CN-121971204-A - Digital design and manufacturing method and system of knee joint tibial prosthesis

CN121971204ACN 121971204 ACN121971204 ACN 121971204ACN-121971204-A

Abstract

The invention discloses a digital design and manufacturing method and system of a knee joint tibial prosthesis, wherein the method comprises the steps of collecting stress distribution and motion tracks of a knee joint in different motion states according to gait analysis data and dynamic mechanical sensor data of a patient to obtain a dynamic mechanical feature set of the knee joint, inputting the dynamic mechanical feature set into a geometric design model based on a generation countermeasure network to generate an initial tibial prosthesis geometric shape to obtain a preliminarily optimized tibial prosthesis geometric model, designing an internal porous structure of the prosthesis according to the tibial prosthesis geometric model by utilizing a material microstructure optimization algorithm based on deep reinforcement learning to obtain a material and tibial prosthesis design scheme, and inputting the tibial prosthesis design scheme into multi-axis additive manufacturing equipment to obtain a final knee joint tibial prosthesis. By utilizing the embodiment of the invention, the precision and consistency of the manufacturing of the prosthesis can be ensured, and the overall quality of the prosthesis can be further improved.

Inventors

ZHAO LIANG
Liao Zheting
ZHAO SHIQI
WANG ZIQI
CHEN YUFAN
WU DESHENG
Qiu Gengtao

Assignees

广州医科大学附属第一医院（广州呼吸中心）

Dates

Publication Date: 20260505
Application Date: 20260123

Claims (10)

1. A method of digitally designing and manufacturing a knee tibial prosthesis, the method comprising: According to gait analysis data and dynamic mechanical sensor data of a patient, stress distribution and motion tracks of the knee joint in different motion states are collected, and a dynamic mechanical feature extraction algorithm based on time sequence decomposition is adopted to separate key stress modes and motion features of the tibia platform, so that a dynamic mechanical feature set of the knee joint is obtained; Inputting the dynamic mechanical feature set into a geometric design model based on a generation countermeasure network to generate an initial tibial prosthesis geometric shape, wherein the geometric design model optimizes the curvature, thickness and contact surface shape of the prosthesis by combining patient anatomical data and biomechanical constraint conditions to obtain a preliminarily optimized tibial prosthesis geometric model; According to the geometrical model of the tibial prosthesis, an internal porous structure of the prosthesis is designed by utilizing a material microstructure optimization algorithm based on deep reinforcement learning, wherein the material microstructure optimization algorithm generates microstructure distribution with gradient porosity and mechanical property by combining mechanical property simulation and multi-objective optimization, and a material and structure integrated tibial prosthesis design scheme is obtained; Inputting the design scheme of the tibial prosthesis into multi-axis additive manufacturing equipment, and adopting an electron beam melting technology to manufacture the prosthesis, wherein the manufacturing process ensures the precision and consistency of the manufacturing of the prosthesis through a real-time deformation monitoring and self-adaptive path planning technology, and after the manufacturing is finished, adopting a dynamic function verification method based on a biomechanical test platform to simulate the daily movement load of the knee joint, verifying the functionality and durability of the prosthesis and obtaining the final knee joint tibial prosthesis.
2. The method according to claim 1, wherein the step of collecting stress distribution and motion trajectories of the knee joint in different motion states according to gait analysis data and dynamic mechanical sensor data of the patient, and separating key stress patterns and motion features of the tibial plateau by using a dynamic mechanical feature extraction algorithm based on time sequence decomposition to obtain a dynamic mechanical feature set of the knee joint comprises: According to gait analysis data and dynamic mechanical sensor data of a patient, a multisource data acquisition frame based on time stamp alignment is adopted to acquire stress distribution and motion trail of the knee joint in different motion states in real time, and time difference of multisource data is eliminated through a data synchronization algorithm to generate a multisource data set with time synchronization; For a time-synchronous multi-source data set, removing noise and abnormal values by adopting a data cleaning method based on a self-adaptive filtering algorithm, and enhancing the characteristic expression of a key stress mode and a motion track by a characteristic enhancement technology to generate a high-quality cleaned data set; The method comprises the steps of separating key stress modes and motion characteristics of a tibial plateau from a cleaned dataset by adopting a dynamic mechanical characteristic extraction algorithm based on variation modal decomposition, capturing key frequencies and amplitudes under different motion states by a self-adaptive modal number selection mechanism, and generating a preliminary dynamic mechanical mode decomposition result; And carrying out weighted fusion on the stress distribution characteristics and the motion track characteristics by adopting a characteristic fusion method based on a multi-head attention mechanism on the primary dynamic mechanical mode decomposition result, capturing the complex dependency relationship between the stress distribution and the motion track through a cross-mode attention mechanism, and generating a final knee joint dynamic mechanical characteristic set.
3. The method of claim 2, wherein the inputting the dynamic mechanical feature set into generating an initial tibial prosthesis geometry based on generating a geometric design model of an antagonism network, wherein the geometric design model optimizes curvature, thickness, and contact surface shape of the prosthesis by combining patient anatomical data and biomechanical constraints, resulting in a preliminarily optimized tibial prosthesis geometry model, comprising: according to the dynamic mechanical feature set, a geometric design model based on a generated countermeasure network is adopted, patient anatomy data and biomechanical constraint conditions are combined, a generator and a discriminator are initialized, the geometric shape generated by the generator is ensured to meet the requirements of anatomy and mechanics through a conditional constraint mechanism, and a preliminary prosthesis geometric shape is generated; For the preliminary prosthesis geometry, a geometry adjustment method based on a curvature optimization algorithm is adopted, the curvature and thickness of the prosthesis are optimized by combining the anatomical features of the tibial plateau and the stress distribution in the dynamic mechanical feature set, and the contact surface shape matching of the prosthesis and the bone is ensured through a self-adaptive curvature adjustment mechanism, so that the preliminary optimized geometry is generated; For the preliminarily optimized geometric shape, a contact surface shape optimization method based on contact mechanics analysis is adopted, the contact surface shape of the prosthesis is optimized by combining the motion track and the stress distribution in the dynamic mechanics characteristic set, and the mechanical property of the contact surface is ensured through a dynamic friction coefficient adjustment mechanism, so that the preliminarily optimized contact surface shape is generated; And (3) for the preliminarily optimized geometric shape and contact surface shape, adopting a verification method based on finite element analysis, combining a patient anatomic model and biomechanical constraint conditions, verifying the mechanical property and biocompatibility of the geometric model, and dynamically adjusting the geometric parameters through a feedback correction mechanism to generate a final preliminarily optimized tibial prosthesis geometric model.
4. The method of claim 3, wherein the designing the internal porous structure of the prosthesis according to the tibial prosthesis geometry model using a material microstructure optimization algorithm based on deep reinforcement learning, wherein the material microstructure optimization algorithm generates a microstructure distribution with gradient porosity and mechanical properties by combining mechanical property simulation and multi-objective optimization, resulting in a material and structure integrated tibial prosthesis design scheme, comprising: according to a tibia prosthesis geometric model, a material microstructure optimization algorithm based on deep reinforcement learning is adopted, the mechanical property requirement and the biocompatibility requirement of the prosthesis are combined, a microstructure design space is initialized, and a porosity, mechanical property and biocompatibility optimization target is defined through a multi-target optimization framework to generate a preliminary microstructure design space; For the preliminary microstructure design space, adopting an optimization method based on gradient porosity distribution, combining stress distribution and motion trail in dynamic mechanical feature set, dynamically adjusting the porosity distribution in the prosthesis, and generating microstructure distribution with gradient porosity through a self-adaptive gradient adjustment mechanism; For microstructure distribution with gradient porosity, adopting a mechanical property simulation method based on finite element analysis, combining a geometrical model and material properties of the prosthesis, simulating the mechanical properties of the prosthesis, dynamically adjusting microstructure parameters through a multi-objective optimization algorithm, and generating primarily optimized microstructure distribution; And (3) for the initially optimized microstructure distribution, adopting a design method based on material and structure integration, and combining the geometry and mechanical property simulation result of the prosthesis to generate a tibial prosthesis design scheme with material and structure integration.
5. The method of claim 4, wherein the inputting the tibial prosthesis design solution to a multi-axis additive manufacturing device uses an electron beam melting technique for manufacturing the prosthesis, wherein the manufacturing process ensures precision and consistency of manufacturing the prosthesis by a real-time deformation monitoring and adaptive path planning technique, and after manufacturing, the method uses a dynamic function verification method based on a biomechanical test platform to simulate daily movement load of the knee joint, verify functionality and durability of the prosthesis, and obtain the final knee joint tibial prosthesis, and comprises: According to the design scheme of the tibial prosthesis, a manufacturing parameter optimization method based on process simulation software is adopted, manufacturing parameters comprising electron beam power, scanning speed and layer thickness are optimized by combining the characteristics of an electron beam melting technology, an optimal manufacturing path is generated through a self-adaptive path planning algorithm, and the precision and consistency of prosthesis manufacturing are ensured; in the manufacturing process, a real-time deformation monitoring system based on an optical sensor is adopted to detect deformation and stress distribution of the prosthesis in the manufacturing process, and manufacturing parameters and path planning are dynamically adjusted through a feedback control mechanism, so that the precision and consistency of the manufacturing of the prosthesis are ensured, and a preliminary prosthesis entity is generated; performing preliminary treatment on the surface of the prosthesis by adopting a post-treatment method based on mechanical polishing to remove burrs and unevenness generated in the manufacturing process, and further optimizing the smoothness of the surface of the prosthesis by adopting a chemical polishing technology to generate the prosthesis with a smooth surface; and simulating daily movement load of the knee joint for the prosthesis with smooth surface by adopting a dynamic function verification method based on a biomechanical test platform, verifying the functionality and durability of the prosthesis, and generating the final knee joint tibial prosthesis through multiple rounds of cyclic test and data analysis.
6. A system for digitally designing and manufacturing a knee tibial prosthesis, the system comprising: The acquisition module is used for acquiring stress distribution and motion trail of the knee joint under different motion states according to gait analysis data and dynamic mechanical sensor data of a patient, and separating key stress modes and motion characteristics of the tibia platform by adopting a dynamic mechanical characteristic extraction algorithm based on time sequence decomposition to obtain a dynamic mechanical characteristic set of the knee joint; the generation module is used for inputting the dynamic mechanical feature set into a geometric design model based on a generation countermeasure network to generate an initial tibial prosthesis geometric shape, wherein the geometric design model optimizes the curvature, thickness and contact surface shape of the prosthesis by combining patient anatomical data and biomechanical constraint conditions to obtain a preliminarily optimized tibial prosthesis geometric model; The design module is used for designing an internal porous structure of the tibial prosthesis by utilizing a material microstructure optimization algorithm based on deep reinforcement learning according to the tibial prosthesis geometric model, wherein the material microstructure optimization algorithm generates microstructure distribution with gradient porosity and mechanical property by combining mechanical property simulation and multi-objective optimization, and a tibial prosthesis design scheme with integrated material and structure is obtained; The manufacturing module is used for inputting the design scheme of the tibial prosthesis into multi-axis additive manufacturing equipment, and manufacturing the prosthesis by adopting an electron beam melting technology, wherein the manufacturing process ensures the precision and consistency of the manufacturing of the prosthesis by adopting a real-time deformation monitoring and self-adaptive path planning technology, and the final knee joint tibial prosthesis is obtained by adopting a dynamic function verification method based on a biomechanical test platform to simulate the daily movement load of the knee joint and verify the functionality and durability of the prosthesis after the manufacturing is completed.
7. The system according to claim 6, wherein the acquisition module is specifically configured to: According to gait analysis data and dynamic mechanical sensor data of a patient, a multisource data acquisition frame based on time stamp alignment is adopted to acquire stress distribution and motion trail of the knee joint in different motion states in real time, and time difference of multisource data is eliminated through a data synchronization algorithm to generate a multisource data set with time synchronization; For a time-synchronous multi-source data set, removing noise and abnormal values by adopting a data cleaning method based on a self-adaptive filtering algorithm, and enhancing the characteristic expression of a key stress mode and a motion track by a characteristic enhancement technology to generate a high-quality cleaned data set; The method comprises the steps of separating key stress modes and motion characteristics of a tibial plateau from a cleaned dataset by adopting a dynamic mechanical characteristic extraction algorithm based on variation modal decomposition, capturing key frequencies and amplitudes under different motion states by a self-adaptive modal number selection mechanism, and generating a preliminary dynamic mechanical mode decomposition result; And carrying out weighted fusion on the stress distribution characteristics and the motion track characteristics by adopting a characteristic fusion method based on a multi-head attention mechanism on the primary dynamic mechanical mode decomposition result, capturing the complex dependency relationship between the stress distribution and the motion track through a cross-mode attention mechanism, and generating a final knee joint dynamic mechanical characteristic set.
8. The system according to claim 7, wherein the generating module is specifically configured to: according to the dynamic mechanical feature set, a geometric design model based on a generated countermeasure network is adopted, patient anatomy data and biomechanical constraint conditions are combined, a generator and a discriminator are initialized, the geometric shape generated by the generator is ensured to meet the requirements of anatomy and mechanics through a conditional constraint mechanism, and a preliminary prosthesis geometric shape is generated; For the preliminary prosthesis geometry, a geometry adjustment method based on a curvature optimization algorithm is adopted, the curvature and thickness of the prosthesis are optimized by combining the anatomical features of the tibial plateau and the stress distribution in the dynamic mechanical feature set, and the contact surface shape matching of the prosthesis and the bone is ensured through a self-adaptive curvature adjustment mechanism, so that the preliminary optimized geometry is generated; For the preliminarily optimized geometric shape, a contact surface shape optimization method based on contact mechanics analysis is adopted, the contact surface shape of the prosthesis is optimized by combining the motion track and the stress distribution in the dynamic mechanics characteristic set, and the mechanical property of the contact surface is ensured through a dynamic friction coefficient adjustment mechanism, so that the preliminarily optimized contact surface shape is generated; And (3) for the preliminarily optimized geometric shape and contact surface shape, adopting a verification method based on finite element analysis, combining a patient anatomic model and biomechanical constraint conditions, verifying the mechanical property and biocompatibility of the geometric model, and dynamically adjusting the geometric parameters through a feedback correction mechanism to generate a final preliminarily optimized tibial prosthesis geometric model.
9. A storage medium having a computer program stored therein, wherein the computer program is arranged to perform the method of any of claims 1-5 when run.
10. An electronic device comprising a memory and a processor, wherein the memory has stored therein a computer program, the processor being arranged to run the computer program to perform the method of any of claims 1-5.

Description

Digital design and manufacturing method and system of knee joint tibial prosthesis Technical Field The invention belongs to the technical field of design and manufacture, and particularly relates to a digital design and manufacture method and system of a knee joint tibial prosthesis. Background Knee joints, which are one of the most complex joints in human motion, are subjected to sustained dynamic pressure and load, and any damage to their structure and function can lead to serious disability in locomotion. With age and increased motor injury, knee joint pathologies such as arthritis and meniscal injury are increasingly severe, and many patients eventually require knee replacement surgery. In the process, the design and the manufacture of the knee joint tibial prosthesis are key links, and the operation effect and the postoperative rehabilitation quality of a patient are directly affected. Traditional knee joint tibial prosthesis design methods generally rely on standard anatomical models, and fail to fully consider stress conditions and motion trajectories of individual patients in different motion states, resulting in discomfort between the prosthesis and the organism of the patient, affecting the function and durability of the prosthesis. Disclosure of Invention The invention aims to provide a digital design and manufacturing method and system of a knee joint tibia prosthesis, which are used for solving the defects in the prior art, ensuring the precision and consistency of the manufacturing of the prosthesis and further improving the overall quality of the prosthesis. One embodiment of the present application provides a method of digitally designing and manufacturing a knee tibial prosthesis, the method comprising: According to gait analysis data and dynamic mechanical sensor data of a patient, stress distribution and motion tracks of the knee joint in different motion states are collected, and a dynamic mechanical feature extraction algorithm based on time sequence decomposition is adopted to separate key stress modes and motion features of the tibia platform, so that a dynamic mechanical feature set of the knee joint is obtained; Inputting the dynamic mechanical feature set into a geometric design model based on a generation countermeasure network to generate an initial tibial prosthesis geometric shape, wherein the geometric design model optimizes the curvature, thickness and contact surface shape of the prosthesis by combining patient anatomical data and biomechanical constraint conditions to obtain a preliminarily optimized tibial prosthesis geometric model; According to the geometrical model of the tibial prosthesis, an internal porous structure of the prosthesis is designed by utilizing a material microstructure optimization algorithm based on deep reinforcement learning, wherein the material microstructure optimization algorithm generates microstructure distribution with gradient porosity and mechanical property by combining mechanical property simulation and multi-objective optimization, and a material and structure integrated tibial prosthesis design scheme is obtained; Inputting the design scheme of the tibial prosthesis into multi-axis additive manufacturing equipment, and adopting an electron beam melting technology to manufacture the prosthesis, wherein the manufacturing process ensures the precision and consistency of the manufacturing of the prosthesis through a real-time deformation monitoring and self-adaptive path planning technology, and after the manufacturing is finished, adopting a dynamic function verification method based on a biomechanical test platform to simulate the daily movement load of the knee joint, verifying the functionality and durability of the prosthesis and obtaining the final knee joint tibial prosthesis. Optionally, according to gait analysis data and dynamic mechanical sensor data of the patient, stress distribution and motion trajectories of the knee joint in different motion states are collected, a dynamic mechanical feature extraction algorithm based on time sequence decomposition is adopted to separate key stress modes and motion features of the tibial plateau, and a dynamic mechanical feature set of the knee joint is obtained, including: According to gait analysis data and dynamic mechanical sensor data of a patient, a multisource data acquisition frame based on time stamp alignment is adopted to acquire stress distribution and motion trail of the knee joint in different motion states in real time, and time difference of multisource data is eliminated through a data synchronization algorithm to generate a multisource data set with time synchronization; For a time-synchronous multi-source data set, removing noise and abnormal values by adopting a data cleaning method based on a self-adaptive filtering algorithm, and enhancing the characteristic expression of a key stress mode and a motion track by a characteristic enhancement technology to generate a high-