Method and device for generating a model of an intervertebral cage

Abstract

The application is suitable for the medical technical field, and provides a model generation method and device of an intervertebral fusion cage. The method acquires vertebral body image data, and determines a target vertebral body region based on the vertebral body image data. The method determines a bone density spatial distribution based on the target vertebral body region, and determines an elastic modulus spatial distribution in the target vertebral body region according to the bone density spatial distribution. The method determines a regional seed point distribution density and a regional structure parameter distribution of a porous structure of an intervertebral fusion cage according to the elastic modulus spatial distribution. The method performs spatial profiling on a design space corresponding to the target vertebral body region based on the regional seed point distribution density, and obtains a plurality of porous units. The method performs parameterized modeling on the plurality of porous units based on the regional structure parameter distribution, and generates an internal porous structure model of the intervertebral fusion cage. The method can improve the pertinence and rationality of intervertebral fusion cage design, and improve the adaptability of the model generation result to actual application requirements.

Description

Technical Field

[0001] This application relates to the field of medical technology, and in particular to a method and apparatus for generating a model of an interbody fusion device. Background Technology

[0002] Interbody fusion cages are commonly used implants in spinal fusion surgery. They are typically used to restore intervertebral height, maintain segmental stability, and provide support for subsequent bony fusion. Current interbody fusion cages are mostly made of metal, polymer materials, or porous materials, and their structural forms and specifications are usually designed according to general needs to meet common implantation requirements.

[0003] In practical applications, different patients often differ in anatomical morphology, implantation site conditions, and intraoperative adaptation requirements. Existing interbody fusion cages mostly employ standardized designs, offering strong versatility but limited specificity. On the other hand, with the development of medical imaging processing and additive manufacturing technologies, some approaches have attempted individualized designs for interbody fusion cages to improve implant usability. However, these design methods still rely on existing design experience or routine parameter adjustments in practical applications. Therefore, a model generation method for interbody fusion cages is urgently needed to improve the specificity and rationality of interbody fusion cage design and enhance the adaptability of the generated model results to practical application needs. Summary of the Invention

[0004] This application provides a method and apparatus for generating a model of an interbody fusion device, which can improve the pertinence and rationality of the design of the interbody fusion device and enhance the adaptability of the model generation results to the needs of actual applications.

[0005] Firstly, a method for generating a model of an interbody fusion device is provided, the method comprising: Acquire vertebral body imaging data and determine the target vertebral body region based on the vertebral body imaging data; The spatial distribution of bone density is determined based on the target vertebral body region, and the spatial distribution of elastic modulus corresponding to the target vertebral body region is determined based on the spatial distribution of bone density. The distribution density of regional seed points and the distribution of regional structural parameters of the porous structure of the interbody fusion device are determined based on the spatial distribution of the elastic modulus, wherein the distribution of regional structural parameters includes at least one of the pore diameter, wire diameter, unit boundary thickness and inscribed circle diameter of the porous unit; Based on the distribution density of seed points in the region, the design space in the target cone region is spatially divided to obtain multiple porous units; Based on the distribution of the regional structural parameters, the multiple porous units are parametrically modeled to generate a porous structure model inside the intervertebral fusion device.

[0006] In the technical solution of this application, firstly, vertebral body imaging data is acquired and the target vertebral body region is determined, enabling the subsequent model generation process to revolve around the actual vertebral body region to be processed, thus providing a foundation for the targeted design of the interbody fusion cage model. Based on this, the spatial distribution of bone density is determined based on the target vertebral body region, and further, the corresponding spatial distribution of elastic modulus is determined based on the spatial distribution of bone density. This allows the distribution characteristics reflecting the internal mechanical differences of the target vertebral body region to be gradually obtained from the imaging information, providing parameter basis corresponding to the actual situation of the vertebral body region for the subsequent porous structure design. Furthermore, the regional seed point distribution density and regional structural parameter distribution of the porous structure of the interbody fusion cage are determined based on the spatial distribution of elastic modulus, thereby enabling… The generation rules and structural parameters of porous structures in different regions are matched with the mechanical distribution characteristics of the corresponding regions, thereby improving the targeting of porous structure design. Subsequently, the design space corresponding to the target vertebral body region is spatially partitioned based on the distribution density of regional seed points to obtain multiple porous units. This allows the design space corresponding to the target vertebral body region to be divided into separately controllable structural units, providing a carrier for the subsequent construction of porous structures with regional differences. Finally, based on the distribution of regional structural parameters, multiple porous units are parametrically modeled to generate an internal porous structure model of the intervertebral fusion device. This enables the formation of an internal porous structure model corresponding to the spatial distribution of the elastic modulus of the target vertebral body region, thereby improving the adaptability and rationality of the generated intervertebral fusion device model.

[0007] In one possible implementation, the method further includes: Based on the internal porous structure model of the interbody fusion device, an implicit function corresponding to the internal porous structure is constructed. Construct the implicit function of the envelope surface corresponding to the macroscopic shape of the interbody fusion cage; Boolean operations are performed based on the implicit functions corresponding to the internal porous structure and the implicit functions of the envelope surface to fuse the internal porous structure model of the intervertebral fusion device with the macroscopic shape of the intervertebral fusion device, thereby generating a solid model of the intervertebral fusion device.

[0008] In this embodiment, by fusing the internal porous structure model of the interbody fusion device with the macroscopic shape of the interbody fusion device, an integrated solid model with an internal porous structure can be generated, thereby improving the integrity of the model representation.

[0009] In one possible implementation, the method further includes: Multiphysics coupled finite element analysis was performed on the solid model of the interbody fusion device to simulate the stress distribution, peak stress and micromotion between the interbody fusion device and the vertebral endplate under complex physiological loads. Based on the stress distribution, peak stress, and micro-motion, a comprehensive performance evaluation is performed on the solid model of the intervertebral fusion device.

[0010] In this embodiment, by performing multiphysics coupled finite element analysis on the solid model of the intervertebral fusion device, the comprehensive performance of the solid model under complex physiological loads can be evaluated, thereby improving the reliability of the model design results.

[0011] In one possible implementation, determining the target vertebral region based on the vertebral imaging data includes: The vertebral body image data is encoded to extract vertebral body features at different levels; The vertebral body features at different levels are fused based on multi-scale skip connections to obtain the fused vertebral body features. The fused vertebral body features are decoded to obtain vertebral body segmentation results; The target vertebral region is determined based on the vertebral segmentation results.

[0012] In this embodiment, the accuracy of determining the target vertebral region can be improved by encoding, feature fusion, and decoding the vertebral image data.

[0013] In one possible implementation, determining the spatial distribution of bone mineral density based on the target vertebral region includes: Based on the vertebral segmentation results, the gray values ​​at each location within the target vertebral body region are mapped to obtain the bone density values ​​corresponding to each location within the target vertebral body region. Based on the bone mineral density values ​​corresponding to each location within the target vertebral body region, regional feature quantification is performed to obtain the vertebral body bone mineral density analysis results. Based on the vertebral bone density analysis results and the spatial parameters of the vertebral imaging data, three-dimensional coordinates of each location within the target vertebral body region are established. By assigning corresponding bone density attribute values ​​to each location within the target vertebral body region, a three-dimensional point cloud model characterizing the spatial distribution features of bone density within the vertebral body is obtained.

[0014] In this implementation, by mapping grayscale values, quantizing regional features, and constructing three-dimensional point clouds, the spatial distribution characteristics of bone density inside the vertebral body can be characterized, thus providing a foundation for subsequent model generation.

[0015] In one possible implementation, determining the spatial distribution of elastic modulus corresponding to the target vertebral region based on the spatial distribution of bone density includes: A three-dimensional point cloud model characterizing the spatial distribution of bone density within the vertebral body is input into a dual-path network; Extracting trabecular microstructure features using local pathways; Extract long-distance dependency features through global paths; Based on the trabecular microstructure features and the long-distance dependence features, the spatial distribution of elastic modulus corresponding to the target vertebral region is output.

[0016] In this embodiment, by combining the trabecular microstructure features extracted from local paths and the long-distance dependency features extracted from global paths, the spatial distribution of elastic modulus corresponding to the target vertebral region can be output, thereby improving the accuracy of determining the spatial distribution of elastic modulus.

[0017] In one possible implementation, determining the regional seed point distribution density of the porous structure of the interbody fusion cage based on the spatial distribution of the elastic modulus includes: Based on the bone density values ​​of each region in the three-dimensional point cloud model, determine the distribution density of the seed points corresponding to each region. Based on the distribution density of seed points in the region, the design space in the target vertebral region is spatially partitioned, including: Based on the distribution density of seed points in each region, seed points are deployed in the design space within the target vertebral region. By introducing boundary constraints, the seed points are divided into Thiessen polygons to obtain multiple independent and tessellated porous elements.

[0018] In this embodiment, by setting seed points based on the bone density values ​​of each region in the three-dimensional point cloud model and performing Thiessen polygon subdivision, multiple porous units can be obtained, thereby improving the targeting of spatial subdivision.

[0019] In one possible implementation, determining the regional structural parameter distribution of the porous structure of the interbody fusion cage based on the spatial distribution of the elastic modulus includes: Establish a quantitative relationship model between bone mineral density and ideal structural stiffness; Based on the spatial distribution of the elastic modulus and the quantization relationship model, determine the element boundary thickness and the inscribed circle diameter of each region. Based on the element boundary thickness and the inscribed circle diameter of each region, determine the corresponding aperture and wire diameter for each region; The distribution of structural parameters for each region is determined based on the corresponding aperture, wire diameter, unit boundary thickness, and inscribed circle diameter.

[0020] In this embodiment, by establishing a quantitative relationship model between bone density and ideal structural stiffness, the distribution of regional structural parameters can be determined, thereby improving the rationality of porous structure parameter design.

[0021] In one possible implementation, the step of parametrically modeling the plurality of porous units based on the distribution of regional structural parameters to generate a porous structure model of the interbody fusion cage includes: The plurality of porous units and the distribution of the regional structural parameters are input into the biomechanical topology optimization calculator; Based on the principle of cellular automata, the biomechanical topology optimization operator iteratively calculates the multiple porous units; While ensuring structural integrity, the morphology and connection method of the unit walls of the multiple porous units are optimized to maximize porosity and pore space for bone tissue ingrowth. Output the porous structure model of the intervertebral fusion device.

[0022] In this embodiment, by using a biomechanical topology optimization calculator to iteratively calculate multiple porous units and optimize the unit wall morphology and connection method, a porous structure model of the intervertebral fusion device can be generated, thereby improving the optimization degree of the internal porous structure model.

[0023] Secondly, a model generation device for an intervertebral fusion cage is provided, the device comprising: The acquisition module is used to acquire vertebral body image data and determine the target vertebral body region based on the vertebral body image data; The distribution determination module is used to determine the spatial distribution of bone density based on the target vertebral body region, and to determine the spatial distribution of elastic modulus corresponding to the target vertebral body region based on the spatial distribution of bone density. The parameter determination module is used to determine the regional seed point distribution density and regional structural parameter distribution of the porous structure of the interbody fusion device based on the spatial distribution of the elastic modulus, wherein the regional structural parameter distribution includes at least one of the porous unit's pore diameter, wire diameter, unit boundary thickness, and unit inscribed circle diameter; The partitioning module is used to spatially partition the design space in the target cone region based on the distribution density of the seed points in the region, thereby obtaining multiple porous units; The modeling module is used to perform parametric modeling of the multiple porous units based on the distribution of the regional structural parameters, and generate a porous structure model of the intervertebral fusion device.

[0024] Thirdly, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor, when executing the computer program, is capable of implementing any of the methods of the first aspect.

[0025] Fourthly, a computer-readable storage medium is provided that stores a computer program that, when executed by a processor, can implement any of the methods of the first aspect.

[0026] Fifthly, a computer program product is provided, comprising a computer program that, when executed by a processor, can implement any of the methods of the first aspect.

[0027] The beneficial effects of the second to fifth aspects mentioned above can be found in the beneficial effects of the first aspect, and will not be repeated here. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating a method for generating a model of an interbody fusion device according to an embodiment of this application; Figure 2 This is a schematic diagram of the vertebral body image data acquisition and preprocessing process provided in the embodiments of this application; Figure 3 This is a schematic diagram of the process for determining the spatial distribution of elastic modulus provided in an embodiment of this application; Figure 4 This is a schematic diagram of the elastic modulus spatial distribution optimization process provided in the embodiments of this application; Figure 5 This is a schematic diagram of the process for determining the spatial distribution of bone density in the target vertebral region provided in the embodiments of this application; Figure 6 This is a schematic diagram of the process for generating the porous structure model of the interbody fusion device provided in the embodiments of this application; Figure 7 This is a flowchart illustrating another method for generating a model of an interbody fusion device provided in an embodiment of this application; Figure 8 This is a schematic diagram illustrating the process of generating a physical model of an interbody fusion device, evaluating its comprehensive performance, and preparing it, as provided in the embodiments of this application. Figure 9 This is a schematic diagram of the structure of a model generation device for an interbody fusion device provided in an embodiment of this application; Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0029] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0030] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0031] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0032] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0033] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0034] It should be understood that the sequence number of each step in this embodiment does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of this application embodiment.

[0035] Interbody fusion cages are commonly used implants in spinal fusion surgery. They are typically used to restore intervertebral height, maintain segmental stability, and provide support for subsequent bony fusion. Current interbody fusion cages are mostly made of metal, polymer materials, or porous materials, and their structural forms and specifications are usually designed according to general needs to meet common implantation requirements.

[0036] In practical applications, different patients often differ in anatomical morphology, implantation site conditions, and intraoperative adaptation requirements. Existing interbody fusion cages mostly employ standardized designs, offering strong versatility but limited specificity. On the other hand, with the development of medical imaging processing and additive manufacturing technologies, some approaches have attempted individualized designs for interbody fusion cages to improve implant usability. However, these design methods still rely on existing design experience or routine parameter adjustments in practical applications. Therefore, a model generation method for interbody fusion cages is urgently needed to improve the specificity and rationality of interbody fusion cage design and enhance the adaptability of the generated model results to practical application needs.

[0037] To address the aforementioned issues, this application proposes a method and apparatus for generating an interbody fusion cage model. The method first acquires vertebral body imaging data and identifies the target vertebral body region, enabling the subsequent model generation process to revolve around the actual vertebral body region to be processed, thus providing a foundation for the targeted design of the interbody fusion cage model. Based on this, the spatial distribution of bone density is determined according to the target vertebral body region, and further, the corresponding spatial distribution of elastic modulus is determined based on the bone density spatial distribution. This allows for the gradual acquisition of distribution characteristics reflecting the internal mechanical differences of the target vertebral body region from the imaging information, providing parameter basis corresponding to the actual situation of the vertebral body region for subsequent porous structure design. Further, the distribution density of regional seed points and the distribution density of the porous structure of the interbody fusion cage are determined based on the elastic modulus spatial distribution. The distribution of domain structural parameters enables the generation rules and structural parameters of porous structures in different regions to match the mechanical distribution characteristics of the corresponding regions, thereby improving the targeting of porous structure design. Subsequently, the design space corresponding to the target vertebral body region is spatially partitioned based on the distribution density of regional seed points to obtain multiple porous units. This allows the design space corresponding to the target vertebral body region to be divided into separately controllable structural units, providing a carrier for the subsequent construction of porous structures with regional differences. Finally, based on the distribution of regional structural parameters, multiple porous units are parametrically modeled to generate an internal porous structure model of the intervertebral fusion device. This results in an internal porous structure model that corresponds to the spatial distribution of the elastic modulus of the target vertebral body region, thereby improving the adaptability and rationality of the generated intervertebral fusion device model.

[0038] The embodiments are described in detail below with reference to the illustrations.

[0039] See Figure 1 The diagram illustrates a flowchart of a method for generating a model of an interbody fusion device according to an embodiment of this application. Figure 1 As shown, the method includes the following steps: Step 101: Obtain vertebral body imaging data and determine the target vertebral body region based on the vertebral body imaging data.

[0040] Vertebral body imaging data refers to the raw image data obtained after a single scan of the relevant vertebral body region. In this application, vertebral body imaging data is used as the basic data source for subsequent data processing and target vertebral body region determination. Based on vertebral body imaging data, high-resolution computed tomography (CT) images and quantitative CT images can be obtained through different data processing methods.

[0041] Here, a high-resolution computed tomography (CT) image refers to an image obtained from the vertebral body image data after appropriate data processing, used to characterize the vertebral body's related anatomical structures. In this embodiment, the high-resolution CT image is used to subsequently determine the target vertebral body region. Optionally, the slice thickness of the high-resolution CT image is less than or equal to 0.5 mm.

[0042] In this context, the quantitative computed tomography (CT) image refers to the image obtained from the vertebral body image data after appropriate data processing, used for subsequent bone mineral density analysis. In this embodiment, the quantitative CT image is used to subsequently determine the spatial distribution of bone mineral density.

[0043] The target vertebral body region refers to the vertebral body region identified in the vertebral body imaging data, which is used for subsequent spatial distribution analysis of bone mineral density and model generation. The target vertebral body region is the object area for subsequent processing steps, and its determination results are used to limit the processing scope of subsequent bone mineral density analysis, determination of spatial distribution of elastic modulus, and generation of porous structure models.

[0044] In step 101, vertebral body imaging data is first acquired. Specifically, a single scan of the spine of the target is performed to obtain raw image data. Subsequently, different data processing is performed based on the raw image data to obtain high-resolution computed tomography (CT) images and quantitative computed tomography (QCT) images, respectively. Then, the target vertebral body region is determined based on the high-resolution CT images. This provides a foundation for subsequent determination of the spatial distribution of bone density around the target vertebral body region and generation of a porous structure model of the interbody fusion cage.

[0045] For example, in one specific implementation, a scan of the spine can be performed first to obtain vertebral body image data. Then, the vertebral body image data undergoes different data processing steps, with one portion of the processing results used to generate a high-resolution computed tomography (CT) image and another portion used to generate a quantitative computed tomography (QCT) image. Subsequently, the target vertebral body region is determined based on the high-resolution CT image. In this way, corresponding data support can be provided for subsequent target vertebral body region determination and bone density analysis based on the same vertebral body image data.

[0046] In one possible implementation, determining the target vertebral region based on vertebral imaging data includes: The vertebral body image data is encoded to extract vertebral body features at different levels; The vertebral body features at different levels are fused based on multi-scale skip connections to obtain the fused vertebral body features. The fused vertebral body features are decoded to obtain the vertebral body segmentation results; The target vertebral region is determined based on the vertebral segmentation results.

[0047] The encoding process refers to the feature extraction process performed on the input vertebral body image data. In this embodiment, encoding processing can extract vertebral body features at different levels from the vertebral body image data.

[0048] Among them, the different levels of vertebral body features refer to the multi-level feature information that characterizes vertebral body-related information obtained during the encoding process of vertebral body image data. The different levels of vertebral body features can be used to reflect the representation results of the vertebral body at different feature levels.

[0049] Multi-scale skip connections refer to a processing method used to connect and fuse vertebral body features at different levels. In this embodiment, multi-scale skip connections can be used to fuse vertebral body features at different levels to obtain fused vertebral body features.

[0050] The fused vertebral features refer to the vertebral features formed after multi-scale skip connection processing. These fused vertebral features are used as input for subsequent decoding to obtain the vertebral segmentation results.

[0051] Decoding refers to the process of processing the fused vertebral body features and outputting the segmentation result. In this embodiment, the corresponding vertebral body segmentation result can be obtained by decoding the fused vertebral body features.

[0052] The vertebral segmentation result refers to the vertebral region identification result obtained after encoding, feature fusion and decoding of vertebral image data.

[0053] For example, in one specific implementation, vertebral body image data of the target's spine can be acquired first; then, the vertebral body image data can be encoded to extract vertebral body features at different levels; then, the vertebral body features at different levels can be fused through multi-scale skip connections to obtain fused vertebral body features; then, the fused vertebral body features can be decoded to obtain vertebral body segmentation results; finally, the target vertebral body region can be determined based on the vertebral body segmentation results.

[0054] In one possible implementation, prior to encoding the vertebral imaging data, the method further includes: Preprocessing of vertebral imaging data based on attention mechanism; Multi-scale feature pyramid network is used to extract multi-level anatomical structures of vertebral body, endplate and pedicle.

[0055] The attention mechanism refers to a processing mechanism used to enhance vertebral body-related areas in vertebral imaging data and weaken information in non-target areas. In this embodiment, the attention mechanism is used to highlight imaging information related to the vertebral body, endplate, and pedicle during the preprocessing stage to facilitate the subsequent extraction of multi-layered anatomical structures.

[0056] The multi-scale feature pyramid network refers to a network structure used to extract features at different scales. In this embodiment, the multi-scale feature pyramid network is used to extract multi-level anatomical structures of the vertebral body, endplate, and pedicle from vertebral imaging data.

[0057] The vertebral body refers to the bony structure in the spine that provides support and load-bearing capacity. The vertebral body is the primary object for determining the target vertebral region in subsequent analyses.

[0058] The endplate refers to the vertebral body end structure adjacent to the intervertebral disc. In this embodiment, the endplate is extracted as one of the vertebral body-related anatomical structures.

[0059] The pedicle refers to the bony part connecting the vertebral body to the posterior structures of the vertebral body. In this embodiment, the pedicle is extracted as one of the anatomical structures related to the vertebral body.

[0060] In this embodiment, the vertebral body image data is first preprocessed based on an attention mechanism to enhance the image features of the corresponding regions of the vertebral body, endplate, and pedicle, while weakening information in non-target regions. Then, based on a multi-scale feature pyramid network, the multi-level anatomical structures of the vertebral body, endplate, and pedicle are extracted from the preprocessed vertebral body image data. After that, the vertebral body image data is further encoded.

[0061] Specifically, see Figure 2 The diagram illustrates a schematic of the vertebral body image data acquisition and preprocessing process provided in an embodiment of this application. In this process, vertebral body image data is first acquired; then, the vertebral body image data is preprocessed based on an attention mechanism to enhance the image features of the vertebral body-related areas and weaken information in non-target areas; next, multi-level anatomical structures of the vertebral body, endplate, and pedicle are extracted based on a multi-scale feature pyramid network; finally, the preprocessed vertebral body image data is output.

[0062] Step 102: Determine the spatial distribution of bone density based on the target vertebral body region, and determine the spatial distribution of elastic modulus in the target vertebral body region based on the spatial distribution of bone density.

[0063] Spatial distribution of bone mineral density (BMD) refers to the spatial distribution of BMD values ​​at different locations within the target vertebral body region. In other words, spatial distribution of BMD not only reflects the existence of BMD differences within the target vertebral body region, but also indicates the locations of these differences within the target vertebral body region. Through spatial distribution of BMD, the spatial variation characteristics of bone quality within the target vertebral body region can be characterized.

[0064] The spatial distribution of elastic modulus refers to the spatial distribution of elastic modulus values ​​at different locations within the target vertebral region. In other words, the spatial distribution of elastic modulus characterizes the differences in mechanical response capabilities at different locations within the target vertebral region. Since the elastic modulus reflects a material's ability to resist elastic deformation, the spatial distribution of elastic modulus can be used to characterize the spatial variations in the mechanical properties within the target vertebral region.

[0065] After identifying the target vertebral body region in step 101, the spatial distribution of bone mineral density (BMD) is first determined based on this region. Specifically, the imaging information corresponding to each location within the target vertebral body region is analyzed to obtain the BMD value for each location, and the spatial distribution of BMD within the target vertebral body region is formed based on the spatial relationship of each location within the region. Subsequently, based on the spatial distribution of BMD, the spatial distribution of elastic modulus corresponding to the target vertebral body region is further determined. In other words, after obtaining the BMD distribution results at different locations within the target vertebral body region, the elastic modulus values ​​at different locations within the target vertebral body region are obtained based on the correspondence between BMD and elastic modulus, thus forming the spatial distribution of elastic modulus. Through the above processing, the spatial distribution of BMD reflecting the differences in bone quality within the target vertebral body region and the spatial distribution of elastic modulus reflecting the differences in mechanical properties can be gradually obtained from the imaging information of the target vertebral body region. This provides a foundation for subsequently determining the regional seed point distribution density and regional structural parameter distribution of the porous structure of the interbody fusion device based on the spatial distribution of elastic modulus.

[0066] For example, in one specific implementation, after determining the target vertebral body region in step 101, the image grayscale information corresponding to each location within the target vertebral body region can be analyzed to obtain the bone mineral density (BMD) values ​​at each location, thus forming a spatial distribution of BMD in the target vertebral body region. Subsequently, the corresponding elastic modulus values ​​are determined based on the BMD values ​​at each location, forming a spatial distribution of elastic modulus corresponding to the target vertebral body region. For instance, if the BMD value at the first location within the target vertebral body region is higher than the BMD value at the second location, then the determined elastic modulus value at the first location can also be higher than the elastic modulus value at the second location. In this way, the determined spatial distribution of elastic modulus can correspond to the actual bone distribution within the target vertebral body region, thereby providing a basis for the subsequent regional design of the porous structure model.

[0067] In one possible implementation, determining the spatial distribution of bone mineral density based on the target vertebral body region includes: Based on the vertebral segmentation results, the gray values ​​of each location within the target vertebral body region are mapped to obtain the bone density values ​​corresponding to each location within the target vertebral body region. Based on the bone mineral density values ​​at each location within the target vertebral body region, regional feature quantification is performed to obtain the vertebral body bone mineral density analysis results. Based on the results of vertebral bone mineral density analysis and the spatial parameters of vertebral imaging data, three-dimensional coordinates of each location within the target vertebral body region are established. By assigning corresponding bone density attribute values ​​to each location within the target vertebral body region, a three-dimensional point cloud model characterizing the spatial distribution features of bone density within the vertebral body is obtained.

[0068] Gray-scale mapping refers to the process of converting the gray-scale information of images corresponding to various locations within a target vertebral body region into bone mineral density values. Through gray-scale mapping, a correspondence can be established between gray-scale information in image data and bone mineral density information, thereby obtaining the bone mineral density values ​​corresponding to different locations within the target vertebral body region.

[0069] Regional feature quantification refers to the process of statistically analyzing the bone mineral density values ​​at various locations within the target vertebral body region. Through regional feature quantification, analytical results characterizing the bone mineral density of the target vertebral body region can be obtained.

[0070] The vertebral body bone mineral density (VMD) analysis results refer to the analysis results obtained after quantifying the regional characteristics based on the bone mineral density values ​​corresponding to each location within the target vertebral body region. The VMD analysis results reflect the distribution characteristics of bone mineral density within the target vertebral body region and serve as the basis for subsequently establishing three-dimensional coordinates and assigning bone mineral density attribute values.

[0071] Spatial parameters refer to parameters used to characterize the spatial relationships between various locations in vertebral imaging data. Spatial parameters are used to determine the corresponding positions of each location within the target vertebral region in three-dimensional space, thus providing a basis for establishing three-dimensional coordinates.

[0072] In this context, three-dimensional coordinates refer to the positional coordinates of various locations within the target vertebral body region in three-dimensional space. By establishing three-dimensional coordinates, bone density information at various locations within the target vertebral body region can be correlated with their corresponding spatial locations.

[0073] Bone mineral density (BMD) attribute values ​​refer to the attribute information assigned to each location within the target vertebral body region to characterize the bone density at that location. These BMD attribute values ​​are combined with corresponding three-dimensional coordinates to form a data model characterizing the spatial distribution of bone density within the vertebral body.

[0074] The 3D point cloud model refers to a three-dimensional data model composed of the three-dimensional coordinates of each location within the target vertebral body and their corresponding bone density attribute values. The 3D point cloud model is used to characterize the spatial distribution of bone density within the vertebral body and provides the input basis for subsequently determining the spatial distribution of the elastic modulus.

[0075] For example, in one specific implementation, after obtaining the vertebral body segmentation results, the grayscale values ​​at each location within the target vertebral body region can be mapped to obtain the corresponding bone density values. Then, the bone density values ​​are quantized using regional features to obtain the vertebral body bone density analysis results. Next, based on the spatial parameters of the vertebral body image data, three-dimensional coordinates of each location within the target vertebral body region are established. Then, the bone density values ​​corresponding to each location are assigned as bone density attribute values ​​to the corresponding three-dimensional coordinate points, forming a three-dimensional point cloud model. For instance, if the grayscale value corresponding to the first location within the target vertebral body region is higher than the grayscale value corresponding to the second location, then after mapping, the bone density value corresponding to the first location can be higher than the bone density value corresponding to the second location. After establishing the three-dimensional coordinates and assigning bone density attribute values, the spatial difference in bone density between the first and second locations can be reflected in the three-dimensional point cloud model. This method allows for a more intuitive representation of the spatial distribution characteristics of bone density within the target vertebral body region.

[0076] Specifically, see Figure 5 This diagram illustrates the process for determining the spatial distribution of bone density in a target vertebral body region according to an embodiment of this application. The process first determines the vertebral body segmentation result corresponding to the target vertebral body region based on vertebral body image data. Then, based on the vertebral body segmentation result, the grayscale values ​​of each location within the target vertebral body region are mapped and combined with regional feature quantization processing to obtain the vertebral body bone density analysis result. Subsequently, based on the vertebral body bone density analysis result and the spatial parameters of the vertebral body image data, three-dimensional coordinates of each location within the target vertebral body region are established, and corresponding bone density attribute values ​​are assigned to each location to generate a three-dimensional point cloud model characterizing the spatial distribution features of bone density within the vertebral body. Finally, the vertebral body segmentation result, the vertebral body bone density analysis result, and the three-dimensional point cloud model are output.

[0077] In one possible implementation, determining the spatial distribution of elastic modulus corresponding to the target vertebral region based on the spatial distribution of bone mineral density includes: A three-dimensional point cloud model characterizing the spatial distribution of bone density within the vertebral body is input into a dual-path network; Extracting trabecular microstructure features using local pathways; Extract long-distance dependency features through global paths; Based on the microstructural features and long-distance dependence features of the trabecular bone, the spatial distribution of the elastic modulus corresponding to the target vertebral body region is output.

[0078] A dual-path network is a network structure that includes both local and global paths. It is used to analyze and process input 3D point cloud models to obtain feature information that characterizes local details and overall relationships, thereby determining the spatial distribution of the elastic modulus of the target cone region.

[0079] In this context, the local path refers to the processing path used to extract local structural features within the target vertebral body region. In this embodiment, the local path is used to extract trabecular microstructural features from the 3D point cloud model to reflect the detailed features of the local bone tissue structure within the target vertebral body region.

[0080] Among them, trabecular microstructure features refer to information used to characterize the details of the trabecular structure within the target vertebral body region. These features reflect subtle differences in the local bone tissue structure within the target vertebral body region and provide local characteristic basis for subsequent determination of the spatial distribution of elastic modulus.

[0081] The global path refers to the processing path used to extract the overall correlation features of the target vertebral region. In this embodiment, the global path is used to extract long-distance dependency features from the 3D point cloud model to reflect the overall correlation between different locations within the target vertebral region.

[0082] Long-distance dependency features refer to the feature information used to characterize the relationship between different locations within the target vertebral region. Long-distance dependency features can reflect the spatial relationship characteristics of the target vertebral region within the overall scope and provide a global feature basis for subsequent determination of the spatial distribution of elastic modulus.

[0083] For example, in one specific implementation, a three-dimensional point cloud model characterizing the spatial distribution of bone density within the target vertebral body region can be first input into a dual-path network. Then, trabecular microstructure features corresponding to each location within the target vertebral body region are extracted via local paths, and long-distance dependency features between different locations within the target vertebral body region are extracted via global paths. Finally, based on the trabecular microstructure features and long-distance dependency features, the spatial distribution of elastic modulus corresponding to the target vertebral body region is output. For instance, if there are differences in bone density properties and spatial relationships between the first and second locations within the target vertebral body region, these differences can be reflected in the features extracted via local and global paths, thereby obtaining different elastic modulus values ​​corresponding to the first and second locations, thus forming the spatial distribution of elastic modulus corresponding to the target vertebral body region.

[0084] In one possible implementation, trabecular microstructural features are extracted via local paths, including: Microstructural features of bone trabeculae were extracted using a three-dimensional convolutional neural network.

[0085] Among them, a three-dimensional convolutional neural network refers to a network structure used to perform convolution processing on three-dimensional data.

[0086] In this embodiment, a three-dimensional point cloud model characterizing the spatial distribution of bone density within the vertebral body can be input into a three-dimensional convolutional neural network within a local path. The three-dimensional convolutional neural network processes the local spatial neighborhood information within the target vertebral body region to extract trabecular microstructural features. For example, the three-dimensional convolutional neural network can perform local convolution processing on the bone density attributes and spatial relationships of adjacent locations to obtain trabecular microstructural features reflecting local structural differences.

[0087] In one possible implementation, long-distance dependency features are extracted via global paths, including: Capture long-distance dependencies using a visual transformation network.

[0088] Among them, visual transformation network refers to a network structure used to model the overall correlation of input data.

[0089] In this embodiment, a 3D point cloud model can be input into a visual transformation network in the global path. The visual transformation network models the spatial relationships between different locations within the target vertebral region, thereby obtaining long-distance dependency features. For example, the visual transformation network can process the relationships between different locations within the entire target vertebral region to extract long-distance dependency features that characterize the overall spatial relationships.

[0090] In one possible implementation, the spatial distribution of the elastic modulus corresponding to the target vertebral region is output, including: The spatial distribution of elastic modulus corresponding to the target vertebral region is predicted based on elastic modulus regression.

[0091] Elastic modulus regression prediction refers to the process of predicting the elastic modulus value at each location within the target vertebral region based on input features.

[0092] In this embodiment, the trabecular microstructural features extracted from local paths and the long-distance dependency features extracted from global paths can be input into the elastic modulus regression prediction model to output the elastic modulus values ​​corresponding to each location within the target vertebral body region, thus forming a spatial distribution of the elastic modulus corresponding to the target vertebral body region. For example, if there are differences in local structural features and spatial correlation features between the first and second locations within the target vertebral body region, the first and second locations in the elastic modulus regression prediction results may correspond to different elastic modulus values.

[0093] In one possible implementation, elastic modulus regression prediction includes at least one of the following auxiliary tasks: bone quality classification; stress distribution prediction; fracture risk assessment.

[0094] Among them, auxiliary tasks refer to tasks used to assist the elastic modulus regression prediction process.

[0095] In this embodiment, at least one of the following tasks—bone quality classification, stress distribution prediction, and fracture risk assessment—can be performed simultaneously while elastic modulus regression prediction is being performed, in order to provide auxiliary characterization of the target vertebral body region from different perspectives. For example, while performing elastic modulus regression prediction, a bone quality classification task can be performed simultaneously to distinguish the bone quality status corresponding to the target vertebral body region; a stress distribution prediction task or a fracture risk assessment task can also be performed simultaneously to assist in the analysis of the relevant biomechanical conditions of the target vertebral body region.

[0096] Specifically, see Figure 3 , Figure 3 This diagram illustrates the process for determining the spatial distribution of elastic modulus according to an embodiment of this application. In this process, a three-dimensional point cloud model characterizing the spatial distribution of bone density within the vertebral body is first input into a dual-path network. Then, trabecular microstructure features are extracted via local paths, and long-distance dependency features are extracted via global paths. Elastic modulus regression prediction is then performed based on the extracted features. In some embodiments, auxiliary tasks 1, 2, and 3 can be combined to assist the elastic modulus regression prediction. Finally, the spatial distribution of the elastic modulus corresponding to the target vertebral body region is output.

[0097] In one possible implementation, before outputting the spatial distribution of the elastic modulus corresponding to the target vertebral region, the method further includes: The dual-path network was pre-trained on a skeletal database.

[0098] Pre-training refers to the process of training the dual-path network using existing sample data before performing elastic modulus spatial distribution prediction; the skeletal database refers to the data set used to provide skeletal-related training samples.

[0099] In this embodiment, the dual-path network can be pre-trained using sample data from a skeletal database. This allows the dual-path network to recognize and represent skeletal features before formally processing the 3D point cloud model that characterizes the spatial distribution of bone density within the vertebral body. For example, training samples from the skeletal database can be input into the dual-path network for training, and then the pre-trained dual-path network can be used in the subsequent process of determining the spatial distribution of elastic modulus corresponding to the target vertebral region.

[0100] In one possible implementation, pre-training the dual-path network includes: Transfer learning was used to pre-train the dual-path network; Adversarial training is employed to enhance the robustness of the dual-path network; Uncertainty quantification is used to output the prediction confidence interval.

[0101] Among them, transfer learning refers to the process of initializing or continuing to train a dual-path network using existing training results; adversarial training refers to a training method that improves network stability by introducing perturbation samples; uncertainty quantification refers to a processing method that outputs the confidence level corresponding to the prediction result while outputting the prediction result.

[0102] In this embodiment, transfer learning can be used to complete the initial pre-training of the dual-path network, followed by adversarial training to enhance the network's adaptability to input changes. Uncertainty quantification is then used to output the prediction confidence interval during the prediction output stage. For example, the dual-path network can be pre-trained based on existing skeletal sample training results, then perturbation samples can be added during training for adversarial training. Finally, when outputting the spatial distribution of the elastic modulus corresponding to the target vertebral region, the corresponding prediction confidence interval is output simultaneously.

[0103] See Figure 4 , Figure 4 A schematic diagram of the elastic modulus spatial distribution optimization process provided in an embodiment of this application is shown. In this process, a physical constraint loss function is first introduced to ensure that the predicted elastic modulus spatial distribution conforms to biomechanical principles; then, multi-physics joint training is performed using finite element simulation data; and graph neural networks are used to model the connectivity of the internal structure of the skeleton; finally, the optimized elastic modulus spatial distribution corresponding to the target vertebral region is output.

[0104] In one possible implementation, before outputting the spatial distribution of the elastic modulus corresponding to the target vertebral region, the method further includes: A physical constraint loss function is introduced to ensure that the prediction results of the spatial distribution of elastic modulus conform to biomechanical principles.

[0105] Among them, the physical constraint loss function refers to the loss function used to impose biomechanical constraints on the prediction results of the spatial distribution of elastic modulus.

[0106] In this embodiment, a physical constraint loss function can be introduced when predicting the spatial distribution of the elastic modulus using a dual-path network, so that the predicted spatial distribution of the elastic modulus is more consistent with biomechanical principles. For example, during network training, the physical constraint loss function and the loss corresponding to the elastic modulus regression prediction can be used together for model optimization, thereby constraining the prediction results.

[0107] In one possible implementation, after introducing the physical constraint loss function, it further includes: Multiphysics joint training is performed using finite element simulation data; Graph neural networks are used to model the connectivity of the internal structure of bones.

[0108] Among them, multi-physics joint training refers to the process of jointly training the network by combining finite element simulation data; the internal structural connectivity of the skeleton refers to the structural relationship between different locations inside the skeleton.

[0109] In this embodiment, after introducing a physical constraint loss function, multiphysics joint training can be further performed using finite element simulation data, and graph neural networks can be used to model the connectivity of the internal structure of the skeleton to further optimize the prediction results of the spatial distribution of the elastic modulus. For example, finite element simulation data can be input into the network as joint training data, and graph neural networks can be used to model the connectivity between different locations inside the skeleton, thereby improving the rationality of the output spatial distribution of the elastic modulus.

[0110] Step 103: Determine the regional seed point distribution density and regional structural parameter distribution of the porous structure of the interbody fusion device based on the spatial distribution of elastic modulus.

[0111] The regional seed point distribution density refers to the density of seed points distributed within different regions corresponding to the porous structure of the intervertebral fusion device. This regional seed point distribution density provides the basis for subsequent spatial subdivision of the design space corresponding to the target vertebral body region, in order to form multiple porous units.

[0112] The regional structural parameter distribution refers to the distribution of structural parameters of the porous structure of the interbody fusion cage in different regions. In this application, the regional structural parameter distribution includes at least one of the following: pore diameter, wire diameter, element boundary thickness, and inscribed circle diameter of the porous element. The regional structural parameter distribution is used to characterize the parameter differences of the porous structure of the interbody fusion cage in different regions, so as to provide a basis for subsequent parametric modeling of multiple porous elements.

[0113] In this context, a porous structure refers to a structural form composed of multiple porous units. Porous structures serve as the basic form for the internal structure of interbody fusion cages, enabling subsequent models of the porous internal structure of interbody fusion cages to exhibit regionally differentiated structural characteristics.

[0114] In this embodiment, after obtaining the spatial distribution of elastic modulus corresponding to the target vertebral body region, the regional seed point distribution density and regional structural parameter distribution of the porous structure of the interbody fusion device are determined based on the spatial distribution of elastic modulus. Specifically, the seed point distribution density of the porous structure of the interbody fusion device in different regions can be determined based on the elastic modulus distribution at different locations within the target vertebral body region, and the corresponding structural parameter distribution in different regions can be further determined. That is, in this step, the regional seed point distribution density for subsequent spatial subdivision and the regional structural parameter distribution for subsequent parametric modeling are obtained from the spatial distribution of elastic modulus within the target vertebral body region. Through the above processing, the generation rules of the porous structure of the interbody fusion device can correspond to the mechanical distribution characteristics within the target vertebral body region, thereby providing a foundation for subsequent spatial subdivision and parametric modeling.

[0115] For example, in one specific implementation, the elastic modulus values ​​corresponding to different locations within the target vertebral body region can be obtained first, forming a spatial distribution of the elastic modulus corresponding to the target vertebral body region. Subsequently, based on the spatial distribution of the elastic modulus, different regions corresponding to the porous structure of the subsequent interbody fusion device can be assigned corresponding regional seed point distribution densities and regional structural parameter distributions. For example, if the elastic modulus distributions of the first and second regions are different, the determined regional seed point distribution densities and regional structural parameter distributions can also be different. In this way, the porous structure model of the subsequently generated interbody fusion device can have regional difference characteristics corresponding to the spatial distribution of the elastic modulus within the target vertebral body region.

[0116] In one possible implementation, determining the regional seed point distribution density of the porous structure of the interbody fusion cage based on the spatial distribution of elastic modulus includes: Based on the bone density values ​​of each region in the 3D point cloud model, determine the distribution density of the corresponding seed points for each region. In this embodiment, the distribution density of seed points for each region is determined based on the bone density values ​​of each region in the three-dimensional point cloud model. In other words, based on the three-dimensional point cloud model characterizing the spatial distribution of bone density within the vertebral body, the density of seed point distribution is determined according to the bone density of different regions.

[0117] In one possible implementation, the distribution of regional structural parameters of the porous structure of the interbody fusion cage is determined based on the spatial distribution of elastic modulus, including: Establish a quantitative relationship model between bone mineral density and ideal structural stiffness; Based on the spatial distribution of elastic modulus and the quantization relationship model, determine the element boundary thickness and the inscribed circle diameter of each region. Based on the element boundary thickness and the inscribed circle diameter of each region, determine the corresponding aperture and wire diameter for each region; Based on the aperture, wire diameter, element boundary thickness, and inscribed circle diameter of each region, the distribution of regional structural parameters is determined.

[0118] See Figure 6 , Figure 6 This illustration shows a schematic diagram of the process for generating a porous structure model of the interbody fusion cage according to an embodiment of this application. In this process, after spatial subdivision, the distribution of regional structural parameters of the porous structure of the interbody fusion cage is further determined based on the spatial distribution of the elastic modulus corresponding to the target vertebral body region. Specifically, a quantitative relationship model between bone density and ideal structural stiffness is first established; then, based on the spatial distribution of the elastic modulus and the quantitative relationship model, the element boundary thickness and the inscribed circle diameter of each region are determined; subsequently, based on the element boundary thickness and the inscribed circle diameter of each region, the pore diameter and wire diameter of each region are determined; finally, based on the pore diameter, wire diameter, element boundary thickness, and inscribed circle diameter of each region, the distribution of regional structural parameters is determined.

[0119] The quantitative relationship model refers to a model used to characterize the correspondence between bone mineral density and ideal structural stiffness. This model links the bone mineral density characteristics within the target vertebral body region with the subsequent determination of porous structure parameters, providing a basis for determining structural parameters in different regions.

[0120] Ideal structural stiffness refers to the structural stiffness target corresponding to the bone condition within the target vertebral body region. Ideal structural stiffness is used to reflect the stiffness requirements of different regions during structural design.

[0121] Here, element boundary thickness refers to the thickness parameter of the boundary portion of the porous element. Element boundary thickness is used to characterize the geometric dimensions of the porous element boundary.

[0122] The inscribed circle diameter refers to the diameter parameter corresponding to the inscribed circle inside the porous element. The inscribed circle diameter is used to characterize the internal spatial scale of the porous element.

[0123] Here, pore size refers to the dimensional parameter corresponding to the pores in a porous unit. Pore size is used to characterize the size of pores in a porous structure.

[0124] Here, wire diameter refers to the dimensional parameter corresponding to the connecting structure in the porous unit. Wire diameter is used to characterize the geometric dimensions of the connecting parts in the porous structure.

[0125] In one specific implementation, a quantitative relationship model between bone density and ideal structural stiffness can first be established for the first and second regions. Then, based on the elastic modulus of the first and second regions, the element boundary thickness and inscribed circle diameter of each region are determined. Next, based on the element boundary thickness and inscribed circle diameter, the aperture and wire diameter of each region are determined. Finally, based on the aperture, wire diameter, element boundary thickness, and inscribed circle diameter of each region, the distribution of regional structural parameters for each region is determined. For example, if the elastic modulus of the first and second regions are different, the element boundary thickness, inscribed circle diameter, aperture, and wire diameter determined for the first and second regions can also be different, thus allowing the subsequently generated porous structure model of the interbody fusion cage to have different structural parameters in different regions.

[0126] Step 104: Based on the distribution density of regional seed points, the design space corresponding to the target cone region is spatially partitioned to obtain multiple porous units.

[0127] In this context, the design space refers to the spatial range corresponding to the target vertebral body region, used for spatial subdivision; the porous element refers to the basic unit obtained after spatial subdivision, used to construct the porous structure inside the intervertebral fusion device. In this embodiment, seed points are first placed in the design space corresponding to the target vertebral body region according to the distribution density of seed points in each region; then, boundary constraints are introduced, and the placed seed points are subjected to Thiessen polygon subdivision to obtain multiple independent and nested porous elements. Through the above method, the design space can be divided into multiple separately controllable structural units, thus providing a foundation for subsequent parametric modeling.

[0128] For example, in one specific implementation, seed points can be first placed in the design space corresponding to the first and second regions respectively; then, the placed seed points are divided into Thiessen polygons to form multiple porous units in the first and second regions respectively. This allows the subsequently generated internal porous structure to have regionally differentiated characteristics.

[0129] See Figure 6 , Figure 6 This illustration shows a spatial subdivision process in the generation of the porous structure model of the intervertebral fusion device provided in this embodiment. In this process, firstly, the distribution density of seed points for each region is determined based on the bone density values ​​of each region in the 3D point cloud model; then, seed points are placed in the design space corresponding to the target vertebral body region based on the distribution density of seed points for each region; subsequently, boundary constraints are introduced to perform Thiessen polygon subdivision on the placed seed points; finally, multiple independent and nested porous units are obtained.

[0130] In one possible implementation, the design space in the target cone region is spatially partitioned based on the distribution density of regional seed points, including: Based on the distribution density of seed points in each region, seed points are deployed in the design space within the target vertebral region. By introducing boundary constraints, the seed points are divided into Thiessen polygons to obtain multiple independent and tessellated porous elements.

[0131] Boundary constraints refer to the constraints applied to the design space boundary during spatial partitioning. In this embodiment, boundary constraints are used to define the element boundaries after seed point partitioning, ensuring that the obtained porous elements are located within the design space corresponding to the target cone region.

[0132] Thiessen polygon partitioning refers to the process of dividing the design space based on seed points. In this embodiment, by performing Thiessen polygon partitioning on the deployed seed points, the design space can be divided into multiple independent and tessellated unit regions.

[0133] In this embodiment, the distribution density of seed points for each region is first determined based on the bone density values ​​of each region in the 3D point cloud model. That is, based on the 3D point cloud model characterizing the spatial distribution of bone density within the vertebral body, the density of seed points is determined according to the bone density of different regions. Then, seed points are placed within the design space corresponding to the target vertebral body region according to the distribution density of seed points for each region. Next, boundary constraints are introduced, and the placed seed points are subjected to Thiessen polygon subdivision to obtain multiple independent and nested porous units. Through the above processing, a spatial partitioning result corresponding to the bone density distribution of the target vertebral body region can be formed within the design space, thus providing a foundation for the subsequent generation of the porous structure model inside the interbody fusion cage.

[0134] For example, in one specific implementation, the distribution density of seed points for the first and second regions in the 3D point cloud model can be determined first, based on the bone density values ​​of the first and second regions, respectively. Then, seed points are deployed within the design space corresponding to the target vertebral body region based on these seed point distribution densities. Next, boundary constraints are introduced, and the deployed seed points are subjected to Thiessen polygon subdivision to obtain multiple independent and nested porous units. For instance, if the bone density values ​​of the first and second regions in the 3D point cloud model are different, the determined distribution densities of seed points for the first and second regions can also be different. Based on this, the distribution of porous units formed in the first and second regions after Thiessen polygon subdivision can also be different accordingly. This method ensures that the obtained porous units correspond to the bone density distribution characteristics within the target vertebral body region.

[0135] Step 105: Parametric modeling of multiple porous units is performed based on the distribution of regional structural parameters to generate a porous structure model inside the intervertebral fusion device.

[0136] Parametric modeling refers to the process of constructing a model of multiple porous units based on pre-determined structural parameters. In this embodiment, parametric modeling is used to align each porous unit with the structural parameters of its corresponding region, thereby forming an overall internal porous structure.

[0137] The porous structure model of the interbody fusion cage refers to a model formed by parametric modeling of multiple porous units, used to characterize the porous structure inside the interbody fusion cage. This porous structure model serves as the basis for subsequently generating the solid model of the interbody fusion cage.

[0138] In this embodiment, after obtaining multiple porous units in step 104, parametric modeling is performed on these porous units based on the distribution of regional structural parameters to generate a porous structure model inside the interbody fusion cage. Specifically, multiple porous units can be associated with corresponding regional structural parameter distributions, allowing porous units in different regions to be modeled according to their respective structural parameters, thereby forming a porous structure model inside the interbody fusion cage. Through this method, the generated internal porous structure model can possess structural features corresponding to the structural parameters of different regions.

[0139] For example, in one specific implementation, multiple porous units formed in the first and second regions can be first input into a biomechanical topology optimization calculator; then, based on the distribution of regional structural parameters corresponding to the first and second regions, parametric modeling and iterative calculations can be performed on the multiple porous units; subsequently, the unit wall morphology and connection method of each porous unit can be optimized; finally, the porous structure model inside the interbody fusion cage is output. Thus, the generated porous structure model inside the interbody fusion cage can have different structural characteristics in different regions.

[0140] In one possible implementation, multiple porous units are parametrically modeled based on the distribution of regional structural parameters to generate a porous structure model of the interbody fusion cage, including: Input multiple porous units and regional structural parameter distributions into the biomechanical topology optimization calculator; Based on the principle of cellular automata, a biomechanical topology optimization operator is used to iteratively calculate multiple porous units; While ensuring structural integrity, the morphology and connection method of the unit walls of multiple porous units are optimized to maximize porosity and pore space for bone tissue ingrowth. Output the porous structure model of the intervertebral fusion device.

[0141] The biomechanical topology optimization calculator is a computational tool used to generate, solve, and optimize multiple porous elements. It receives multiple porous elements and their corresponding regional structural parameter distributions, and performs subsequent iterative calculations and structural optimizations to output a porous structure model of the intervertebral fusion cage. In other words, the biomechanical topology optimization calculator is a processing function that "transforms the defined porous elements and determined structural parameters into a usable internal porous structure model."

[0142] The cellular automata principle refers to the process of dividing the structure to be optimized into multiple local units and updating these local units step by step according to a preset update rule. In this embodiment, multiple porous units can be regarded as local structural units that can participate in evolution, and the state of each local structural unit is updated according to the distribution of regional structural parameters corresponding to its location. In other words, the cellular automata principle emphasizes that the final structure is not obtained directly all at once, but rather the overall structure gradually approaches the target state through multiple, step-by-step local updates.

[0143] Iterative calculation refers to the process of repeatedly calculating and updating multiple porous elements according to preset rules. In this embodiment, iterative calculation is not an isolated single-time processing of a porous element, but rather a multi-round update of multiple porous elements by combining the adjacency relationships between multiple porous elements, the structural parameters of their respective regions, and the overall structural requirements. Through iterative calculation, the local structural form of the porous elements can be gradually adjusted, making the final output internal porous structure model more in line with the expected requirements.

[0144] Structural integrity refers to the state in which the generated porous structure can maintain continuity, stability, and form a complete structure as a whole. In other words, when optimizing the unit wall morphology and connection method of multiple porous units, it is not about arbitrarily increasing the porosity or arbitrarily reducing the connections, but about ensuring that each porous unit can form a complete structural system, thereby avoiding structural interruptions, local detachment, or overall discontinuity.

[0145] The unit wall morphology refers to the geometric shape of the boundary portion of the porous unit. The unit wall morphology reflects the specific structural form of the porous unit boundary, such as variations in the boundary contour and the local thickness distribution. In this embodiment, optimizing the unit wall morphology aims to further obtain a more suitable internal structural form for the porous unit while meeting the requirements for regional structural parameter distribution.

[0146] The connection method refers to the structural form in which multiple porous units are interconnected. The connection method reflects how different porous units form an overall structure. In this embodiment, the connection method is optimized to establish a more reasonable connection relationship between multiple porous units, enabling both the formation of a continuous structure and the consideration of pore space formation.

[0147] Porosity refers to the proportion of pores in a porous structure. It characterizes the amount of voids within a porous structure. In this embodiment, maximizing porosity does not simply mean pursuing as many voids as possible, but rather, while ensuring structural integrity, maximizing the degree to which voids can form within the structure to facilitate the subsequent formation of an internal space more suitable for bone tissue ingrowth.

[0148] In this context, the porous space for bone tissue ingrowth refers to the space formed within the porous structure that allows bone tissue to enter and grow. In other words, the porous space not only requires the existence of voids, but also requires that these voids structurally serve as internal spaces for subsequent bone tissue entry and growth. Therefore, in this embodiment, when optimizing multiple porous units, not only is the geometric shape considered, but also whether the optimized porous space is more conducive to bone tissue ingrowth.

[0149] In this embodiment, multiple porous elements and the distribution of regional structural parameters are first input into the biomechanical topology optimization calculator. Here, the multiple porous elements can be understood as the basic structural units obtained after the previous spatial partitioning, and the distribution of regional structural parameters is used to specify the parameter requirements such as pore size, filament diameter, element boundary thickness, and inscribed circle diameter for each region. After receiving the above input, the biomechanical topology optimization calculator does not directly output the final result, but rather performs iterative calculations on the multiple porous elements based on the principle of cellular automata. That is, the calculator will gradually update the local state of each porous element in combination with the structural parameter requirements of the region where each porous element is located, so that the structural form of the porous element is continuously adjusted.

[0150] For example, in one specific implementation, multiple porous units in the first and second regions, along with the corresponding regional structural parameter distributions, can be input into a biomechanical topology optimization calculator. Subsequently, the calculator performs multiple iterative calculations on the porous units in the first and second regions based on cellular automata principles. During the iteration process, the unit wall morphology and connection method of each porous unit can be gradually adjusted according to the parameter requirements of the first and second regions. When the overall structure meets the integrity requirements, the relevant structures are further optimized to increase porosity and create a larger pore space for bone tissue ingrowth. Finally, the internal porous structure model of the interbody fusion cage is output. This method ensures that the obtained internal porous structure model maintains overall structural integrity while possessing a more suitable internal pore space for bone tissue ingrowth.

[0151] In summary, this embodiment of the application acquires vertebral body imaging data, sequentially determines the target vertebral body region, the spatial distribution of bone density, and the spatial distribution of elastic modulus, and further determines the regional seed point distribution density and regional structural parameter distribution of the porous structure of the interbody fusion cage. Based on this, spatial subdivision and parametric modeling are completed, thereby generating a porous structure model of the interbody fusion cage. Thus, the generated porous structure model of the interbody fusion cage corresponds to the spatial distribution characteristics within the target vertebral body region.

[0152] See Figure 7 This document illustrates a flowchart of another method for generating a model of an interbody fusion device according to an embodiment of this application; as shown below. Figure 7 As shown, the method may include the following steps: Step 701: Obtain vertebral body imaging data and determine the target vertebral body region based on the vertebral body imaging data.

[0153] Step 702: Determine the spatial distribution of bone density based on the target vertebral body region, and determine the spatial distribution of elastic modulus in the target vertebral body region based on the spatial distribution of bone density.

[0154] Step 703: Determine the regional seed point distribution density and regional structural parameter distribution of the porous structure of the interbody fusion device based on the spatial distribution of elastic modulus.

[0155] Step 704: Based on the distribution density of regional seed points, the design space in the target cone region is spatially partitioned to obtain multiple porous units.

[0156] Step 705: Parametric modeling of multiple porous units is performed based on the distribution of regional structural parameters to generate a porous structure model inside the intervertebral fusion device.

[0157] The implementation process and principle of steps 701 to 705 are similar to steps 101 to 105 in the above embodiments, and can be referred to each other. The embodiments of this application will not be repeated here.

[0158] Step 706: Generate a solid model of the interbody fusion device based on the porous structure model inside the device.

[0159] The interbody fusion cage solid model refers to a three-dimensional solid model formed by fusing the internal porous structure model of the interbody fusion cage with its macroscopic shape. The solid model of the interbody fusion cage can serve as the basis for subsequent comprehensive performance evaluation and fabrication.

[0160] See Figure 8 , Figure 8 This illustration shows a schematic diagram of the process for generating, comprehensively evaluating, and fabricating a solid model of an interbody fusion device (IBDD) according to an embodiment of this application. In this process, firstly, implicit functions corresponding to the internal porous structure are constructed based on the IBDD internal porous structure model, and implicit functions of the envelope surface corresponding to the macroscopic shape of the IBDD are also constructed. Then, Boolean operations are performed based on the implicit functions corresponding to the internal porous structure and the implicit functions of the envelope surface to fuse the IBDD internal porous structure model with the macroscopic shape of the IBDD, generating a solid model of the IBDD. Subsequently, multiphysics coupled finite element analysis is performed on the solid model of the IBDD to conduct a comprehensive performance evaluation. In some embodiments, if the evaluation results do not meet preset conditions, the distribution of regional structural parameters can be adjusted and the solid model of the IBDD can be regenerated. Finally, based on the solid model of the IBDD with comprehensive performance evaluation, the IBDD is fabricated using laser selective melting technology.

[0161] In one possible implementation, step 706 above may specifically include: Implicit functions corresponding to the internal porous structure are constructed based on the model of the internal porous structure of the interbody fusion device. Construct the implicit function of the envelope surface corresponding to the macroscopic shape of the interbody fusion cage; Boolean operations are performed based on the implicit functions corresponding to the internal porous structure and the implicit functions of the envelope surface to fuse the internal porous structure model of the interbody fusion device with the macroscopic shape of the interbody fusion device, thereby generating a solid model of the interbody fusion device.

[0162] Here, implicit functions refer to the functional expressions used to characterize the geometric boundaries of the target structure. In this embodiment, the implicit function corresponding to the internal porous structure is used to characterize the geometric boundaries of the internal porous structure model of the interbody fusion device, and the implicit function of the envelope surface is used to characterize the geometric boundaries of the macroscopic shape of the interbody fusion device.

[0163] The macroscopic shape of the interbody fusion cage refers to the geometric shape corresponding to the overall outer contour of the cage. The macroscopic shape defines the overall shape range of the interbody fusion cage to facilitate integration with the internal porous structure model.

[0164] The envelope surface refers to the surface used to characterize the macroscopic outer boundary of the interbody fusion cage. The envelope surface serves as the basis for the external boundary of the cage during subsequent fusion with the internal porous structure.

[0165] Boolean operations refer to the process of performing calculations based on the combination relationships between different geometric objects. In this embodiment, Boolean operations are used to fuse the porous internal structure model of the interbody fusion device with the macroscopic shape of the interbody fusion device to form a unified solid model.

[0166] For example, in one specific implementation, the internal porous structure model of the interbody fusion cage can first be expressed as a function to form an implicit function corresponding to the internal porous structure; then, the macroscopic shape of the interbody fusion cage can be expressed as a function to form an implicit function of the envelope surface; finally, Boolean operations are used to fuse the internal porous structure and the macroscopic shape to generate a solid model of the interbody fusion cage. Thus, the resulting model can simultaneously possess the characteristics of the internal porous structure and the overall external morphology.

[0167] In one possible implementation, the method further includes: Multiphysics coupled finite element analysis was performed on the solid model of the interbody fusion device to simulate the stress distribution, peak stress and micromotion between the interbody fusion device and the vertebral endplate under complex physiological loads. Based on stress distribution, peak stress, and micromotion, a comprehensive performance evaluation of the interbody fusion device physical model was conducted.

[0168] Multiphysics coupled finite element analysis (MPA) refers to the process of numerically analyzing a solid model of an interbody fusion device under complex physiological load conditions. In this embodiment, MPA is used to evaluate the force and micromotion between the solid model of the interbody fusion device and the vertebral endplate.

[0169] Complex physiological loads refer to load conditions used to simulate the actual stress state after interbody fusion device implantation. Complex physiological loads are used as external action conditions in finite element analysis.

[0170] For example, in one specific implementation, the solid model of the interbody fusion device can be used for finite element analysis first, and the corresponding stress distribution, peak stress and micromotion can be obtained under complex physiological load conditions. Then, based on the above results, a comprehensive performance evaluation of the solid model of the interbody fusion device can be carried out to determine whether the solid model meets the requirements for subsequent use.

[0171] In one possible implementation, if the stress distribution, peak stress, or interface micromotion does not meet the preset conditions, the distribution of regional structural parameters is adjusted, and the solid model of the interbody fusion device is regenerated.

[0172] In this embodiment, when the stress distribution, peak stress, or fretting conditions do not meet the preset conditions, the distribution of regional structural parameters can be adjusted. Subsequently, based on the adjusted regional structural parameter distribution, parametric modeling and solid model generation are performed again to regenerate the solid model of the interbody fusion device. In this way, the model can be adjusted and generated again when the comprehensive performance evaluation results do not meet the requirements.

[0173] For example, in one specific implementation, if the finite element analysis results show that the peak stress in a certain area is large or the micro-motion does not meet the preset conditions, the distribution of regional structural parameters in that area can be adjusted, and the solid model of the interbody fusion device can be regenerated based on the adjusted distribution of regional structural parameters.

[0174] In one possible implementation, the method further includes: Based on a physical model of an interbody fusion device that has undergone comprehensive performance evaluation, an interbody fusion device was fabricated using laser selective melting technology.

[0175] In one possible implementation, before fabricating the interbody fusion device using laser selective melting technology, the following steps are also included: Optimize printing direction, support structure, and process parameters.

[0176] In one possible implementation, the intervertebral fusion device is a titanium alloy fusion device.

[0177] In one possible implementation, the internal pore permeability of the interbody fusion device is greater than 85%.

[0178] In one possible implementation, the dimensional accuracy of the interbody fusion device is controlled within ±0.1 mm.

[0179] In summary, in this embodiment, an implicit function corresponding to the internal porous structure of the interbody fusion cage is first constructed based on the model of the porous structure inside the cage. This function is then combined with the implicit function of the envelope surface corresponding to the macroscopic shape of the interbody fusion cage to generate a solid model of the cage through Boolean operations. Subsequently, multiphysics coupled finite element analysis is performed on the solid model of the interbody fusion cage, and a comprehensive performance evaluation is conducted based on stress distribution, peak stress, and micromotion. If the evaluation results do not meet the preset conditions, the distribution of regional structural parameters can be adjusted, and a new solid model of the interbody fusion cage can be generated. Based on this, the interbody fusion cage is fabricated using laser selective melting technology, based on the solid model of the interbody fusion cage with comprehensive performance evaluation. Thus, a complete process from generating the internal porous structure model, constructing the solid model, performance evaluation to final fabrication can be achieved.

[0180] See Figure 9 A schematic diagram of a model generation device for an intervertebral fusion device provided in this application is shown; for ease of explanation, only the parts related to the embodiments of this application are shown.

[0181] The interbody fusion cage model generation device 900 includes: The acquisition module 901 is used to acquire vertebral body image data and determine the target vertebral body region based on the vertebral body image data; The distribution determination module 902 is used to determine the spatial distribution of bone density based on the target vertebral body region, and to determine the spatial distribution of elastic modulus corresponding to the target vertebral body region based on the spatial distribution of bone density. The parameter determination module 903 is used to determine the regional seed point distribution density and regional structural parameter distribution of the porous structure of the interbody fusion device based on the spatial distribution of elastic modulus. The regional structural parameter distribution includes at least one of the porous unit's pore diameter, wire diameter, unit boundary thickness, and unit inscribed circle diameter. The partitioning module 904 is used to spatially partition the design space in the target cone region based on the distribution density of regional seed points, and obtain multiple porous elements. Modeling module 905 is used to parametrically model multiple porous units based on the distribution of regional structural parameters, and generate a porous structure model of the intervertebral fusion device.

[0182] In this embodiment of the application, the model generation device 900 for intervertebral fusion cages further includes a solid model generation module, which is used for: Implicit functions corresponding to the internal porous structure are constructed based on the model of the internal porous structure of the interbody fusion device. Construct the implicit function of the envelope surface corresponding to the macroscopic shape of the interbody fusion cage; Boolean operations are performed based on the implicit functions corresponding to the internal porous structure and the implicit functions of the envelope surface to fuse the internal porous structure model of the interbody fusion device with the macroscopic shape of the interbody fusion device, thereby generating a solid model of the interbody fusion device.

[0183] In this embodiment, the interbody fusion cage model generation device 900 further includes a comprehensive performance evaluation module, which is used for: Multiphysics coupled finite element analysis was performed on the solid model of the interbody fusion device to simulate the stress distribution, peak stress and micromotion between the interbody fusion device and the vertebral endplate under complex physiological loads. Based on stress distribution, peak stress, and micromotion, a comprehensive performance evaluation of the interbody fusion device physical model was conducted.

[0184] In this embodiment of the application, the acquisition module 901 includes: The encoding unit is used to encode vertebral body image data and extract vertebral body features at different levels; The feature fusion unit is used to fuse vertebral body features at different levels based on multi-scale skip connections to obtain fused vertebral body features. The decoding unit is used to decode the fused vertebral body features to obtain the vertebral body segmentation result; The region determination unit is used to determine the target vertebral region based on the vertebral segmentation results.

[0185] In this embodiment of the application, the distribution determination module 902 includes: The bone mineral density determination unit is used to map the gray values ​​of each location within the target vertebral body region based on the vertebral body segmentation results, so as to obtain the bone mineral density values ​​corresponding to each location within the target vertebral body region. The bone mineral density analysis unit is used to quantify regional features based on the bone mineral density values ​​corresponding to each location within the target vertebral body region, and obtain the vertebral body bone mineral density analysis results. The three-dimensional coordinate establishment unit is used to establish the three-dimensional coordinates of each location within the target vertebral body region based on the spatial parameters of vertebral body bone density analysis results and vertebral body imaging data. The point cloud generation unit is used to assign corresponding bone density attribute values ​​to each location within the target vertebral body region, thereby obtaining a three-dimensional point cloud model that characterizes the spatial distribution features of bone density within the vertebral body.

[0186] In this embodiment of the application, the distribution determination module 902 includes: The model input unit is used to input a three-dimensional point cloud model representing the spatial distribution characteristics of bone density inside the vertebral body into the dual-path network; The local feature extraction unit is used to extract the microstructural features of trabecular bone through local paths; The global feature extraction unit is used to extract long-distance dependent features through global paths; The modulus distribution output unit is used to output the spatial distribution of elastic modulus corresponding to the target vertebral region based on the microstructural features and long-distance dependence features of the trabecular bone.

[0187] In this embodiment of the application, the parameter determination module 903 includes: The seed point distribution density determination unit is used to determine the regional seed point distribution density of each region based on the bone density values ​​of each region in the 3D point cloud model. Correspondingly, the subdivision module 904 includes: The seed point deployment unit is used to deploy seed points in the design space of the target cone region according to the seed point distribution density of each region. Spatial subdivision elements are used to introduce boundary constraints and perform Thiessen polygon subdivision on the seed points after they are laid out, resulting in multiple independent and tessellated porous elements.

[0188] In this embodiment of the application, the parameter determination module 903 includes: The quantitative relationship establishment unit is used to establish a quantitative relationship model between bone density and ideal structural stiffness; The boundary parameter determination element is used to determine the element boundary thickness and the inscribed circle diameter of each region based on the spatial distribution of the elastic modulus and the quantization relationship model. The aperture and wire diameter determination unit is used to determine the aperture and wire diameter of each region based on the unit boundary thickness and the inscribed circle diameter of each region. The regional structural parameter distribution determination unit is used to determine the regional structural parameter distribution based on the corresponding aperture, wire diameter, unit boundary thickness, and inscribed circle diameter of each region.

[0189] In this embodiment of the application, the modeling module 905 includes: The model input unit is used to input multiple porous units and the distribution of regional structural parameters into the biomechanical topology optimization calculator. The iterative computation unit is used to perform iterative computations on multiple porous units based on the principle of cellular automata and through a biomechanical topology optimization operator; The morphology connection optimization unit is used to optimize the morphology and connection method of multiple porous units while ensuring structural integrity, so as to maximize porosity and pore space for bone tissue ingrowth. The model output unit is used to output the porous structure model inside the intervertebral fusion device.

[0190] The interbody fusion model generation device 900 provided in this application embodiment can be applied to the interbody fusion model generation method provided in the foregoing embodiment. For details, please refer to the description of the interbody fusion model generation method provided in the above embodiment, which will not be repeated here.

[0191] See Figure 10This illustration shows a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device 1000 of this embodiment includes: at least one processor 1010 (… Figure 10 The processor 1010 contains only one of the following: a memory 1020 and a computer program 1021 stored in the memory 1020 and executable on the processor 1021. When the processor executes the computer program 1021, it implements the above-described method for generating a model of an intervertebral fusion device.

[0192] The electronic device 1000 may be a server, a physical server, or a computing device, etc. This electronic device may include, but is not limited to, a processor 1010 and a memory 1020. Those skilled in the art will understand that... Figure 10 This is merely an example of electronic device 1000 and does not constitute a limitation on electronic device 1000. It may include more or fewer components than shown, or combine certain components, or use different components.

[0193] The processor 1010 can be a Central Processing Unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0194] In some embodiments, the memory 1020 may be an internal storage unit of the electronic device 1000, such as a hard disk or memory of the electronic device 1000. In other embodiments, the memory 1020 may be an external storage device of the electronic device 1000, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the electronic device 1000. Furthermore, the memory 1020 may include both internal and external storage units of the electronic device 1000. The memory 1020 is used to store operating systems, applications, boot loaders, data, and other programs, such as the program code of computer programs. The memory 1020 can also be used to temporarily store data that has been output or will be output.

[0195] In specific implementations, the processor 1010, memory 1020, and computer program 1021 described in the embodiments of this application can execute the embodiments of the model generation method for the intervertebral fusion device of this application, which will not be repeated here.

[0196] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0197] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0198] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0199] In the embodiments provided in this application, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings or direct couplings or communication connections may be through some interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0200] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0201] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0202] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0203] The implementation of all or part of the processes in the methods of the above embodiments can also be accomplished by a computer program product. When the computer program product is run on an electronic device, the electronic device can implement the steps in the various method embodiments described above.

[0204] The embodiments described above are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for generating a model of an interbody fusion device, characterized in that, The method includes: Acquire vertebral body imaging data and determine the target vertebral body region based on the vertebral body imaging data; The spatial distribution of bone density is determined based on the target vertebral body region, and the spatial distribution of elastic modulus corresponding to the target vertebral body region is determined based on the spatial distribution of bone density. The distribution density of regional seed points and the distribution of regional structural parameters of the porous structure of the interbody fusion device are determined based on the spatial distribution of the elastic modulus, wherein the distribution of regional structural parameters includes at least one of the pore diameter, wire diameter, unit boundary thickness and inscribed circle diameter of the porous unit; Based on the distribution density of seed points in the region, the design space in the target cone region is spatially divided to obtain multiple porous units; Based on the distribution of the regional structural parameters, the multiple porous units are parametrically modeled to generate a porous structure model inside the interbody fusion device. Wherein, determining the spatial distribution of elastic modulus corresponding to the target vertebral body region based on the spatial distribution of bone density includes: A three-dimensional point cloud model characterizing the spatial distribution of bone density within the vertebral body is input into a dual-path network; Extracting trabecular microstructure features using local pathways; Long-distance dependency features are extracted through global path extraction. These long-distance dependency features refer to feature information used to characterize the relationship between different locations within the target vertebral region. Based on the trabecular microstructure features and the long-distance dependence features, the spatial distribution of elastic modulus corresponding to the target vertebral region is output.

2. The method as described in claim 1, characterized in that, The method further includes: Based on the internal porous structure model of the interbody fusion device, an implicit function corresponding to the internal porous structure is constructed. Construct the implicit function of the envelope surface corresponding to the macroscopic shape of the interbody fusion cage; Boolean operations are performed based on the implicit functions corresponding to the internal porous structure and the implicit functions of the envelope surface to fuse the internal porous structure model of the intervertebral fusion device with the macroscopic shape of the intervertebral fusion device, thereby generating a solid model of the intervertebral fusion device.

3. The method as described in claim 2, characterized in that, The method further includes: Multiphysics coupled finite element analysis was performed on the solid model of the interbody fusion device to simulate the stress distribution, peak stress and micromotion between the interbody fusion device and the vertebral endplate under complex physiological loads. Based on the stress distribution, peak stress, and micro-motion, a comprehensive performance evaluation is performed on the solid model of the intervertebral fusion device.

4. The method as described in claim 1, characterized in that, Determining the target vertebral region based on the vertebral imaging data includes: The vertebral body image data is encoded to extract vertebral body features at different levels; The vertebral body features at different levels are fused based on multi-scale skip connections to obtain the fused vertebral body features. The fused vertebral body features are decoded to obtain vertebral body segmentation results; The target vertebral region is determined based on the vertebral segmentation results.

5. The method as described in claim 4, characterized in that, Determining the spatial distribution of bone mineral density based on the target vertebral body region includes: Based on the vertebral segmentation results, the gray values ​​at each location within the target vertebral body region are mapped to obtain the bone density values ​​corresponding to each location within the target vertebral body region. Based on the bone mineral density values ​​corresponding to each location within the target vertebral body region, regional feature quantification is performed to obtain the vertebral body bone mineral density analysis results. Based on the vertebral bone density analysis results and the spatial parameters of the vertebral imaging data, three-dimensional coordinates of each location within the target vertebral body region are established. By assigning corresponding bone density attribute values ​​to each location within the target vertebral body region, a three-dimensional point cloud model characterizing the spatial distribution features of bone density within the vertebral body is obtained.

6. The method as described in claim 5, characterized in that, The determination of the regional seed point distribution density of the porous structure of the interbody fusion cage based on the spatial distribution of the elastic modulus includes: Based on the bone density values ​​of each region in the three-dimensional point cloud model, determine the distribution density of the seed points corresponding to each region. Based on the distribution density of seed points in the region, the design space in the target vertebral region is spatially partitioned, including: Based on the distribution density of seed points in each region, seed points are deployed in the design space within the target vertebral region. By introducing boundary constraints, the seed points are divided into Thiessen polygons to obtain multiple independent and tessellated porous elements.

7. The method as described in claim 6, characterized in that, The determination of the regional structural parameter distribution of the porous structure of the interbody fusion cage based on the spatial distribution of the elastic modulus includes: Establish a quantitative relationship model between bone mineral density and ideal structural stiffness; Based on the spatial distribution of the elastic modulus and the quantization relationship model, determine the element boundary thickness and the inscribed circle diameter of each region. Based on the element boundary thickness and the inscribed circle diameter of each region, determine the corresponding aperture and wire diameter for each region; The distribution of structural parameters for each region is determined based on the corresponding aperture, wire diameter, unit boundary thickness, and inscribed circle diameter.

8. The method as described in claim 7, characterized in that, The step of parametrically modeling the multiple porous units based on the distribution of the regional structural parameters to generate a porous structure model inside the interbody fusion cage includes: The plurality of porous units and the distribution of the regional structural parameters are input into the biomechanical topology optimization calculator; Based on the principle of cellular automata, the biomechanical topology optimization operator iteratively calculates the multiple porous units; While ensuring structural integrity, the morphology and connection method of the unit walls of the multiple porous units are optimized to maximize porosity and pore space for bone tissue ingrowth. Output the porous structure model of the intervertebral fusion device.

9. A model generation device for an interbody fusion device, characterized in that, include: The acquisition module is used to acquire vertebral body image data and determine the target vertebral body region based on the vertebral body image data; The distribution determination module is used to determine the spatial distribution of bone density based on the target vertebral body region, and to determine the spatial distribution of elastic modulus corresponding to the target vertebral body region based on the spatial distribution of bone density. The parameter determination module is used to determine the regional seed point distribution density and regional structural parameter distribution of the porous structure of the interbody fusion device based on the spatial distribution of the elastic modulus, wherein the regional structural parameter distribution includes at least one of the following: pore diameter, wire diameter, unit boundary thickness, and inscribed circle diameter of the unit. The partitioning module is used to spatially partition the design space corresponding to the target cone region based on the distribution density of seed points in the region, thereby obtaining multiple porous units. The modeling module is used to perform parametric modeling of the multiple porous units based on the distribution of the regional structural parameters, and generate a porous structure model inside the interbody fusion device. The distribution determination module includes: The model input unit is used to input a three-dimensional point cloud model representing the spatial distribution characteristics of bone density inside the vertebral body into the dual-path network; The local feature extraction unit is used to extract the microstructural features of trabecular bone through local paths; The global feature extraction unit is used to extract long-distance dependency features through a global path. The long-distance dependency features refer to feature information used to characterize the relationship between different locations within the target vertebral region. The modulus distribution output unit is used to output the spatial distribution of elastic modulus corresponding to the target vertebral body region based on the trabecular microstructure features and the long-distance dependence features.

Images