design defects in advance, thereby significantly shortening the research and development cycle and reducing research and development risks and costs.
[0077] In one embodiment, the step of constructing a skeletal digital twin further includes:
[0078] Based on the patient's medical imaging data, the three-dimensional anatomical geometry of the target bone is reconstructed;
[0079] Based on individual bone density distribution, determine the material property gradient between cortical bone and cancellous bone in the target bone;
[0080] By assigning the material property gradient to the three-dimensional anatomical geometry, a skeletal digital twin that conforms to the biomechanical characteristics of the individual is obtained, thus overcoming the shortcomings of traditional homogenized models in representing individual anatomical variations.
[0081] Multi-scale modeling technology: Based on patient medical imaging data, it integrates real-time sensor information and physiological load spectrum to construct a high-fidelity, dynamically responsive, shape-and-structure integrated digital twin of the skeleton. It accurately simulates individualized characteristics from static anatomical structure to dynamic biomechanical behavior, precisely reconstructs the three-dimensional geometric morphology of the target bone, and endows it with biomechanical properties that conform to the individual bone density distribution and the gradient of cortical / cancellous bone material properties, overcoming the shortcomings of traditional homogenized models in representing individual anatomical variations.
[0082] Effects: The digital twin display system includes real-time status display of surgical instruments, real-time rendering of a skeletal digital twin, and real-time display of the main mechanical properties of orthopedic medical devices. It accurately and clearly displays the status of surgical instruments through a virtual 3D model. The real-time rendering of the skeletal digital twin uses computer graphics technology to construct a high-fidelity virtual 3D skeletal model, and then uses computer rendering technology to render the simulation results of the medical device's function onto the skeletal 3D model. The real-time display of the main mechanical properties of the bone is achieved through charts and scale lines, displaying the values of various biomechanical properties of the medical device during the research and development process.
[0083] In one embodiment, a data preprocessing and fusion step is included before the computational-measurement fusion simulation step, including:
[0084] Collect raw monitoring data generated by various sensors on surgical instruments and transmit it to the computing terminal;
[0085] The original monitoring data is cleaned to remove duplicate, redundant and abnormal data, and the cleaned original monitoring data is sorted and formatted according to the dependency of the numerical calculation process.
[0086] Based on biomechanical algorithms, the processed data is solved and analyzed to deduce the load distribution and mechanical constraints of medical devices.
[0087] The cleaned sensor data is integrated and fused with the load distribution and mechanical constraints to form a unified data set, providing input for subsequent simulation and prediction.
[0088] In one embodiment, the computational-measurement fusion simulation step further includes: using contact pressure and strain field data obtained from in vitro biomechanical experiments or limited in vivo monitoring to continuously optimize the boundary conditions, contact friction coefficient, and bone material constitutive model parameters in the simulation online, so as to minimize the deviation between the simulation results and the actual physiological response.
[0089] The computational-measurement fusion simulation engine integrates finite element analysis, multibody dynamics calculations, and machine learning-driven proxy models to achieve real-time simulation and prediction of the mechanical performance of implanted devices under complex physiological environments. This significantly improves the realism and efficiency of the simulation. The engine does not simply rely on mechanistic models based on physical laws; instead, it deeply integrates and dynamically calibrates numerical calculations based on nonlinear finite element analysis and multibody dynamics with real-time or historical measured data of the implant's in-situ state. After importing implant design parameters, including geometry, materials, surface conditions, and target physiological load conditions, the engine performs real-time or near-real-time biomechanical simulations in a virtual environment. It accurately calculates the contact pressure, micro-displacement, stress shielding, and other microscopic interactions between the orthopedic medical device and bone tissue, the stress distribution within the bone tissue, and the mechanical response of the implant itself.
[0090] Effects: The data association system comprises a data transmission system, a data processing system, a numerical calculation system, and a data fusion system connected in sequence. The data transmission system collects and aggregates information from various sensors on surgical instruments, establishing a connection with a computer via wireless communication methods such as Bluetooth to achieve real-time transmission of sensor data to the data processing system. After receiving the raw data, the data processing system first cleans the data, removing duplicate, redundant, and erroneous data to ensure data quality and the reliability of subsequent processing. Then, based on the dependencies and computational complexity of the numerical calculation process, it sorts and preprocesses the valid data, providing efficient and standardized input to the numerical calculation system. The numerical calculation system, based on biomechanical algorithms, solves and analyzes the processed data to derive the load distribution and mechanical constraints of the medical instruments during surgery. The data fusion system integrates and fuses the cleaned sensor data with the mechanical data output from the numerical calculation system, forming a unified, multi-dimensional dataset to provide accurate and complete input data support for downstream simulation and prediction systems.
[0091] In one embodiment, the step of providing 3D visualization feedback further includes:
[0092] It dynamically displays the stress, strain distribution, and strain energy density inside bones and implants in the form of color cloud maps, displays the displacement field and interface micro-motion direction in the form of vector maps, and supports viewing internal details through interactive cutting, perspective, or cross-sectional viewing methods.
[0093] In the form of digital panels, dynamic curves or dashboards, key biomechanical performance parameters such as contact pressure, micro-motion amplitude, osseointegration potential, long-term fatigue life and stress shielding effect are displayed in real time in a quantitative manner.
[0094] Automatically generate a structured simulation verification report containing the evaluation conclusions of biomechanical safety and effectiveness based on the visualization rendering results and quantization parameters.
[0095] In one embodiment, the three-dimensional visualization feedback step further includes:
[0096] In response to the user's interaction operation on a specific area in the engineering cloud map, dynamically display the biomechanical performance values corresponding to that area and their change curves over time;
[0097] Simultaneously display the time history curve of the physiological load spectrum corresponding to the current simulation condition, and support the user to customize and modify the load spectrum to trigger a resimulation.
[0098] Three-dimensional visualization feedback: Through a professional three-dimensional visualization engine, the platform dynamically renders the spatial positions and interaction states of orthopedic medical devices in the individualized bones, and intuitively and quantitatively presents key biomechanical indicators such as stress concentration areas, strain energy density distributions, and micro-motion amplitudes at the bone-medical device interface in the form of engineering cloud maps, providing clear and operable engineering basis for design decisions. These functions together constitute an intelligent environment that supports rapid virtual iteration of design schemes and early safety verification.
[0099] Effect: Virtually assemble the three-dimensional model of the medical device to be verified with the above individualized bone model. Under the drive of the imported physiological load spectrum, calculate the biomechanical responses at the interaction interface between the orthopedic medical device and the bone. Using data such as contact pressure and strain fields obtained from ex vivo biomechanical tests or limited in vivo monitoring, continuously optimize the boundary conditions, contact friction coefficients, bone material constitutive model parameters, etc. in the simulation online to ensure that the simulation results have a high degree of confidence with the real physiological responses. This platform can quickly predict and perform sensitivity analysis on key long-term performance indicators such as the bone integration potential, long-term fatigue life, stress shielding effect, and interface fretting wear of orthopedic medical devices within seconds. R & D personnel can quickly explore and perform multi-objective optimization within a large design space based on this surrogate model, and automatically find the optimal design scheme of orthopedic medical devices with better performance. Finally, through the three-dimensional visualization rendering engine, the optimized design scheme and its biomechanical simulation results in the individualized bone model are visualized with high fidelity. The system can dynamically and intuitively display the stress and strain distributions inside the bone and the implant in the form of color cloud maps, display the displacement and micro-motion directions in vector diagrams, and can view the internal details through section tools. At the same time, all key biomechanical performance parameters are displayed in real time in the form of numbers, curve graphs or dashboards. This visualization system provides intuitive decision support for R & D engineers and clinical experts, facilitating the final evaluation and confirmation of the biomechanical safety and effectiveness of the design scheme, and generating a structured simulation verification report.
[0100] In one embodiment, a long-cycle performance prediction step is also included:
[0101] Based on the proxy model generated by the computational-measurement fusion simulation engine, at least one parameter among the following in orthopedic medical devices—osseointegration potential, long-term fatigue life, stress shielding effect, and interface fretting wear—can be rapidly predicted and analyzed for sensitivity within a preset short time threshold:
[0102] In another embodiment, the method further includes a long-cycle performance prediction step.
[0103] Specifically, based on the proxy model pre-generated by the computational-measurement fusion simulation engine, the system can quickly predict and perform sensitivity analysis on at least one parameter among the osseointegration potential, long-term fatigue life, stress shielding effect, and interface fretting wear of orthopedic medical devices within a preset short time threshold (e.g., seconds or milliseconds).
[0104] The surrogate model is a lightweight computational model trained on a large number of finite element simulation results using machine learning algorithms, which can significantly improve computational speed while maintaining accuracy. Sensitivity analysis is used to evaluate the impact of various design parameters (such as geometric dimensions, material properties, surface coatings, etc.) on the above performance indicators, identify key influencing factors, and provide guidance for subsequent design optimization.
[0105] In one embodiment, a design space exploration and optimization step is also included:
[0106] Based on the surrogate model, a multi-objective optimization algorithm is executed in the design space containing geometric parameters, material parameters and surface feature parameters to automatically find the optimal design scheme of orthopedic medical devices that meets the biomechanical safety constraints.
[0107] The optimization objective of the multi-objective optimization algorithm includes at least one of the following: minimizing stress shielding effect and interface fretting wear, maximizing osseointegration potential and long-term fatigue life.
[0108] The constraints include the physiological load-bearing threshold of bone and / or the strength limit of implant materials.
[0109] In another embodiment, the method further includes a design space exploration and optimization step.
[0110] This step is based on a proxy model generated by the computational-test-simulation engine. Because the proxy model has extremely high computational efficiency, it can complete the performance evaluation of a large number of design schemes in a short time, thus providing a feasible basis for design space exploration.
[0111] Specifically, the system executes a multi-objective optimization algorithm within a pre-defined design space. This design space includes variable parameters of the medical device to be optimized, including geometric parameters (such as implant size, curvature, and thickness), material parameters (such as elastic modulus and Poisson's ratio), and surface characteristic parameters (such as coating type and porosity). The multi-objective optimization algorithm aims to balance multiple potentially conflicting performance indicators, such as minimizing stress shielding effects and interfacial fretting wear, and maximizing osseointegration potential and long-term fatigue life. Simultaneously, the optimization process must meet pre-defined constraints, including the bone physiological load-bearing threshold (ensuring bone does not suffer damage due to excessive stress) and the implant material strength limit (ensuring implant does not experience fatigue failure), thereby ensuring the safety and effectiveness of the optimization scheme.
[0112] Through the above optimization steps, the system can automatically find the optimal design scheme of medical devices that meets biomechanical safety constraints, replacing the traditional experience-based manual trial and error process, and significantly improving design efficiency and scheme quality.
[0113] In one embodiment, the optimization scheme verification and feedback steps are also included:
[0114] The optimized design scheme is re-imported into the virtual assembly environment, and the above-mentioned calculation and measurement fusion simulation and three-dimensional visualization feedback steps are executed.
[0115] The key biomechanical indicators before and after optimization were compared and analyzed, and the optimization effect was presented in a visual form.
[0116] In another embodiment, the method further includes an optimization scheme verification and feedback step.
[0117] This step aims to perform closed-loop verification of the design scheme obtained from the aforementioned optimization steps, ensuring the effectiveness and reliability of the optimization results.
[0118] Specifically, the optimized design scheme is first re-imported into the virtual assembly environment. The system automatically assembles the optimized 3D model of the medical device with the individualized digital twin of the skeleton, maintaining the same assembly relationship and boundary conditions as the initial scheme. Subsequently, the aforementioned computational-measurement fusion simulation and 3D visualization feedback steps are executed, and under the same physiological load spectrum, the various biomechanical performance indicators of the optimized scheme are recalculated.
[0119] Based on this, the system will conduct a comparative analysis of key biomechanical indicators before and after optimization. The comparative analysis indicators include, but are not limited to, the degree of stress shielding effect, the amplitude of interface micromotion, the osseointegration potential score, and the predicted long-term fatigue life. The comparison results are presented in a visual format, such as through side-by-side or overlaid engineering cloud maps to intuitively compare stress distribution changes, through bar charts or line graphs to show the improvement of each indicator's quantitative value, or through tables summarizing the changes in key parameters before and after optimization.
[0120] Through the verification and feedback steps described above, R&D personnel can intuitively confirm the effectiveness of the optimization scheme, forming a complete closed loop from "modeling—simulation—optimization—verification". This step not only enhances the credibility of the optimization results but also provides quantitative evidence for subsequent design iterations, further improving R&D efficiency and design quality.
[0121] In one embodiment, a digital twin simulation system for orthopedic medical devices is also provided for performing the method of the above embodiments, comprising:
[0122] A skeletal digital twin module is constructed to reconstruct the three-dimensional anatomical geometry of the skeleton based on the patient's medical imaging data. Real-time sensor information and physiological load spectrum are integrated to endow the three-dimensional anatomical geometry with individualized biomechanical properties and dynamic response characteristics, thereby obtaining a form- and property-integrated skeletal digital twin.
[0123] The computational-measurement fusion simulation module is used to build a computational-measurement fusion simulation engine, which integrates finite element analysis, multibody dynamics calculation and machine learning-driven surrogate models to obtain measured data of the implant in situ state, and fuses and dynamically calibrates the physics-based numerical calculation results with the measured data of the implant in situ state; so as to realize the real-time simulation and prediction of the mechanical performance of implanted devices under complex physiological environments.
[0124] The 3D visualization feedback module is used to construct a virtual assembly environment containing the medical device to be validated and the individualized skeleton. Driven by the physiological load spectrum, it uses a 3D visualization engine to dynamically render the spatial position and interaction state of the two, and presents key biomechanical indicators intuitively and quantitatively in the form of engineering cloud maps.
[0125] In one embodiment, the system further includes a data processing module for performing the data preprocessing and fusion steps described above.
[0126] In one embodiment, the system further includes an optimization calculation module for performing the aforementioned long-cycle performance prediction and design space exploration optimization steps.
[0127] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
Claims
1. A digital twin simulation method for orthopedic medical devices, characterized in that, include: Constructing a skeletal digital twin: Reconstructing the three-dimensional anatomical geometry of the skeleton based on the patient's medical imaging data, and integrating real-time sensor information and physiological load spectrum, endowing the three-dimensional anatomical geometry with individualized biomechanical properties and dynamic response characteristics, thereby obtaining a skeletal digital twin that integrates form and function; Fusion simulation of computation and measurement: Construct a simulation engine that integrates finite element analysis, multibody dynamics calculation and machine learning-driven proxy model to obtain measured data of implant in-situ state, and fuse and dynamically calibrate the physics-based numerical calculation results with the measured data of implant in-situ state; so as to realize real-time simulation and prediction of the mechanical performance of implanted devices under complex physiological environment. 3D visualization feedback: Construct a virtual assembly environment containing the medical device to be validated and a personalized skeleton. Driven by the physiological load spectrum, use a 3D visualization engine to dynamically render the spatial position and interaction state of the two, and present key biomechanical indicators intuitively and quantitatively in the form of engineering cloud maps.
2. The method according to claim 1, characterized in that, The step of constructing a skeletal digital twin further includes: Based on the patient's medical imaging data, the three-dimensional anatomical geometry of the target bone is reconstructed; Based on individual bone density distribution, determine the material property gradient between cortical bone and cancellous bone in the target bone; By assigning the material property gradient to the three-dimensional anatomical geometry, a skeletal digital twin that conforms to the biomechanical characteristics of the individual is obtained, thus overcoming the shortcomings of traditional homogenized models in representing individual anatomical variations.
3. The method according to claim 1, characterized in that, The computational-measurement fusion simulation step is preceded by a data preprocessing and fusion step, including: Collect raw monitoring data generated by various sensors on surgical instruments and transmit it to the computing terminal; The original monitoring data is cleaned to remove duplicate, redundant and abnormal data, and the cleaned original monitoring data is sorted and formatted according to the dependency of the numerical calculation process. Based on biomechanical algorithms, the processed data is solved and analyzed to deduce the load distribution and mechanical constraints of medical devices. The cleaned sensor data is integrated and fused with the load distribution and mechanical constraints to form a unified data set, providing input for subsequent simulation and prediction.
4. The method according to claim 1, characterized in that, The steps of the 3D visualization feedback further include: It dynamically displays the stress, strain distribution, and strain energy density inside bones and implants in the form of color cloud maps, displays the displacement field and interface micro-motion direction in the form of vector maps, and supports viewing internal details through interactive cutting, perspective, or cross-sectional viewing methods. In the form of digital panels, dynamic curves or dashboards, key biomechanical performance parameters such as contact pressure, micro-motion amplitude, osseointegration potential, long-term fatigue life and stress shielding effect are displayed in real time in a quantitative manner. Based on the visualization rendering results and quantitative parameters, a structured simulation verification report containing biomechanical safety and effectiveness assessment conclusions is automatically generated.
5. The method according to claim 1, characterized in that, The computational-measurement fusion simulation step also includes: using contact pressure and strain field data obtained from in vitro biomechanical experiments or limited in vivo monitoring to continuously optimize the boundary conditions, contact friction coefficient, and bone material constitutive model parameters in the simulation online, so as to minimize the deviation between the simulation results and the actual physiological response.
6. The method according to claim 1, characterized in that, It also includes long-cycle performance prediction steps: Based on the proxy model generated by the computational-measurement fusion simulation engine, at least one parameter among the following in orthopedic medical devices—osseointegration potential, long-term fatigue life, stress shielding effect, and interface fretting wear—can be rapidly predicted and analyzed for sensitivity within a preset short time threshold:
7. The method according to claim 6, characterized in that, It also includes design space exploration and optimization steps: Based on the surrogate model, a multi-objective optimization algorithm is executed in the design space containing geometric parameters, material parameters and surface feature parameters to automatically find the optimal design scheme of orthopedic medical devices that meets the biomechanical safety constraints. The optimization objective of the multi-objective optimization algorithm includes at least one of the following: minimizing stress shielding effect and interface fretting wear, maximizing osseointegration potential and long-term fatigue life. The constraints include the physiological load-bearing threshold of bone and / or the strength limit of implant materials.
8. The method according to claim 7, characterized in that, It also includes the steps of optimizing the solution verification and feedback: The optimized design scheme is re-imported into the virtual assembly environment, and the calculation and measurement fusion simulation and three-dimensional visualization feedback steps described in claim 1 are executed. The key biomechanical indicators before and after optimization were compared and analyzed, and the optimization effect was presented in a visual form.
9. The method according to claim 1, characterized in that, The 3D visualization feedback step further includes: In response to user interaction with a specific area in the engineering cloud map, the corresponding biomechanical performance values and their time-varying curves for that area are dynamically displayed. The system synchronously displays the physiological load spectrum time history curve corresponding to the current simulation condition and supports user-defined modification of the load spectrum to trigger a resimulation.
10. A digital twin simulation system for orthopedic medical devices, used to perform the method according to any one of claims 1-9, characterized in that, include: A skeletal digital twin module is constructed to reconstruct the three-dimensional anatomical geometry of the skeleton based on the patient's medical imaging data. Real-time sensor information and physiological load spectrum are integrated to endow the three-dimensional anatomical geometry with individualized biomechanical properties and dynamic response characteristics, thereby obtaining a form- and property-integrated skeletal digital twin. The computational-measurement fusion simulation module is used to build a computational-measurement fusion simulation engine, which integrates finite element analysis, multibody dynamics calculation and machine learning-driven surrogate models to obtain measured data of the implant in situ state, and fuses and dynamically calibrates the physics-based numerical calculation results with the measured data of the implant in situ state; so as to realize the real-time simulation and prediction of the mechanical performance of implanted devices under complex physiological environments. The 3D visualization feedback module is used to construct a virtual assembly environment containing the medical device to be validated and the individualized skeleton. Driven by the physiological load spectrum, it uses a 3D visualization engine to dynamically render the spatial position and interaction state of the two, and presents key biomechanical indicators intuitively and quantitatively in the form of engineering cloud maps.