Orthopedic implant adaptive 3D printing control method and system

Abstract

The application discloses an orthopedic implant adaptive 3D printing control method and system, wherein the method comprises the following steps: acquiring medical image data of a target bone region of a patient to determine a bone defect site, establishing a three-dimensional model corresponding to the bone defect site, and determining initial design parameters of an implant; based on the initial design parameters, a topology optimization algorithm based on a SIMP model is used to analyze the three-dimensional model to obtain a printing material distribution density cloud map; a Voronoi polygon diagram algorithm is used to plan seed point distribution of the density cloud map to generate a bionic structure processing map corresponding to different regions; the bionic structure processing map is subjected to printing layering processing to form a plurality of printing layers, and a printing strategy corresponding to each printing layer is generated; and based on ideal topography parameters corresponding to the bionic structure processing map, printing parameters are adjusted in real time. The application can meet the individual processing requirements of different orthopedic implants, effectively improve the overall printing efficiency, and ensure the consistency of the printing quality.

Description

Technical Field

[0001] This invention relates to the field of additive manufacturing technology for orthopedic implants, and in particular to an adaptive 3D printing control method and system for orthopedic implants. Background Technology

[0002] Mechanical fit and biocompatibility are key indicators for evaluating orthopedic implants. Mechanical fit is mainly determined by the molding process of orthopedic implants. Since different patients have different requirements for the shape and performance of the orthopedic implants they need, 3D printing technology has made it possible to manufacture orthopedic implants in a personalized manner, and has therefore become the main molding method for orthopedic implants.

[0003] However, existing 3D printing technologies mostly use Gaussian beams, which have high energy centers and rapid edge decay, easily leading to unstable molten pools, spatter, and thermal stress concentration, affecting density and surface quality. Furthermore, they are less efficient when printing large-format structures. Additionally, existing 3D printing technologies typically use the same laser process parameters for different areas of the same orthopedic implant, failing to achieve on-demand printing and resulting in insufficient printing accuracy.

[0004] Therefore, it is necessary to improve existing 3D printing molding processes to meet the personalized processing needs of different orthopedic implants. Summary of the Invention

[0005] In order to overcome the shortcomings of the prior art, the present invention aims to provide an adaptive 3D printing control method and system for orthopedic implants, which has the advantages of meeting the personalized processing needs of different orthopedic implants and improving the accuracy of 3D printing.

[0006] The objective of this invention is achieved through the following technical solution: According to a first aspect of the present disclosure, an adaptive 3D printing control method for orthopedic implants is provided, comprising: Acquire medical imaging data of the patient's target bone region to determine the location of bone defects, establish a three-dimensional model of the corresponding bone defect location, and determine the initial design parameters of the implant; Based on the initial design parameters, the three-dimensional model is analyzed using a topology optimization algorithm based on the SIMP model to obtain a density cloud map of the printing material. The functional regions of the density cloud map include at least a high-density region as the main load-bearing structure, a low-density region as the porous bone ingrowth structure, and a transition region. The Thiessen polygon graph algorithm is used to plan the distribution of seed points in the density cloud map to generate biomimetic structure processing maps for different regions. The biomimetic structure fabrication map is printed in layers to form several printing layers, and a printing strategy is generated for each printing layer. The printing strategy is to switch the Gaussian beam and flat-top beam to perform printing actions based on the functional regions on the density cloud map corresponding to each printing layer in order to match the seed point distribution plan of the biomimetic structure fabrication map. During the printing process, the printing scanning path and molten pool parameters are acquired in real time to predict the actual morphological parameters of the formed structure, and the printing parameters are adjusted in real time based on the ideal morphological parameters corresponding to the biomimetic structure processing map.

[0007] In some exemplary embodiments, when determining the initial design parameters of the implant, an implant design space is also determined based on the initial design parameters, and boundary conditions are defined within the implant design space to provide constraint parameters. The boundary conditions include: human skeletal biomechanical parameters corresponding to the implant, bone regeneration requirement parameters, printing material characteristic parameters, and 3D printing equipment operating parameters.

[0008] In some exemplary embodiments, the step of using a topology optimization algorithm based on a SIMP model to analyze the 3D model to obtain a printing material distribution density cloud map based on the initial design parameters specifically includes: The implant design space is discretized into N finite elements based on the boundary conditions determined by the initial design parameters, and a pseudo-density parameter is assigned to each finite element. Using the minimization of structural flexibility as the objective function and the material volume fraction as a constraint, the SIMP model is used to establish the correspondence between pseudo-density parameters and material physical parameters; Finite element analysis is performed on each of the finite elements, and the derivatives of the objective function and constraints with respect to each design variable are iteratively calculated to perform sensitivity analysis; An optimization algorithm is used to update the pseudo-density parameters of all the finite elements until the calculation results converge, so as to finally obtain the density cloud map of the printed material.

[0009] In some exemplary embodiments, the step of using the Thiessen polygon graph algorithm to plan the seed point distribution of the density cloud map to generate biomimetic structure fabrication maps corresponding to different regions specifically includes: Obtain the pseudo-density parameters of each finite element corresponding to the density cloud map of the printing material distribution, and use the Poisson disk sampling method to plan the distribution of seed points to generate a seed point set; Based on the set of seed points, construct Thiessen polyhedra corresponding to various sub-points to form an interconnected three-dimensional network pore space; The dual graph of the Thiessen polyhedron is calculated to determine the trabecular edges between two seed points that are located on the same shared face, and the diameter of each trabecular edge is determined based on the pseudo-density parameter corresponding to the seed point to generate a trabecular network with a gradient diameter. The skeletal network is scanned and shaped to generate a biomimetic structural fabrication atlas that can be used for 3D printing.

[0010] In some exemplary embodiments, the step of switching between Gaussian beams and flat-top beams based on the functional regions on the density contour maps corresponding to each printing layer to perform the printing action specifically involves: If the printed layer corresponds to a high-density area, it is mapped to the first printing mode that uses a flat-top beam to perform the printing action; If the printed layer corresponds to a low-density region, it is mapped to a second printing mode that uses a Gaussian beam to perform the printing action; If the printed layer corresponds to a transition region, it is mapped to a third printing mode: first, the main body region is printed using a flat-top beam, and then the boundary region is printed using a Gaussian beam.

[0011] In some exemplary embodiments, during the printing layering process, each printing layer is assigned a beam pattern identifier and initial process parameters, the beam pattern identifier being used to map the printing pattern.

[0012] In some exemplary embodiments, the real-time acquisition of the printing scan path and molten pool parameters to predict the actual formed structure morphology parameters specifically includes: The system acquires the spatial coordinates of the current printing point, scanning speed, scanning direction, and beam mode identifier in real time to determine the printing scanning path, and acquires the molten pool shape information, molten pool temperature information, and molten pool height information in real time to output molten pool parameters. The theoretical molten pool size and heat-affected zone are calculated based on the printing scanning path, molten pool parameters and initial process parameters, and the calculation results are corrected and compensated in real time using a neural network algorithm model. Based on the correction and compensation results, output the actual morphological parameters of the formed structure predicted within the current printing point and the predetermined time period.

[0013] In some exemplary embodiments, when obtaining the print scan path and melt pool parameters, timestamps and spatial coordinate labels are also marked on the print points to make the melt pool parameters correspond to the print position.

[0014] In some exemplary embodiments, the real-time adjustment of printing parameters based on the ideal morphology parameters corresponding to the biomimetic structural processing atlas specifically includes: After extracting the ideal morphological parameters corresponding to the current printing point from the biomimetic structure processing map, the predicted actual morphological parameters of the formed structure are compared with the ideal morphological parameters to calculate the deviation value. The printing parameters are adjusted in real time according to the deviation value. The printing parameters include: laser power, scanning speed and spot diameter.

[0015] According to a second aspect of the present disclosure, an adaptive 3D printing control system for orthopedic implants is provided, comprising: The model building unit is used to acquire medical imaging data of the patient's target bone region to determine the location of bone defects, build a three-dimensional model of the corresponding bone defect location, and determine the initial design parameters of the implant. The topology analysis unit is used to analyze the three-dimensional model based on the initial design parameters using a topology optimization algorithm based on the SIMP model to obtain a density cloud map of the printing material distribution. The functional areas of the density cloud map include at least a high-density area as the main load-bearing structure, a low-density area as the porous bone ingrowth structure, and a transition area. The biomimetic generation unit is used to plan the distribution of seed points on the density cloud map using the Thiessen polygon graph algorithm to generate biomimetic structure processing maps corresponding to different regions. A printing layer unit is used to perform printing layer processing on the biomimetic structure processing map to form several printing layers, and generate a printing strategy corresponding to each printing layer. The printing strategy is: based on the functional area switching of the density cloud map corresponding to each printing layer, the Gaussian beam and the flat top beam are used to perform printing actions to match the seed point distribution plan of the biomimetic structure processing map. The dynamic adjustment unit is used to acquire the printing scan path and molten pool parameters in real time during the printing task to predict the actual formed structure morphology parameters, and adjust the printing parameters in real time based on the ideal morphology parameters corresponding to the biomimetic structure processing map.

[0016] In summary, compared with the prior art, the present invention has the following beneficial effects: This invention provides an adaptive 3D printing control method and system for orthopedic implants. By using a topology optimization algorithm based on the SIMP model to analyze the 3D model, the resulting density cloud map accurately reflects the density distribution of the implant. The density cloud map is divided into high-density, low-density, and transitional regions. The Thiessen polygon map algorithm is then used to process the density cloud map, resulting in a biomimetic structural fabrication map suitable for 3D printing. The macroscopic density field obtained from the topology optimization algorithm is used as a guiding signal to control the microscopic porosity distribution of the Thiessen polygon, thus ensuring the internal structure of the implant is controlled both macroscopically and microscopically. At the microscale, it meets biomechanical adaptation requirements and can satisfy the personalized processing needs of different orthopedic implants. During printing, it switches between Gaussian beams and flat-top beams to perform printing actions corresponding to different functional areas. It makes full use of the energy concentration characteristics of Gaussian beams and the energy uniform distribution characteristics of flat-top beams. While ensuring the forming quality of complex biomimetic structures, it can effectively improve the overall printing efficiency. Furthermore, during the printing process, it acquires the printing scanning path and melt pool parameters in real time to predict the actual formed structure morphology parameters. It then compares the actual morphology parameters with the ideal morphology parameters and dynamically adjusts the printing parameters to further improve the 3D printing accuracy and form a closed-loop control to ensure the consistency of printing quality. Attached Figure Description

[0017] Figure 1 This is a flowchart of the adaptive 3D printing control method for orthopedic implants in an embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram of medical imaging data acquired by the adaptive 3D printing control method for orthopedic implants in an embodiment of the present invention.

[0019] The numbers and letters in the diagram represent the names of the corresponding components: 10. Model building unit; 20. Topology analysis unit; 30. Bionic generation unit; 40. Printing layer unit; 50. Dynamic adjustment unit. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] like Figure 1 and 2 As shown, the first aspect of this invention provides an adaptive 3D printing control method for orthopedic implants, comprising: S100. Obtain medical imaging data of the patient's target bone region to determine the location of the bone defect, establish a three-dimensional model of the corresponding bone defect location, and determine the initial design parameters of the implant.

[0022] Specifically, medical imaging data, such as CT or MRI images, can clearly show the actual image of the target bone region, thereby determining the defect location. Then, a three-dimensional model of the bone defect is reconstructed using reverse engineering software such as Mimics or Geomagic. The shape and outline of the implant can be determined through this three-dimensional model, and the initial design parameters of the implant are the design parameters determined based on this three-dimensional model.

[0023] Simultaneously, when determining the initial design parameters of the implant, the implant design space is also determined based on the initial design parameters, and boundary conditions are defined within the implant design space to provide constraint parameters. The boundary conditions include: the human skeletal mechanical parameters corresponding to the implant, bone regeneration requirement parameters, printing material characteristic parameters, and 3D printing equipment operating parameters. The human skeletal mechanical parameters corresponding to the implant include parameters such as elastic modulus and compressive strength; the bone regeneration requirement parameters include parameters such as target porosity gradient and pore size range; the printing material characteristic parameters include the physical property parameters of the selected printing material; and the 3D printing equipment operating parameters include parameters such as laser power range, printing scanning speed range, and layer thickness.

[0024] The implant design space is used to define the design boundaries, such as limiting the printing start point, printing end point, and printing range. The implant design space strictly matches the patient's bone defect site and provides a spatial basis for the subsequent topology optimization process.

[0025] S200. Based on the initial design parameters, the three-dimensional model is analyzed using a topology optimization algorithm based on the SIMP model to obtain a density cloud map of the printing material distribution. The functional areas of the density cloud map include at least a high-density area as the main load-bearing structure, a low-density area as the porous bone ingrowth structure, and a transition area.

[0026] The SIMP model (Solid Isotropic Material with Penalization) is a density-stiffness interpolation model commonly used in topology optimization. By introducing a continuous pseudo-density variable, it transforms the discrete choice problem of "with material (1) or without material (0)" in the design of printed material distribution into a continuous and differentiable optimization problem, and forces the final result to approximate a clear 0-1 distribution.

[0027] Specifically, the S200 includes: S201. Based on the boundary conditions determined by the initial design parameters, the implant design space is discretized into N finite elements, and a pseudo-density parameter is assigned to each finite element. The boundary conditions and the implant design space define the boundaries of the division. The finite elements are typically tetrahedral or hexahedral meshes, and the value of the pseudo-density parameter is set to continuously vary between 0 and 1.

[0028] S202. Using the minimization of structural flexibility as the objective function and the material volume fraction as a constraint, the SIMP model is used to establish the correspondence between pseudo-density parameters and material physical parameters.

[0029] When establishing the correspondence between pseudo-density parameters and material physical parameters using the SIMP model, the SIMP interpolation relationship is as follows: ; Where: E0 is the elastic modulus of the printing material, such as the elastic modulus of titanium alloy being 110 GPa; E min Take a very small positive integer to represent the elastic modulus without material, to avoid singularity in the overall stiffness matrix, typically E min Take 10 of E0 -3 Up to 10 -9 The default setting is E. min =10 -9 E0;ρ e The pseudo-density parameter of the finite element takes values ​​[0, 1]; p is a penalty factor, which is usually set to p ≥ 3. When p > 1, the equivalent modulus corresponding to the intermediate density value is... It will be much smaller than the linear value This makes the use of intermediate densities mechanically unreasonable, thus penalizing and eliminating them in optimization, ultimately driving the density of most finite elements to a clear boundary of 0 or 1.

[0030] S203. Perform finite element analysis on each of the finite elements, and iteratively calculate the derivatives of the objective function and constraints with respect to each design variable to conduct sensitivity analysis. During the finite element analysis, calculate the displacement and compliance, then iteratively calculate the derivatives of the objective function and constraints with respect to each design variable, and finally perform compliance minimization calculation to achieve sensitivity analysis.

[0031] S204. An optimization algorithm is used to update the pseudo-density parameters of all the finite elements until the calculation results converge, ultimately obtaining the density cloud map of the printed material. The optimization algorithm sets and updates design variables to increase the density of finite elements with high sensitivity and decrease the density of finite elements with low sensitivity. When the change in the objective function or the maximum density change in two consecutive iterations is less than a predetermined threshold, the result is considered converged. This predetermined threshold can be set according to actual needs, for example, to 0.01. After optimization convergence, a model with optimal pseudo-density parameters for each finite element is obtained, which is the density cloud map.

[0032] High-density regions, low-density regions, and transition regions can be distinguished based on preset density thresholds. Since the relative density value of each finite element is 0-1 after topology optimization, high-density and low-density thresholds are set for region division. For example, the high-density threshold can be set to 0.7, and the low-density threshold can be set to 0.3. The pseudo-density parameter of the finite element is ρ, and the high-density threshold is ρ. H The low density threshold is ρ L When ρ≥ρ H When ρ ≤ ρ, the finite element is determined to belong to a high-density region. L When this happens, the finite element is determined to belong to a low-density region, ρ H ≥ρ≥ρ L If the condition is met, then the finite element is determined to belong to the transition region.

[0033] S300. The Thiessen polygon graph algorithm is used to plan the distribution of seed points on the density cloud map to generate biomimetic structure processing maps for different regions.

[0034] The Thiessen polygon graph algorithm divides the entire implant design space into multiple regions by providing several discrete seed points. The distance from any point in each region to the seed point in that region is less than the distance to any other seed point. In the process of orthopedic implant design, the Thiessen polygon graph algorithm can simulate the trabecular network of natural cancellous bone, which has excellent characteristics such as high porosity, low elastic modulus and high specific strength.

[0035] Specifically, the S300 includes: S301. Obtain the pseudo-density parameters of each finite element corresponding to the density cloud map of the printing material distribution, and use the Poisson disk sampling method to plan the distribution of seed points to generate a seed point set.

[0036] When planning the distribution of seed points using the Poisson disk sampling method, the distance between each seed point and any other seed point must be greater than a certain minimum distance. The higher the density, the larger the required minimum distance, and the sparser the actual number of seed points generated. As a result, larger pores are generated in high-density regions and transition regions, while denser seed points and smaller pores are generated in low-density regions. Finally, a set of seed points that satisfy the above rules is generated within the region.

[0037] S302. Based on the set of seed points, construct Tyson polyhedra corresponding to various sub-points to form an interconnected three-dimensional network pore space; each seed point corresponds to a Tyson polyhedron, and the distance from any spatial point within the Tyson polyhedron to its corresponding seed point is less than the distance to any other seed point. These Tyson polyhedra are stacked to fill the implant design space, and the interface between adjacent Tyson polyhedra naturally defines an interconnected three-dimensional network pore space.

[0038] S303. Calculate the dual graph of the Thiessen polyhedron to determine the bone beam edges located on the same shared surface between two seed points, and determine the diameter of each bone beam edge based on the pseudo density parameter corresponding to the seed point to generate a bone beam network with a gradient diameter.

[0039] In this step, the interface of the Thiessen polyhedron is transformed into a trabecular network similar to bone trabeculae. In three-dimensional space, any two seed points are connected. If their Thiessen polyhedra share a face, then there is an edge between these two seed points. Calculating the dual graph of the Thiessen polyhedron yields the corresponding trabecular edges. The set of all trabecular edges constitutes a three-dimensional, interconnected wireframe network, where each trabecular edge is a potential bone trabeculae. Next, the diameter of the bone trabeculae is determined based on the pseudo-density parameters of the two seed points. Each bone trabeculae is not fixed; the higher the average density of the two types of seed points, the larger the diameter of the generated bone trabeculae, and the greater its corresponding stiffness. Let the diameter of the bone trabeculae be d, and the pseudo-density parameters of the two seed points be ρ. i and ρ j , d=d min +β (ρ) i +ρ j ) / 2, where d min The minimum printable diameter is determined by the precision of the 3D printing equipment, and β is the scaling factor. After determining the diameter of all trabeculae, the entire trabeculae network is determined. In this trabeculae network, the trabeculae diameter is larger near the transition region and smaller near the low-density region.

[0040] S304. Perform a scanning and forming operation on the skeleton network to generate a biomimetic structure fabrication map that can be used for 3D printing. This biomimetic structure fabrication map can be, for example, a STEP or STL file. Typically, after generating the biomimetic structure fabrication map, its connectivity is checked to ensure there are no isolated islands, its minimum feature size is checked to ensure it is printable, and its porosity distribution is checked to determine whether it is consistent with the design gradient. In addition, a second-order finite element analysis can be performed on the skeleton network to verify whether its elastic modulus, yield strength, and other parameters meet the design requirements and compare them with the macroscopic target of topology optimization.

[0041] In some embodiments, the Thiessen polygon graph algorithm can be used to plan the distribution of seed points in low-density and transitional regions. For high-density regions, since they are the main mechanical load-bearing structures, the printing material should exist in the form of a completely dense solid or a regular lightweight structure with extremely high relative density, such as a honeycomb structure or a diamond lattice structure. It can be directly set to perform continuous dense printing, which is beneficial to improving printing efficiency.

[0042] S400. The biomimetic structure processing map is processed by printing layering to form several printing layers, and a printing strategy corresponding to each printing layer is generated. The printing strategy is: based on the functional area switching of the density cloud map corresponding to each printing layer, the Gaussian beam and the flat top beam are used to perform printing actions to match the seed point distribution plan of the biomimetic structure processing map.

[0043] After layered printing, the printing strategy for each layer can be precisely controlled. Specifically, the printing action is executed by switching between Gaussian beams and flat-top beams based on the functional areas on the density contour map corresponding to each printing layer: If the printed layer corresponds to a high-density region, it is mapped to the first printing mode that uses a flat-top beam to perform the printing action. Since the flat-top beam has uniform energy, it can greatly improve the stability of the molten pool, increase the scanning speed, and achieve high printing efficiency, thus forming a stable and dense structure. If the printed layer corresponds to a low-density area, it is mapped to a second printing mode that uses a Gaussian beam to perform the printing action. Since the Gaussian beam has strong focusing ability and small spot size, it is suitable for high-precision forming. Using a Gaussian beam to print the skeletal network structure in a low-density area can effectively ensure printing accuracy. If the printed layer corresponds to a transition area, it is mapped to a third printing mode: first, a flat-top beam is used to print the main body area, and then a Gaussian beam is used to print the boundary area. Usually, during printing, a flat-top beam is used to print the main body area to ensure high efficiency, and then a Gaussian beam is used to print the edges, corners and other boundary areas for fine finishing, thereby ensuring that the past is taken.

[0044] Simultaneously, during the layered printing process, each printing layer is assigned a beam mode identifier and initial process parameters. The beam mode identifier is used to map the printing mode, thereby quickly and initially determining which printing mode to switch to when printing the current layer. The initial process parameters may include parameters such as laser power, scanning speed, and spot diameter. In a specific example, when using a flat-top beam, the initial process parameters are set as follows: laser power P = 300W, scanning speed V = 1200mm / s, and spot diameter d = 200μm; when using a Gaussian beam, the initial process parameters are set as follows: laser power P = 150W, scanning speed V = 800mm / s, and spot diameter d = 80μm.

[0045] S500: During the printing process, the printing scanning path and molten pool parameters are acquired in real time to predict the actual morphological parameters of the formed structure, and the printing parameters are adjusted in real time based on the ideal morphological parameters corresponding to the biomimetic structure processing map.

[0046] Specifically, the real-time acquisition of printing scan path and molten pool parameters to predict the actual formed structure morphology parameters includes: S501: Real-time acquisition of the spatial coordinates, scanning speed, scanning direction, and beam mode identifier of the current printing point to determine the printing scanning path; real-time acquisition of molten pool shape information, molten pool temperature information, and molten pool height information to output molten pool parameters; various parameters of the printing scanning path can be directly obtained through the CNC system of the 3D printing equipment, while molten pool parameters are acquired through various sensors configured in the 3D printing equipment. Molten pool shape information, such as the length, width, area, and brightness of the molten pool, can be acquired through a high-speed coaxial vision system. Molten pool temperature information, such as the peak temperature, temperature distribution, and cooling rate of the molten pool, can be acquired through an infrared thermal imager or photodiode. Molten pool height information can be acquired through a laser rangefinder or confocal displacement sensor.

[0047] In some embodiments, when obtaining the printing scan path and molten pool parameters, timestamps and spatial coordinate labels are also marked on the printing points to make the molten pool parameters correspond to the printing position. By marking timestamps and spatial coordinate labels, the molten pool parameters and the printing position can be accurately matched, thereby effectively improving the accuracy of 3D printing.

[0048] S502. Calculate the theoretical molten pool size and heat-affected zone based on the printing scan path, molten pool parameters, and initial process parameters, and use a neural network algorithm model to correct and compensate the calculation results in real time. The theoretical molten pool size and heat-affected zone can be calculated quickly, for example, using the Rosenthal analytical model. The neural network algorithm model can be used for online learning, thereby effectively correcting and compensating the calculation results.

[0049] S503. Based on the correction and compensation results, output the actual molding structure morphology parameters predicted for the current printing point and within the predetermined time period. The actual molding structure morphology parameters usually include parameters such as melt channel geometry and surface roughness.

[0050] Furthermore, the real-time adjustment of printing parameters based on the ideal morphology parameters corresponding to the biomimetic structure processing map specifically includes: After extracting the ideal morphological parameters corresponding to the current printing point from the biomimetic structure processing map, the predicted actual morphological parameters of the formed structure are compared with the ideal morphological parameters to calculate the deviation value. The printing parameters are adjusted in real time according to the deviation value. The printing parameters include: laser power, scanning speed and spot diameter.

[0051] The ideal morphology parameters may include parameters such as the target trabecular diameter, target surface roughness, and target porosity. The deviation value between the actual formed structure morphology parameters and the ideal morphology parameters is obtained by comparing the actual morphology parameters with the ideal morphology parameters. It is then determined whether the deviation value exceeds the deviation threshold. For example, the deviation threshold for the trabecular diameter is set to ±5%. When the deviation value exceeds this deviation threshold, the printing parameters need to be adjusted, such as by rapidly fine-tuning the laser power. By adjusting the printing parameters in real time, closed-loop control of printing is achieved, thereby improving the consistency of printing quality.

[0052] A second aspect of this invention also provides an adaptive 3D printing control system for orthopedic implants, comprising: a model building unit, configured to acquire medical imaging data of a patient's target bone region to determine the bone defect location, build a three-dimensional model corresponding to the bone defect location, and determine the initial design parameters of the implant; a topology analysis unit, configured to analyze the three-dimensional model using a topology optimization algorithm based on a SIMP model according to the initial design parameters to obtain a density cloud map of the printing material distribution, wherein the functional regions of the density cloud map include at least a high-density region serving as the main load-bearing structure, a low-density region serving as a porous bone ingrowth structure, and a transition region; and a biomimetic generation unit, configured to use a Thiessen polygon graph calculation... The method performs seed point distribution planning on the density cloud map to generate biomimetic structure processing maps corresponding to different regions; a printing layer unit is used to perform printing layer processing on the biomimetic structure processing map to form several printing layers, and generate printing strategies corresponding to each printing layer. The printing strategy is: based on the functional area on the density cloud map corresponding to each printing layer, switch the Gaussian beam and flat-top beam to perform printing actions to match the seed point distribution planning of the biomimetic structure processing map; a dynamic adjustment unit is used to acquire the printing scanning path and molten pool parameters in real time during the execution of the printing task to predict the actual formed structure morphology parameters, and adjust the printing parameters in real time based on the ideal morphology parameters corresponding to the biomimetic structure processing map.

[0053] This invention analyzes a 3D model using a topology optimization algorithm based on the SIMP model. The resulting density cloud map accurately reflects the density distribution of the implant. The density cloud map is divided into high-density, low-density, and transitional regions. The Thiessen polygon map algorithm is then used to process the density cloud map, yielding a biomimetic structure fabrication map suitable for 3D printing. The macroscopic density field obtained from the topology optimization algorithm serves as a guiding signal for controlling the microscopic porosity distribution of the Thiessen polygon, ensuring that the implant's internal structure meets biomechanical adaptation requirements at both macroscopic and microscopic scales and satisfies the personalized processing needs of different orthopedic implants. During printing, Gaussian and flat-top beams are switched for different functional regions, fully utilizing the energy concentration and uniform energy distribution characteristics of the Gaussian beam and the flat-top beam. This ensures the quality of complex biomimetic structure formation while effectively improving overall printing efficiency. Furthermore, during printing, the printing scanning path and molten pool parameters are acquired in real-time to predict the actual formed structure's morphology parameters. These parameters are then dynamically adjusted by comparing them with the ideal morphology parameters, further improving 3D printing accuracy and forming a closed-loop control system to guarantee consistent printing quality.

[0054] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that, for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention. These are all equivalent modifications and improvements made to the above embodiments based on the essential technology of the present invention, and all of these fall within the protection scope of the present invention.

Claims

1. An adaptive 3D printing control method for orthopedic implants, characterized in that, include: Acquire medical imaging data of the patient's target bone region to determine the location of bone defects, establish a three-dimensional model of the corresponding bone defect location, and determine the initial design parameters of the implant; Based on the initial design parameters, the three-dimensional model is analyzed using a topology optimization algorithm based on the SIMP model to obtain a density cloud map of the printing material. The functional regions of the density cloud map include at least a high-density region as the main load-bearing structure, a low-density region as the porous bone ingrowth structure, and a transition region. The Thiessen polygon graph algorithm is used to plan the distribution of seed points in the density cloud map to generate biomimetic structure processing maps for different regions. The biomimetic structure fabrication map is printed in layers to form several printing layers, and a printing strategy is generated for each printing layer. The printing strategy is to switch the Gaussian beam and flat-top beam to perform printing actions based on the functional regions on the density cloud map corresponding to each printing layer in order to match the seed point distribution plan of the biomimetic structure fabrication map. During the printing process, the printing scanning path and molten pool parameters are acquired in real time to predict the actual morphological parameters of the formed structure, and the printing parameters are adjusted in real time based on the ideal morphological parameters corresponding to the biomimetic structure processing map.

2. The adaptive 3D printing control method for orthopedic implants according to claim 1, characterized in that, When determining the initial design parameters of the implant, the implant design space is also determined based on the initial design parameters, and boundary conditions are defined in the implant design space to provide constraint parameters. The boundary conditions include: human skeletal biomechanical parameters corresponding to the implant, bone regeneration requirement parameters, printing material characteristic parameters, and 3D printing equipment operating parameters.

3. The adaptive 3D printing control method for orthopedic implants according to claim 2, characterized in that, The process of using a topology optimization algorithm based on a SIMP model to analyze the 3D model based on the initial design parameters to obtain the printing material distribution density cloud map specifically includes: The implant design space is discretized into N finite elements based on the boundary conditions determined by the initial design parameters, and a pseudo-density parameter is assigned to each finite element. Using the minimization of structural flexibility as the objective function and the material volume fraction as a constraint, the SIMP model is used to establish the correspondence between pseudo-density parameters and material physical parameters; Finite element analysis is performed on each of the finite elements, and the derivatives of the objective function and constraints with respect to each design variable are iteratively calculated to perform sensitivity analysis; An optimization algorithm is used to update the pseudo-density parameters of all the finite elements until the calculation results converge, so as to finally obtain the density cloud map of the printed material.

4. The adaptive 3D printing control method for orthopedic implants according to claim 3, characterized in that, The step of using the Thiessen polygon graph algorithm to plan the seed point distribution of the density cloud map to generate biomimetic structure processing maps for different regions specifically includes: Obtain the pseudo-density parameters of each finite element corresponding to the density cloud map of the printing material distribution, and use the Poisson disk sampling method to plan the distribution of seed points to generate a seed point set; Based on the set of seed points, construct Thiessen polyhedra corresponding to various sub-points to form an interconnected three-dimensional network pore space; The dual graph of the Thiessen polyhedron is calculated to determine the trabecular edges between two seed points that are located on the same shared face, and the diameter of each trabecular edge is determined based on the pseudo-density parameter corresponding to the seed point to generate a trabecular network with a gradient diameter. The skeletal network is scanned and shaped to generate a biomimetic structural fabrication atlas that can be used for 3D printing.

5. The adaptive 3D printing control method for orthopedic implants according to claim 4, characterized in that, The specific steps for performing the printing action by switching between the Gaussian beam and the flat-top beam based on the functional regions on the density cloud map corresponding to each printing layer are as follows: If the printed layer corresponds to a high-density area, it is mapped to the first printing mode that uses a flat-top beam to perform the printing action; If the printed layer corresponds to a low-density region, it is mapped to a second printing mode that uses a Gaussian beam to perform the printing action; If the printed layer corresponds to a transition region, it is mapped to a third printing mode: first, the main body region is printed using a flat-top beam, and then the boundary region is printed using a Gaussian beam.

6. The adaptive 3D printing control method for orthopedic implants according to claim 1 or 5, characterized in that, During the printing layering process, each printing layer is assigned a beam pattern identifier and initial process parameters. The beam pattern identifier is used to map the printing mode.

7. The adaptive 3D printing control method for orthopedic implants according to claim 6, characterized in that, The real-time acquisition of printing scan path and molten pool parameters to predict the actual formed structure morphology parameters specifically includes: The system acquires the spatial coordinates of the current printing point, scanning speed, scanning direction, and beam mode identifier in real time to determine the printing scanning path, and acquires the molten pool shape information, molten pool temperature information, and molten pool height information in real time to output molten pool parameters. The theoretical molten pool size and heat-affected zone are calculated based on the printing scanning path, molten pool parameters and initial process parameters, and the calculation results are corrected and compensated in real time using a neural network algorithm model. Based on the correction and compensation results, the system outputs the actual morphological parameters of the formed structure predicted for the current printing point and within the predetermined time period.

8. The adaptive 3D printing control method for orthopedic implants according to claim 7, characterized in that, When obtaining the print scan path and melt pool parameters, timestamps and spatial coordinate labels are also marked on the print points to make the melt pool parameters correspond to the print position.

9. The adaptive 3D printing control method for orthopedic implants according to claim 7, characterized in that, The real-time adjustment of printing parameters based on the ideal morphology parameters corresponding to the biomimetic structural processing map specifically includes: After extracting the ideal morphological parameters corresponding to the current printing point from the biomimetic structure processing map, the predicted actual morphological parameters of the formed structure are compared with the ideal morphological parameters to calculate the deviation value. The printing parameters are adjusted in real time according to the deviation value. The printing parameters include: laser power, scanning speed and spot diameter.

10. An adaptive 3D printing control system for orthopedic implants, characterized in that, include: The model building unit is used to acquire medical imaging data of the patient's target bone region to determine the location of bone defects, build a three-dimensional model of the corresponding bone defect location, and determine the initial design parameters of the implant. The topology analysis unit is used to analyze the three-dimensional model based on the initial design parameters using a topology optimization algorithm based on the SIMP model to obtain a density cloud map of the printing material distribution. The functional areas of the density cloud map include at least a high-density area as the main load-bearing structure, a low-density area as the porous bone ingrowth structure, and a transition area. The biomimetic generation unit is used to plan the distribution of seed points on the density cloud map using the Thiessen polygon graph algorithm to generate biomimetic structure processing maps corresponding to different regions. A printing layer unit is used to perform printing layer processing on the biomimetic structure processing map to form several printing layers, and generate a printing strategy corresponding to each printing layer. The printing strategy is: based on the functional area switching of the density cloud map corresponding to each printing layer, the Gaussian beam and the flat top beam are used to perform printing actions to match the seed point distribution plan of the biomimetic structure processing map. The dynamic adjustment unit is used to acquire the printing scan path and molten pool parameters in real time during the printing task to predict the actual formed structure morphology parameters, and adjust the printing parameters in real time based on the ideal morphology parameters corresponding to the biomimetic structure processing map.

Images