Arch expander adaptation simulation system and method based on 5D bionic technology

Abstract

The invention provides a 5D bionic technology-based arch expander adaptive simulation system and method, and the method comprises the steps: generating an oral cavity three-dimensional model according to pre-collected oral scanning data, and carrying out the organ separation of the oral cavity three-dimensional model, so as to obtain an oral cavity organ model group; performing targeted organ model recombination on the oral organ model group based on a preset arch expander adapter list to obtain a recombined oral model; performing multi-modal data registration on the recombined oral cavity model and an oral cavity CBCT obtained in advance to obtain a registered oral cavity model; carrying out anchorage nail point location matching on the registered oral cavity model to obtain anchorage nail point location information, and initializing a bow expander model and an anchorage nail shoulder model according to the anchorage nail point location information; according to the registered oral cavity model, generating a bow expansion screw model and an anchorage tooth band model for the bow expansion device model, and according to the bow expansion device model, the bow expansion screw model, the anchorage nail shoulder model and the anchorage tooth band model, generating a bow expansion adaptive model.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of expansion arch device adaptation, and particularly relates to an expansion arch device adaptation simulation system and method based on 5D bionic technology. BACKGROUND

[0002] With the progress of orthodontic technology, skeletal expansion arch treatment is widely used in the correction of maxillary narrowness, dental arch crowding and other deformities. The traditional expansion arch device design mainly relies on two-dimensional drawings and doctor experience, that is, the doctor manually designs the expansion arch device based on oral scan data, and displays the parameters such as the position of the anchorage pin and the size of the expansion arch device through two-dimensional drawings. However, the traditional method has significant limitations. First, two-dimensional drawings cannot fully present three-dimensional anatomical structures, and doctors have difficulty in accurately judging the implantation angle and depth of the anchorage pin, which may lead to improper implantation position due to visual angle deviation, thereby increasing the risk of surgery. Second, the adaptability of the expansion arch device to the mucosa of the palatine vault is poor, and it relies on standardized templates, which are difficult to fit complex anatomical shapes such as high arch palatal cover, and may cause soft tissue damage or uneven distribution of correction force, which may result in low efficiency of expansion arch device adaptation. SUMMARY

[0003] (I) Technical problems solved In view of the deficiencies in the prior art, the present application provides an expansion arch device adaptation simulation system and method based on 5D bionic technology, which has the advantages of intuitively displaying the three-dimensional relationship between the expansion arch device and the oral cavity, high design efficiency of the expansion arch device, and high adaptation degree of the expansion arch device, and solves the problems of inaccurate three-dimensional space judgment, low individuality and low design efficiency in the design and adaptation process of the expansion arch device.

[0004] (II) Technical solutions To achieve the above-mentioned purpose, the present application provides the following technical solutions: The present application provides an expansion arch device adaptation simulation system based on 5D bionic technology, comprising a model separation module, a model recombination module, a multi-modal registration module, an anchorage pin matching module and an expansion arch device adaptation module, wherein: The model separation module is used to generate a three-dimensional oral cavity model according to pre-acquired oral scan data, and to separate the three-dimensional oral cavity model at multiple levels to obtain a set of oral organ models; The model recombination module is used to recombine the target organ models of the set of oral organ models based on a pre-set expansion arch device adapter organ list to obtain a recombined oral cavity model; The multi-modal registration module is used to register the recombined oral cavity model with pre-acquired oral CBCT to obtain a registered oral cavity model; The anchorage pin matching module is used to match the anchorage pin point position of the registered oral cavity model to obtain anchorage pin point position information, and to initialize the expansion arch device model and the anchorage pin shoulder model according to the anchorage pin point position information. an expander adapter module configured to generate an expander screw model and a bracket tube model based on the registered dental model, and generate an expander adapter model based on the expander model, the expander screw model, the anchorage abutment model, and the bracket tube model.

[0005] According to one of the preferred embodiments of the present application, the model separation module comprises the following steps when performing multi-level organ separation on the three-dimensional dental model to obtain a dental organ model group: performing multi-view projection on the three-dimensional dental model to obtain a dental projection atlas; performing block feature mapping and position convolution on the dental projection atlas to obtain a dental two-dimensional feature set, and performing two-dimensional organ segmentation on the dental projection atlas based on the dental two-dimensional feature set to obtain a two-dimensional segmentation label; projecting the two-dimensional segmentation label on the three-dimensional dental model to obtain a primary segmentation label; performing voxel weighted convolution on the three-dimensional dental model based on the primary segmentation label to obtain dental voxel features; performing transpose convolution and skip connection on the dental voxel features to obtain a secondary segmentation label; performing local attention feature extraction and global feature aggregation on the three-dimensional dental model based on the secondary segmentation label to obtain aggregated model point features, and performing multi-level organ separation on the three-dimensional dental model based on the aggregated model point features to obtain a dental organ model group.

[0006] According to another preferred embodiment of the present application, the model separation module comprises the following steps when performing local attention feature extraction and global feature aggregation on the three-dimensional dental model based on the secondary segmentation label to obtain aggregated model point features: performing point cloud sampling on the three-dimensional dental model, and performing point cloud rigid registration on the three-dimensional dental model after point cloud sampling to obtain a rigid dental point cloud; and performing point feature extraction on the rigid dental point cloud based on the secondary segmentation label to obtain dental point cloud features; performing local attention feature extraction on the dental point cloud features based on the neighborhood relative coordinates of the dental point cloud features to obtain point cloud attention features; performing feature fusion on the point cloud attention features based on the skip connection method, and performing global attention weighted aggregation on the point cloud attention features after feature fusion to obtain aggregated model point features.

[0007] According to another preferred embodiment of the present application, the multi-modal registration module comprises the following steps when performing multi-modal data registration of the recombined oral model and the pre-acquired oral CBCT to obtain a registered oral model: performing three-dimensional volume conversion on the pre-acquired oral CBCT to obtain an oral three-dimensional volume; extracting a bony region from the oral three-dimensional volume based on a threshold segmentation method, and performing local feature extraction on the bony region to obtain an oral volume feature; performing normal estimation and down-sampling on the recombined oral model to obtain an oral model feature; performing geometric feature matching and geometric constraint registration on the oral volume feature and the oral model feature to obtain a primary registration relationship; establishing a distance field function based on the oral three-dimensional volume, and performing parameter updating on the distance field function based on the primary registration relationship to obtain an updated distance field function; performing non-rigid geometric registration on the oral volume feature and the oral model feature based on the updated distance field function to obtain a secondary registration relationship; mapping the oral three-dimensional volume into the recombined oral model based on the secondary registration relationship to obtain a fused oral model; performing spatial coincidence verification on the fused oral model, and performing registration optimization on the fused oral model according to the verification result to obtain a registered oral model.

[0008] According to another preferred embodiment of the present application, the bracket pin matching module comprises the following steps when performing bracket pin point position matching on the registered oral model to obtain bracket pin point position information: performing voxel quantization on the registered oral model, and generating a bone density distribution map, a bone thickness distribution map and a soft tissue thickness distribution map of the registered oral model according to the voxel quantization result; screening a candidate point position region from the registered oral model based on the bone density distribution map, the bone thickness distribution map and the soft tissue thickness distribution map; extracting an oral key structure from the registered oral model, calculating a key structure distance according to the oral key structure, and establishing a point position constraint relationship based on the key structure distance; performing grid point sampling on the candidate point position region to obtain a sampling point set, and performing global weighted scoring on each sampling point in the sampling point set based on the point position constraint relationship to obtain a point position score set; screening a bracket pin point position group from the sampling point set according to the point position score set, and performing heuristic optimization and information labeling on the bracket pin point position group to obtain bracket pin point position information.

[0009] According to another preferred embodiment of the present application, the model separation module comprises the following steps when generating the three-dimensional oral cavity model according to the pre-acquired oral scanning data: performing point cloud noise filtering on the pre-acquired oral scanning data to obtain denoised oral scanning data; performing neighborhood search on the denoised oral scanning data to obtain neighborhood point cloud data, and performing plane fitting and feature decomposition on the denoised oral scanning data based on the neighborhood point cloud data to obtain primary normal vector features; performing uniformization operation on the primary normal vector features to obtain standard normal vector features; performing local topological analysis on the denoised oral scanning data to obtain boundary point cloud, and constructing a boundary ring based on the boundary point cloud; performing surface fitting on the boundary ring based on the standard normal vector features to obtain a fitted surface, and performing point cloud filling on the denoised oral scanning data using the fitted surface to obtain completed oral scanning data; performing triangular faceting operation on the completed oral scanning data to obtain a three-dimensional oral cavity model.

[0010] According to another preferred embodiment of the present application, the model recombination module comprises the following steps when performing targeted organ model recombination on the oral organ model group based on the pre-set expander adapter organ list to obtain a recombined oral cavity model: extracting Gaussian curvature features of the oral organ model group, and performing redundancy detection and volume culling on the oral organ model group based on the Gaussian curvature features to obtain a denoised organ model group; performing organ screening on the denoised organ model group based on the expander adapter organ list to obtain an adapter organ model group; performing coordinate alignment on the adapter organ model group based on organ anatomical relationship constraints to obtain an aligned organ model; performing bilateral smoothing on the aligned organ model, and matching an organ interface from the bilateral smoothed aligned organ model, performing local vertex resampling on the organ interface to obtain a smoothed organ model; performing curvature optimization on the smoothed organ model, and performing triangular face topological optimization on the curvature optimized smoothed organ model to obtain an optimized organ model; performing triangular face self-intersection detection on the optimized organ model, and performing mesh surface repair on the optimized organ model according to the results of the triangular face self-intersection detection to obtain a recombined oral cavity model.

[0011] According to another preferred embodiment of the present application, the anchorage peg matching module comprises the following steps when initializing the expander model and the anchorage peg shoulder model according to the anchorage peg point information: According to the anchorage nail point position information, an anchorage nail center and an expander frame size are calculated; According to the anchorage nail point position information, an anchorage nail center and an expander frame size are calculated; Based on the elastic stress of the expander contour line, the expander contour line is updated to obtain an updated contour line; According to the registration oral model, the bone thickness and the soft tissue thickness corresponding to the anchorage nail point position information are calculated; According to the bone thickness and the soft tissue thickness, an anchorage nail model is matched, and an expander model is initialized according to the anchorage nail model, the expander frame size and the updated contour line; According to the expander model and the anchorage nail model, an initial abutment model is initialized. From the registration oral model, a palatal plane is extracted, and the initial abutment model is updated based on the palatal plane to obtain an anchorage nail abutment model.

[0012] According to another preferred embodiment of the application, when the expander fitting module generates an expander screw model and an anchorage band loop model according to the registration oral model, it includes: From the registration oral model, a palatal vault surface is extracted, and an expander screw model is initialized on the surface of the expander model based on the palatal vault surface; The registration oral model is matched with the molar to obtain an oral molar model, and a band loop model is initialized according to the oral molar model; From the expander model, an expander connecting arm is extracted, and a band loop connecting line is fitted according to the expander connecting arm and the band loop model; Based on the palatal plane of the registration oral model, the band loop connecting line is matched with the line to obtain a fitted connecting line, and the fitted connecting line is connected with the band loop model to obtain an anchorage band loop model.

[0013] The application provides an expander fitting simulation method based on 5D biomimetic technology, which includes: According to the pre-acquired oral scanning data, an oral three-dimensional model is generated, and the oral three-dimensional model is separated into multiple levels of organs to obtain an oral organ model group; Based on a preset expander fitting organ list, the oral organ model group is reorganized into a target organ model to obtain a reorganized oral model; The reorganized oral model is registered with pre-acquired oral CBCT to obtain a registration oral model; The registration oral model is subjected to anchorage peg point position matching, anchorage peg point position information is obtained, and an expansion arch device model and an anchorage peg shoulder model are initialized according to the anchorage peg point position information; An expansion arch screw model and an anchorage band ring model are generated for the registration oral model, and an expansion arch fitting model is generated according to the expansion arch device model, the expansion arch screw model, the anchorage peg shoulder model and the anchorage band ring model.

[0014] (Three) beneficial effects Compared with the prior art, the expansion arch device fitting simulation system based on 5D bionic technology has the following beneficial effects: The expansion arch device fitting simulation system based on 5D bionic technology realizes multi-level organ segmentation from two dimensions to three dimensions to point cloud subdivision by the method of triple segmentation, fuses information of three dimensions of visual semantics, voxel structure and geometric topology, makes the separation of complex oral structure more accurate and more robust, realizes separate viewing and analysis of organs, and improves the intuitiveness of expansion arch device fitting and the further registration work of doctors.

[0015] The expansion arch device fitting simulation system based on 5D bionic technology realizes accurate quantitative analysis of the oral bone structure by voxel quantification and multi-modal fusion through anchorage peg point position matching of the registration oral model, significantly improves the scientificity and safety of the anchorage peg implantation point position, ensures that the selected point meets the mechanical stability and clinical operability by combining key structure constraints and global weighted scoring, and improves the symmetry and efficiency of expansion arch force transmission by heuristic optimization, thereby improving the efficiency of expansion arch device fitting.

[0016] The expansion arch device fitting simulation system based on 5D bionic technology ensures accurate matching of the expansion arch device and the jawbone anatomical structure by initializing the expansion arch device system model based on the anchorage peg point position information, avoids implantation deviation, enhances treatment safety, generates expansion arch screws and anchorage band ring components by using the registration oral model, realizes highly personalized fitting, reduces the risk of soft tissue injury caused by palatine vault morphological variation, and thereby improves the efficiency of expansion arch device fitting. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 A structure diagram of the expansion arch device fitting simulation system based on 5D bionic technology is shown.

[0018] Figure 2 A model schematic diagram of an oral organ model group of the expansion arch device fitting simulation system based on 5D bionic technology is shown.

[0019] Figure 3A model schematic diagram of a reorganization oral model of an arch expander adaptive simulation system based on 5D bionic technology is shown.

[0020] Figure 4 A structure schematic diagram of an arch expander system of an arch expander adaptive simulation system based on 5D bionic technology is shown.

[0021] Figure 5 A flow chart of an arch expander adaptive simulation method based on 5D bionic technology is shown.

[0022] Wherein, a is a model of a tooth root organ, b is a model of a jaw bone organ, c is an arch expander model, d is a molar model of the oral cavity, e is a model of an anchorage tooth ring, f is an arch expansion connecting arm, and g is an arch expansion screw model. DETAILED DESCRIPTION

[0023] The following description is provided to enable any person skilled in the art to make and use the present application. The preferred embodiments in the following description are only examples of the present application and modifications can be made by those skilled in the art without departing from the spirit and scope of the present application. The basic principles defined in the following description can be applied to other embodiments, modifications, improvements, equivalents and other technical solutions without departing from the spirit and scope of the present application.

[0024] It can be understood that the term "one" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of one element can be one, and in another embodiment, the number of the element can be multiple, and the term "one" cannot be understood as a limitation on the number.

[0025] Embodiment one: Please refer to Figure 1 The present application discloses an arch expander adaptive simulation system based on 5D bionic technology, which mainly comprises a model separation module, a model reorganization module, a multi-modal registration module, an anchorage nail matching module and an arch expander adaptive module, wherein: The model separation module is used for generating an oral cavity three-dimensional model according to pre-acquired oral scanning data, and performing multi-level organ separation on the oral cavity three-dimensional model to obtain an oral organ model group.

[0026] In detail, the oral scanning data refers to oral cavity scanning data, which can be oral cavity three-dimensional point cloud data collected by an intraoral scanner. When the intraoral scanner projects structured light into the oral cavity, the oral scanning data is recorded by capturing the reflected light signal. The oral scanning data is data in the form of point cloud for recording the three-dimensional geometric information of the tooth surface. The oral cavity three-dimensional model is a model obtained after three-dimensional modeling according to the oral scanning data.

[0027] In the embodiment of the present application, the model separation module comprises the following steps when generating the oral cavity three-dimensional model according to the pre-acquired oral scan data: performing point cloud noise filtering on the pre-acquired oral scan data to obtain denoised oral scan data; performing neighborhood search on the denoised oral scan data to obtain neighborhood point cloud data, and performing plane fitting and feature decomposition on the denoised oral scan data based on the neighborhood point cloud data to obtain primary normal vector features; performing uniformization operation on the primary normal vector features to obtain standard normal vector features; performing local topological analysis on the denoised oral scan data to obtain boundary point cloud, and constructing a boundary ring based on the boundary point cloud; performing surface fitting on the boundary ring based on the standard normal vector features to obtain a fitted surface, and performing point cloud filling on the denoised oral scan data using the fitted surface to obtain completed oral scan data; performing triangularization operation on the completed oral scan data to obtain an oral cavity three-dimensional model.

[0028] The point cloud noise filtering can be performed using a neighborhood statistical filtering algorithm or a moving least square method, the neighborhood search refers to selecting neighboring point clouds within a preset radius range to form neighborhood point cloud data, the plane fitting refers to constructing a covariance matrix corresponding to the neighborhood point cloud data to fit a corresponding point cloud plane, the feature decomposition of the plane fitted point cloud plane can be performed using a principal component analysis method to obtain corresponding primary normal vector features, and the uniformization operation refers to unifying the signs of the primary normal vector features, for example, unifying the directions of the standard normal vector features from a viewpoint direction.

[0029] In detail, the local topological analysis refers to analyzing the geometric and topological characteristics of each point cloud and its neighborhood in the denoised oral scan data to identify point cloud defects corresponding to the denoised oral scan data, the local topological analysis can be performed using a K-nearest neighbor algorithm or a covariance algorithm, the boundary ring is constructed by performing connectivity analysis on the boundary point cloud, the surface fitting can be performed using a weighted least square fitting surface constrained by the standard normal vector features, the point cloud filling refers to adding point cloud data to the surface fitted surface, and the triangularization operation can be performed using a Delaunay triangulation scheme.

[0030] Specifically, the obtained oral cavity three-dimensional model is a whole model, and in order to fit the oral cavity with an expansion arch device, it is necessary to clearly understand the specific conditions of the dental crown, dental root, and jaw bone tissues in the oral cavity, therefore, it is necessary to separate the organs of the oral cavity three-dimensional model for further analysis.

[0031] In detail, the model separation module comprises the following steps when performing multi-level organ separation on the oral cavity three-dimensional model to obtain an oral cavity organ model group: Projecting the oral cavity three-dimensional model from multiple perspectives to obtain an oral cavity projection atlas; Performing block feature mapping and position convolution on the oral cavity projection atlas to obtain an oral cavity two-dimensional feature set, and performing two-dimensional organ segmentation on the oral cavity projection atlas based on the oral cavity two-dimensional feature set to obtain a two-dimensional segmentation label; Projecting the oral cavity three-dimensional model using the two-dimensional segmentation label to obtain a primary segmentation label; Performing voxel weighted convolution on the oral cavity three-dimensional model based on the primary segmentation label to obtain oral cavity voxel features; Performing transpose convolution and skip connection on the oral cavity voxel features to obtain a secondary segmentation label; Based on the secondary segmentation label, performing local attention feature extraction and global feature aggregation on the oral cavity three-dimensional model to obtain aggregated model point features, and performing multi-level organ separation on the oral cavity three-dimensional model based on the aggregated model point features to obtain an oral cavity organ model group.

[0032] The multiple perspective projection refers to projecting the oral cavity three-dimensional model from multiple predetermined angles. Each oral cavity projection image in the oral cavity projection atlas can be a front view projection, a top view projection, or a side view projection of the oral cavity three-dimensional model at different angles, or a slice projection of the oral cavity three-dimensional model after layer-by-layer cutting. The block feature mapping and position convolution on the oral cavity projection atlas to obtain an oral cavity two-dimensional feature set comprises image blocking and high-dimensional projection on each oral cavity projection image in the oral cavity projection atlas to obtain an oral cavity projection block feature set, and position encoding and self-attention encoding on the oral cavity projection block feature set to obtain an oral cavity two-dimensional feature set.

[0033] Specifically, the two-dimensional organ segmentation on the oral cavity projection atlas based on the oral cavity two-dimensional feature set to obtain a two-dimensional segmentation label refers to cross-feature fusion or splicing fusion of the oral cavity two-dimensional features and user input prompt features, generating a segmentation mask using the fused features, and segmenting the oral cavity projection atlas using the segmentation mask to obtain a two-dimensional segmentation label. The two-dimensional segmentation label includes the position area of different organs in each oral cavity projection image and the corresponding label and confidence. The label projection refers to inversely projecting the label and confidence onto the three-dimensional oral cavity model according to the projection direction of the multiple perspective projection.

[0034] Specifically, when performing voxel weighting convolution on the oral cavity three-dimensional model based on the primary segmentation label to obtain oral cavity voxel features, the model separation module comprises: performing sparse voxel conversion on the oral cavity three-dimensional model to obtain oral cavity model voxels, and performing multi-scale sparse convolution on the oral cavity model voxels to obtain sparse voxel features; performing multi-channel feature fusion on the sparse voxel features to obtain fused voxel features; constructing a weighted loss function of the fused voxel features based on the primary segmentation label, and updating the fused voxel features into oral cavity voxel features based on the weighted loss function.

[0035] wherein, the sparse voxelization is a process of converting three-dimensional point cloud or triangular mesh data into sparse voxel representation, and the sparse voxelization only stores non-empty voxels to reduce memory and computing overhead, the multi-scale sparse convolution is a multi-scale convolution operation applied on the sparse voxel grid to extract features of different spatial scales, the multi-channel feature fusion refers to fusing features after different scale sparse convolution, the cross-entropy loss function can be used to construct the weighted loss function, that is, the corresponding regions and classes in the primary segmentation label are used as region weights, spatial weights and class weights for weighting operation, and the back propagation algorithm can be used to update the fused voxel features into oral cavity voxel features based on the weighted loss function.

[0036] The transposed convolution up-samples the oral cavity voxel features into high-resolution segmentation masks and confidence, and the skip connection can realize feature alignment, and the skip connection is to directly transmit the shallow features in the encoder into the decoder, fuse with the up-sampled features, and thus enhance the detail retention capability of the segmentation model.

[0037] In detail, when performing local attention feature extraction and global feature aggregation on the oral cavity three-dimensional model based on the secondary segmentation label to obtain aggregated model point features, the model separation module comprises: performing point cloud sampling on the oral cavity three-dimensional model, and performing point cloud rigid registration on the oral cavity three-dimensional model after point cloud sampling to obtain rigid oral cavity point cloud; and performing point feature extraction on the rigid oral cavity point cloud based on the secondary segmentation label to obtain oral cavity point cloud features; performing local attention feature extraction on the oral cavity point cloud features based on the neighborhood relative coordinates of the oral cavity point cloud features to obtain point cloud attention features; performing feature fusion on the point cloud attention features in a skip connection manner, and performing global attention weighted aggregation on the point cloud attention features after feature fusion to obtain aggregated model point features.

[0038] The uniform sampling refers to maintaining the global uniformity of the point cloud by using Farthest Point Sampling (FPS) or Poisson Disk sampling, the point cloud rigid registration refers to aligning the point cloud of the oral three-dimensional model with a standard oral template or bone structure by using a feature matching-based iterative closest point algorithm or a feature-based registration method, and the point feature extraction refers to extracting the class features and confidence features of each point cloud from the secondary segmentation label and combining the position features and normal vector features of the point cloud for feature extraction.

[0039] In detail, the local attention feature extraction includes local neighborhood search on each point of the oral point cloud feature, and calculating geometric coding of relative coordinates and normal difference of each oral point cloud feature according to the result of the local neighborhood search, using a linear projection method to perform geometric attention embedding on the oral point cloud feature, and updating the feature of the oral point cloud feature based on the attention weight, the feature fusion of the point cloud attention feature refers to multi-scale fusion from low-level geometric detail features to high-level semantic context, through local attention feature extraction and global feature aggregation, the position-aware attention method can be used to capture the relationship between the microstructures of each organ, thereby improving the accuracy of organ separation, and the organ separation refers to attention decoding of the oral three-dimensional model based on the aggregated model point feature to obtain the regions and classes corresponding to each organ in the oral three-dimensional model, and splitting the oral three-dimensional model according to the regions and classes to obtain an oral organ model group composed of multiple oral organ models.

[0040] Referring to Figure 2 is a model schematic diagram of the oral organ model group after multi-level organ separation, wherein a is a model of a tooth root organ, and b is a model of a jaw bone organ, the organ separation of the oral three-dimensional model is realized by a triple segmentation method, realizing multi-level organ segmentation from two-dimensional to three-dimensional to point cloud subdivision, fusing information in three dimensions of visual semantics, voxel structure and geometric topology, making the oral complex structure separation more accurate and more robust, and realizing separate viewing and analysis of organs, improving the intuitiveness of the expansion arch adapter fitting and facilitating the further registration work of doctors.

[0041] The model recombination module is configured to perform targeted organ model recombination on the oral organ model group based on a preset expansion arch adapter organ list to obtain a recombined oral model.

[0042] In the embodiment of the present application, in the process of model recombination of the oral organ model group, due to metal reflection, saliva reflection, scanning dead angle or structure overlap, redundant model volume may appear, and therefore when re-performing model recombination, these models need to be removed, only organ models required in the expansion arch adapter fitting are retained, and the final recombined oral model is recombined.

[0043] In the embodiment of the present application, when the model recombination module performs targeted organ model recombination on the oral organ model group based on the preset expansion arch adapter organ list to obtain a recombined oral model, the following steps are included: Gaussian curvature features of the oral organ model group are extracted, and the oral organ model group is subjected to redundancy detection and volume removal based on the Gaussian curvature features to obtain a denoised organ model group; The denoised organ model group is subjected to organ screening based on the expansion arch adapter organ list to obtain an adapter organ model group; The adapter organ model group is subjected to coordinate alignment based on organ anatomical relationship constraints to obtain an aligned organ model; The aligned organ model is subjected to bilateral smoothing, and an organ interface is matched from the aligned organ model after bilateral smoothing, the organ interface is subjected to local vertex resampling to obtain a smoothed organ model; The smoothed organ model is subjected to curvature optimization, and the smoothed organ model after curvature optimization is subjected to triangular face topology optimization to obtain an optimized organ model; The optimized organ model is subjected to triangular face self-intersection detection, and the optimized organ model is subjected to grid face repair according to the result of triangular face self-intersection detection to obtain a recombined oral model.

[0044] In detail, please refer to Figure 3 is a model schematic diagram of the recombined oral model, the Gaussian curvature features can reflect sharp edges and redundant areas in the oral organ model group, and in combination with the processing result of the previous point cloud noise filtering, sharp areas generated by metal reflection or saliva reflection can be filtered and removed, the organ screening refers to screening organ models belonging to the expansion arch adapter organ list from the denoised organ model group as the adapter organ model group.

[0045] The organ anatomical relationship constraint refers to coordinate alignment based on constraints such as crown, root alignment, geometric continuity of the maxillary bone and the bottom of the nasal cavity, and position coordinate relationship of the oral cavity scanning data; organ interface can be matched by using grid Boolean operation or distance field; local vertex resampling can be performed by using a moving least square method; curvature optimization can be performed by using Laplace smoothing or mean curvature flow; the triangular face topology optimization is performed by detecting non-manifold edges or island faces of the triangular face, and by using edge flipping or vertex merging; self-intersection detection of the triangular face can be performed by using a ray projection algorithm; and the mesh face repair refers to repairing self-intersection regions in the self-intersection detection of the triangular face by using local re-meshing.

[0046] A multi-modal registration module is configured to perform multi-modal data registration between the reconstructed oral cavity model and a pre-acquired oral cavity CBCT to obtain a registered oral cavity model.

[0047] The oral cavity CBCT (Cone Beam Computed Tomography) is a three-dimensional imaging technology for the oral cavity and maxillofacial region. By using a cone-shaped X-ray beam to scan the oral cavity, teeth, jaw bones and surrounding structures of a patient, a high-resolution three-dimensional image is generated. By performing multi-modal data registration between the reconstructed oral cavity model and the oral cavity CBCT, the position of the alveolar bone and nerves in the oral cavity CBCT and the shape of the crown surface of the reconstructed oral cavity model can be combined to design anchorage pin points, and the limitations of a single modality are overcome, thereby improving the accuracy of the expansion arch adapter.

[0048] In detail, when performing multi-modal data registration between the reconstructed oral cavity model and a pre-acquired oral cavity CBCT to obtain a registered oral cavity model, the multi-modal registration module includes the following steps: performing three-dimensional volume conversion on the pre-acquired oral cavity CBCT to obtain an oral cavity three-dimensional volume; extracting a bony region from the oral cavity three-dimensional volume based on a threshold segmentation method, and performing local feature extraction on the bony region to obtain an oral cavity volume feature; performing normal estimation and downsampling on the reconstructed oral cavity model to obtain an oral cavity model feature; performing geometric feature matching and geometric constraint registration on the oral cavity volume feature and the oral cavity model feature to obtain a primary registration relationship; establishing a distance field function based on the oral cavity three-dimensional volume, and updating the distance field function based on the primary registration relationship to obtain an updated distance field function; performing non-rigid geometric registration on the oral cavity volume feature and the oral cavity model feature based on the updated distance field function to obtain a secondary registration relationship; mapping the oral three-dimensional object into the recombined oral model based on the secondary registration relationship to obtain a fused oral model; performing spatial coincidence verification on the fused oral model, and performing registration optimization on the fused oral model according to a verification result to obtain a registered oral model.

[0049] Wherein, the threshold segmentation-based method extracts the bony region from the oral three-dimensional object, and the local feature extraction on the bony region refers to extracting the bony region of the three-dimensional model of the oral three-dimensional object by using the threshold segmentation method, and extracting the features of the bony region by using the fast point feature histogram or the deep network, the normal estimation and the down-sampling can extract the key geometric feature points of the recombined oral model, such as the crown edge and the palatal cover surface, the feature types of the oral object features and the oral model features are the same, the distance field function is a mathematical function for describing the distance of any point in space to the surface of an object, the value returns positive, negative or zero according to the position relationship of the point and the object surface, the parameter updating refers to updating the rigid or non-rigid transformation parameters of the distance field function according to the primary registration relationship, and the non-rigid geometric registration refers to taking the distance field corresponding to the updated distance field function as a similarity measure, and performing non-rigid registration by using the B-spline or Demons algorithm.

[0050] The anchorage pin matching module is configured to perform anchorage pin point matching on the registered oral model to obtain anchorage pin point information, and initialize an expander model and an anchorage pin abutment model according to the anchorage pin point information.

[0051] Wherein, the anchorage pin point information refers to the position of the matching corresponding anchorage pin of the registered oral model, the anchorage pin is a micro-implant pin used in the treatment of deformity correction, the anchorage pin is implanted into the alveolar bone or the palatal bone to provide stable absolute anchorage and avoid interference with adjacent teeth during tooth movement, the expander model is a model of an expander selected for expander adaptation of the oral cavity, the expander is an orthodontic treatment device used to expand the maxillary dental arch or the palate to correct dental and skeletal problems related to maxillary constriction, and gradually expand the maxillary bone width or the dental arch width by applying a lateral force to improve the tooth arrangement and the occlusal relationship, and the anchorage pin abutment model is a model of an anchorage pin abutment of the expander, the anchorage pin abutment is used to realize the fixed connection of the anchorage pin and the expander, the expander frame is adapted to the abutment through the hole or the clamping structure to fix the device to the palate or the alveolar bone to provide bony anchorage.

[0052] Specifically, when the anchorage pin matching module performs anchorage pin point matching on the registered oral model to obtain anchorage pin point information, the anchorage pin matching module includes: perform voxel quantification on the registered oral model, and generate a bone density distribution map, a bone thickness distribution map and a soft tissue thickness distribution map of the registered oral model according to the voxel quantification result; screen a candidate point area from the registered oral model according to the bone density distribution map, the bone thickness distribution map and the soft tissue thickness distribution map; extract an oral key structure from the registered oral model, calculate a key structure distance according to the oral key structure, and establish a point constraint relationship based on the key structure distance; perform grid point sampling on the candidate point area to obtain a sampling point set, and perform global weighted scoring on each sampling point in the sampling point set based on the point constraint relationship to obtain a point score set; screen a support anchor point group from the sampling point set according to the point score set, and perform heuristic optimization and information labeling on the support anchor point group to obtain support anchor point information.

[0053] The voxel quantification refers to quantitative statistics according to density information in voxel information in the registered oral model, the bone density distribution map is a density distribution map layer of a bone part in the registered oral model, the bone thickness distribution map is a thickness distribution map layer of the bone part in the registered oral model, and the soft tissue thickness distribution map is a thickness distribution map layer of a soft tissue part in the registered oral model. The bone density distribution map, the bone thickness distribution map and the soft tissue thickness distribution map of the registered oral model can be generated by generating a density threshold distribution of voxels after the voxel quantification. The candidate point area is screened from the registered oral model according to the bone density distribution map, the bone thickness distribution map and the soft tissue thickness distribution map, for example, a part voxel area greater than 300HU in the bone density distribution map is taken as the candidate point area.

[0054] In the embodiment of the application, the oral key structure refers to key structures such as a lower pressure groove nerve tube, a sinus floor and a tooth root vertex in the oral cavity, the key structure distance refers to a distance between each point in the candidate point area and the oral key structure, and the point constraint relationship refers to a distance constraint interval of each key structure distance, for example, a point needs to be greater than 1.5mm away from a tooth root, greater than 2mm away from a maxillary sinus floor, implanted in a bone layer depth between 8 and 10mm, a bone cortex thickness is greater than 1mm, and a degree angle needs to be 60 to 80 degrees with a palatal plane.

[0055] In detail, the grid point sampling refers to selecting a corresponding point group from the candidate point area as a sampling point set, the point group is a point group composed of a preset number of candidate points determined according to the model of the expander, the global weighted score refers to taking the importance of each constraint relationship in the point constraint relationship as a weight, and calculating the corresponding point score according to the weighted value of each constraint relationship, and the step of screening the anchorage peg point group from the sampling point set according to the point score set refers to screening the sampling point corresponding to the point score in the point score set whose point score is greater than a preset threshold to form an anchorage peg point group, and the heuristic optimization refers to further optimizing the anchorage peg point in the anchorage peg point group by using a genetic algorithm or a particle swarm algorithm to obtain the best anchorage peg point.

[0056] In the embodiment of the present application, by matching the anchorage peg point of the registration oral model, the voxel quantification and multi-modal fusion are realized to achieve accurate quantitative analysis of the oral bone structure, and the scientificity and safety of the anchorage peg implant point are significantly improved; by combining the key structure constraint and the global weighted score, it is ensured that the selected point meets the mechanical stability and clinical operability at the same time; the heuristic optimization makes the distribution of the anchorage peg more balanced, improves the symmetry and efficiency of the expansion force transmission, and further improves the efficiency of the expander fitting.

[0057] In detail, the anchorage peg matching module includes the following steps when initializing the expander model and the anchorage peg shoulder model according to the anchorage peg point information: The anchorage peg center and the expander frame size are calculated according to the anchorage peg point information; The anchorage peg point information is distributed fitting according to the expander frame size and the anchorage peg center to obtain an expansion contour line; The expansion contour line is updated based on the elastic stress of the expansion contour line to obtain an updated contour line; The bone thickness and the soft tissue thickness corresponding to the anchorage peg point information are calculated according to the registration oral model; The anchorage peg model is matched according to the bone thickness and the soft tissue thickness, and the expander model is initialized according to the anchorage peg model, the expander frame size and the updated contour line; The initial shoulder model is initialized according to the expander model and the anchorage peg model; The palate plane is extracted from the registration oral model, and the fitting degree of the initial shoulder model is updated based on the palate plane to obtain an anchorage peg shoulder model.

[0058] The calculating of the anchorage pin center and the expansion arch frame size according to the anchorage pin point information refers to calculating the center coordinates as the anchorage pin center by using the geometric mean algorithm of the point position, and the expansion arch frame size is determined by calculating the transverse distance between the anchorage pins to determine the frame coverage range, and then the expansion arch frame size is determined; the distribution fitting can be performed by using the spline curve fitting method, and the updating of the expansion arch contour line based on the elastic stress of the expansion arch contour line refers to establishing the elastic stress model of the expansion arch contour line, updating the coordinates of the contour line nodes by using the finite element analysis method, and then re-fitting the updated contour line.

[0059] The bone thickness refers to the bone thickness in the registration oral model corresponding to the anchorage pin point information, and the soft tissue thickness refers to the soft tissue thickness at the position corresponding to the anchorage pin point information; the anchorage pin model refers to the model of the anchorage pin with a length and a size meeting the bone thickness and the soft tissue thickness; the initial expansion arch model refers to selecting an expansion arch three-dimensional model meeting the anchorage pin model, the expansion arch frame size and the updated contour line requirement from the expansion arch models; the initial abutment model refers to the model of the anchorage pin abutment initially meeting the expansion arch model and the anchorage pin model; the palatal plane refers to the plane corresponding to the upper wall of the oral cavity, and the fit degree updating refers to replacing or updating the fitting surface of the initial abutment model according to the surface shape of the palatal plane, so that the initial abutment model is more fitted to the registration oral model.

[0060] The expansion arch adapter module is used for generating an expansion screw model and an anchorage band ring model for the expansion arch model according to the registration oral model, and generating an expansion arch adapter model according to the expansion arch model, the expansion screw model, the anchorage pin abutment model and the anchorage band ring model.

[0061] The expansion screw model is the model corresponding to the expansion screw, the expansion screw is a core component in the expansion arch, which is usually a rotatable metal screw installed on the palatal device to expand the maxillary dental arch or the palatal bone by applying a transverse force by regularly tightening; the anchorage band ring model is the model corresponding to the anchorage band ring, the anchorage band ring is a metal ring fixed on the first molar to connect the expansion arch and the tooth, and the expansion arch, the expansion screw, the anchorage pin abutment and the anchorage band ring jointly constitute an expansion arch system, and the expansion arch adapter model corresponds to the system.

[0062] In the embodiment of the application, when the expansion arch adapter module executes the expansion screw model and the anchorage band ring model for the expansion arch model according to the registration oral model, the expansion arch adapter module comprises: extract a palatal vault surface from the registration dental model, and initialize an expansion screw model on the surface of the expander model based on the palatal vault surface; molar matching is performed on the registration dental model to obtain a dental molar model, and a band ring model is initialized according to the dental molar model; extract an expansion connecting arm from the expander model, and fit a band ring connecting line according to the expansion connecting arm and the band ring model; perform line-fitting matching on the band ring connecting line based on the palatal plane of the registration dental model to obtain a fitted connecting line, and connect the fitted connecting line with the band ring model to obtain a band ring model of the anchorage.

[0063] Wherein, please refer to Figure 4 is a structural schematic diagram of the expander system, wherein c is the expander model, d is the dental molar model, e is the band ring model of the anchorage, f is the expansion connecting arm, g is the expansion screw model, the palatal vault surface is the surface of the middle palatal vault region of the registration dental model, and the initialization of the expansion screw model on the surface of the expander model based on the palatal vault surface means that the model of the initial expansion screw is at the corresponding position on the surface of the expander model, so that the expansion screw model is symmetrically balanced based on the palatal vault surface, the molar matching means identifying and matching the model region corresponding to the molar in the registration dental model, and the initialization of the band ring model according to the dental molar model means initializing the band ring model that fits the dental molar model on the periphery of the dental molar model; the expansion connecting arm is a protruding model region on the expander model for expansion connection, and the fitting of the band ring connecting line according to the expansion connecting arm and the band ring model means initializing the wire model connecting the expansion connecting arm and the band ring model, and the line-fitting matching means matching the shape of the band ring connecting line with the palatal plane and maintaining a safe distance from the palatal plane.

[0064] In detail, by initializing the model based on the anchorage pin point information, the accurate matching of the expander and the jaw bone anatomical structure is ensured, implantation deviation is avoided, treatment safety is enhanced, the expansion screw and the band ring assembly of the anchorage are generated by using the registration dental model, highly personalized adaptation is realized, the risk of soft tissue injury caused by variation of the palatal vault shape is reduced, and the efficiency of expander adaptation is improved.

[0065] Embodiment two: Please refer to Figure 5 The application discloses an expander adaptation simulation method based on 5D bionic technology, which comprises the following steps: generating a three-dimensional oral cavity model according to pre-acquired oral scanning data, and performing multi-level organ separation on the three-dimensional oral cavity model to obtain an oral cavity organ model group; performing targeted organ model recombination on the oral cavity organ model group based on a preset expander adapter organ list to obtain a recombinant oral cavity model; performing multi-modal data registration on the recombinant oral cavity model and pre-acquired oral CBCT to obtain a registered oral cavity model; performing anchorage peg point matching on the registered oral cavity model to obtain anchorage peg point information, and initializing an expander model and an anchorage abutment model according to the anchorage peg point information; generating an expander screw model and an anchorage band ring model for the expander model according to the registered oral cavity model, and generating an expander fitting model according to the expander model, the expander screw model, the anchorage abutment model, and the anchorage band ring model.

[0066] The processes described above with reference to the flowcharts can be implemented as computer software programs in accordance with embodiments of the present disclosure. Embodiments of the present disclosure include a computer program product comprising a computer program carried on a computer readable medium, the computer program comprising program code for performing the methods illustrated by the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication section, and / or installed from a detachable medium. When the computer program is executed by a central processing unit (CPU), the above-described functions defined in the methods of the present application are performed. It should be noted that the computer readable medium of the present application can be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may, for example, but not limited to, be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any suitable combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus or device. In the present application, the computer readable signal medium can include a data signal carried in a baseband or as part of a carrier wave, in which the computer readable program code is carried. Such a propagated data signal can take many forms, including but not limited to, an electromagnetic signal, an optical signal or any suitable combination of the above. The computer readable signal medium can also be any computer readable medium that can send, propagate or transfer a program for use by or in connection with an instruction execution system, apparatus or device. The program code contained on the computer readable medium can be transmitted by any suitable medium, including but not limited to, wireless, wire, optical cable, RF or the like, or any suitable combination of the above.

[0067] The computer program product of the present application can be a computer program product comprising a computer-readable medium bearing computer program code embodied therein for use with a computer. The computer program code can be code defining and / or implementing the present application. The computer program code can be written in any suitable computer readable programming language. The computer program code can be stored in a computer- readable storage medium, such as, but not limited to, any type of disk including an optical disk, a CD-ROM, a CD-R, a CD-RW, a DVD, a flash memory, a ROM, a RAM, a magnetic disk or hard drive, or any other suitable type of medium including a medium that holds the software for a particular or specialized computing purpose, or any suitable combination of media. The computer program product can be a computer program product distributed to end users, whether as a stand-alone program, as part of a physical system, or as a software download. The computer program product can be distributed on a physical medium, such as, but not limited to, a floppy disk, a CD-ROM, a CD-R, a CD-RW, a DVD, a flash memory, a ROM, a RAM, a magnetic disk or hard drive, or any other suitable type of medium, or any suitable combination of media. The computer program product can be distributed from a program distribution center, either as a tangible medium or via electronic delivery, such as from a Web site via the Internet, or from one computer to another via electronic transfer, such as by e-mail. The computer program product can be distributed in an encrypted manner, such as via encryption or via password protection.

[0068] Those skilled in the art will understand that the application described above and illustrated in the accompanying drawings is presented by way of example only and is not limiting as to the present application. The intent is to cover all modifications and alternatives of the present application falling within the scope of the application.

Claims

1. A bow expander adaptation simulation system based on 5D bionic technology, characterized in that, The system includes a model separation module, a model recombination module, a multimodal registration module, a bracing nail matching module, and a bow expander adaptation module, wherein: The model separation module is used to generate a three-dimensional oral cavity model based on pre-collected oral scan data, and to perform multi-level organ separation on the three-dimensional oral cavity model to obtain an oral cavity organ model group. The model recombination module is used to perform targeted organ model recombination on the oral organ model group based on a preset list of expander adapter organs to obtain a reconstructed oral model. A multimodal registration module is used to register the reconstructed oral cavity model with pre-acquired oral CBCT data in a multimodal manner to obtain a registered oral cavity model. The anchorage screw matching module is used to match the anchorage screw positions on the registered oral cavity model, obtain the anchorage screw position information, and initialize the expander model and the anchorage screw shoulder model according to the anchorage screw position information. The expander adapter module is used to generate an expander screw model and an anchorage dental band model for the expander model based on the registered oral model, and to generate an expander adapter model based on the expander model, the expander screw model, the anchorage screw shoulder model, and the anchorage dental band model.

2. The bow expander adaptation simulation system based on 5D bionic technology according to claim 1, characterized in that, When the model separation module performs multi-level organ separation on the oral cavity 3D model to obtain an oral cavity organ model group, it includes: The oral cavity 3D model is projected from multiple perspectives to obtain an oral cavity projection atlas; The oral cavity projection map is subjected to block feature mapping and positional convolution to obtain a two-dimensional oral cavity feature set. Based on the two-dimensional oral cavity feature set, the oral cavity projection map is subjected to two-dimensional organ segmentation to obtain two-dimensional segmentation labels. The oral cavity 3D model is projected using the two-dimensional segmentation labels to obtain primary segmentation labels; Based on the primary segmentation labels, voxel-weighted convolution is performed on the oral cavity 3D model to obtain oral cavity voxel features; The oral voxel features are transposed and convolved with skip connections to obtain secondary segmentation labels; Based on the secondary segmentation labels, local attention features are extracted and global features are aggregated on the oral cavity 3D model to obtain aggregated model point features. Based on the aggregated model point features, multi-level organ separation is performed on the oral cavity 3D model to obtain oral cavity organ model group.

3. The bow expander adaptation simulation system based on 5D bionic technology according to claim 2, characterized in that, When the model separation module performs local attention feature extraction and global feature aggregation on the oral cavity 3D model based on the secondary segmentation labels to obtain aggregated model point features, it includes: Point cloud sampling is performed on the oral cavity 3D model, and rigid registration of the point cloud sampling oral cavity 3D model is performed to obtain rigid oral cavity point cloud; Based on the secondary segmentation labels, point features are extracted from the rigid oral cavity point cloud to obtain oral cavity point cloud features; Local attention features are extracted from the oral cavity point cloud features based on the relative coordinates of the neighborhood of the oral cavity point cloud features to obtain point cloud attention features. The point cloud attention features are fused using a skip connection approach, and then the fused point cloud attention features are aggregated with global attention weighting to obtain aggregated model point features.

4. The bow expander adaptation simulation system based on 5D bionic technology according to claim 1, characterized in that, The multimodal registration module, when performing multimodal data registration between the reconstructed oral model and pre-acquired oral CBCT to obtain a registered oral model, includes: The pre-acquired oral CBCT was converted into a three-dimensional body to obtain an oral three-dimensional body; The bony regions are extracted from the three-dimensional oral cavity body using a threshold segmentation method, and local feature extraction is performed on the bony regions to obtain oral cavity body features. Normal estimation and downsampling are performed on the reconstructed oral cavity model to obtain oral cavity model features; Geometric feature matching and geometric constraint registration are performed on the oral cavity body features and the oral cavity model features to obtain a primary registration relationship; A distance field function is established based on the oral cavity three-dimensional body, and the parameters of the distance field function are updated according to the primary registration relationship to obtain the updated distance field function; Based on the updated distance field function, non-rigid geometric registration is performed on the oral cavity body features and the oral cavity model features to obtain a secondary registration relationship; Based on the secondary registration relationship, the three-dimensional oral cavity body is mapped to the reconstructed oral cavity model to obtain a fused oral cavity model; The spatial overlap of the fused oral cavity model was verified, and the registration optimization was performed on the fused oral cavity model based on the verification results to obtain a registered oral cavity model.

5. The bow expander adaptation simulation system based on 5D bionic technology according to claim 1, characterized in that, When the anchorage screw matching module performs anchorage screw location matching on the registered oral model to obtain anchorage screw location information, it includes: The registered oral cavity model is voxel-quantized, and bone density distribution map, bone thickness distribution map and soft tissue thickness distribution map of the registered oral cavity model are generated based on the voxel-quantization results. Candidate point regions are selected from the registered oral model based on the bone density distribution map, the bone thickness distribution map, and the soft tissue thickness distribution map. Key oral structures are extracted from the registered oral model, key structure distances are calculated based on the key oral structures, and point constraint relationships are established based on the key structure distances. The candidate point area is sampled by grid points to obtain a set of sampling points, and each sampling point in the set of sampling points is globally weighted and scored based on the point constraint relationship to obtain a point score set; Based on the point score set, the anchorage nail point group is selected from the sampling point set, and the anchorage nail point group is heuristically optimized and information labeled to obtain the anchorage nail point information.

6. The bow expander adaptation simulation system based on 5D bionic technology according to claim 1, characterized in that, The model separation module, when generating a three-dimensional oral cavity model based on pre-collected intraoral scan data, includes: Point cloud noise filtering is performed on the pre-collected orifice scan data to obtain denoised orifice scan data; The denoised scanning data is subjected to a neighborhood search to obtain neighborhood point cloud data. Based on the neighborhood point cloud data, the denoised scanning data is subjected to plane fitting and feature decomposition to obtain primary normal vector features. The primary normal vector features are uniformized to obtain the standard normal vector features; Local topology analysis is performed on the denoised port scan data to obtain boundary point clouds, and a boundary loop is constructed based on the boundary point clouds; The boundary loop is fitted with a surface based on the standard normal vector features to obtain a fitted surface, and the fitted surface is used to fill the point cloud of the denoised orifice scan data to obtain the completed orifice scan data. The completed oral scan data is triangulated to obtain a three-dimensional oral cavity model.

7. The bow expander adaptation simulation system based on 5D bionic technology according to claim 1, characterized in that, When the model recombination module performs targeted organ model recombination on the oral organ model group based on a preset list of expander adapter organs to obtain a reconstructed oral model, it includes: The Gaussian curvature features of the oral organ model group are extracted, and redundancy detection and volume removal are performed on the oral organ model group based on the Gaussian curvature features to obtain the noise-reduced organ model group. Based on the list of bow expander adapters, organ screening is performed on the noise-reducing organ model group to obtain the adapter organ model group. Based on the constraints of organ anatomical relationships, the adapter organ model group is coordinate aligned to obtain the aligned organ model; The aligned organ model is smoothed on both sides, and the organ interface is matched from the smoothed aligned organ model. The organ interface is then resampled locally to obtain a smoothed organ model. The curvature of the smooth organ model is optimized, and the triangular topology of the smooth organ model after curvature optimization is optimized to obtain the optimized organ model. The optimized organ model is subjected to triangular self-intersection detection, and the mesh surface of the optimized organ model is repaired based on the results of the triangular self-intersection detection to obtain a reconstructed oral cavity model.

8. The bow expander adaptation simulation system based on 5D bionic technology according to claim 1, characterized in that, When the anchorage nail matching module initializes the expander model and the anchorage nail shoulder model based on the anchorage nail location information, it includes: The center of the anchor nail and the dimensions of the expander frame are calculated based on the anchor nail location information. Based on the dimensions of the expander frame and the center of the anchor pin, the distribution of the anchor pin location information is fitted to obtain the expander outline; The bow expansion profile is updated based on the elastic stress of the bow expansion profile to obtain an updated profile. The bone thickness and soft tissue thickness corresponding to the anchorage screw location information are calculated based on the registered oral model. The anchorage screw model is matched according to the bone thickness and the soft tissue thickness, and the expander model is initialized according to the anchorage screw model, the expander frame size, and the updated contour line. Initialize the initial shoulder platform model based on the bow expander model and the anchor nail model; The palatal plane is extracted from the registered oral cavity model, and the fit of the initial shoulder model is updated based on the palatal plane to obtain the anchorage screw shoulder model.

9. The bow expander adaptation simulation system based on 5D bionic technology according to claim 1, characterized in that, When the expander adapter module generates expander screw models and anchorage band models for the expander model based on the registered oral model, it includes: The palatal dome surface is extracted from the registered oral cavity model, and the expander screw model is initialized on the surface of the expander model based on the palatal dome surface; The registered oral model is matched with molars to obtain an oral molar model, and the dental band model is initialized based on the oral molar model; Extract the expander connecting arm from the expander model, and fit the dental band loop connecting line based on the expander connecting arm and the dental band loop model; Based on the palatal plane of the registered oral model, the dental band connection line is matched and fitted to obtain the fitting connection line. The fitting connection line is then connected to the dental band model to obtain the anchorage dental band model.

10. A method for simulating bow expander adaptation based on 5D bionic technology, characterized in that, The method includes: A three-dimensional oral cavity model is generated based on pre-collected oral scan data, and multi-level organ separation is performed on the three-dimensional oral cavity model to obtain an oral cavity organ model group. Based on a preset list of expander adapters, the oral organ model group is reconstructed using targeted organ model recombination to obtain a reconstructed oral model. The reconstructed oral cavity model is registered with pre-acquired oral CBCT data using multimodal data to obtain a registered oral cavity model; The registered oral cavity model is matched with anchorage screw locations to obtain anchorage screw location information, and the expander model and anchorage screw shoulder model are initialized according to the anchorage screw location information. Based on the registered oral model, an expander screw model and an anchorage band model are generated for the expander model, and an expander fit model is generated based on the expander model, the expander screw model, the anchorage screw shoulder model, and the anchorage band model.

Images