Intelligent generation method, media and equipment for personalized dental implant abutments
By generating the abutment gingival margin line and transgingival contour surface, calculating the height difference of the internal space of the crown, and setting a safety design coefficient, the problems of spatial interference and insufficient retention force in the design of abutment preparations are solved, and the accurate generation and efficient processing of abutment preparations are realized.
Patent Information
- Application Number
- CN202611357632.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-03
- Publication Date
- 2026-11-17
AI Technical Summary
In the current process of designing personalized dental implant abutments, there is a lack of precise detection and boundary calculation mechanisms based on the actual anatomical space inside the crown. This leads to blind setting of the height of the abutment preparation body, which can easily cause spatial interference with the restorative crown or insufficient retention.
By acquiring the registered intraoral scanning mesh, base mesh, and initial crown mesh, the abutment gingival margin line and transgingival contour surface are generated. The height difference between the intersection of the probe line and the crown mesh is calculated, and the minimum value is taken as the maximum usable space height. This value is then multiplied by a safety design factor to generate the abutment preparation body, ensuring that the height setting of the abutment preparation body is accurate and avoiding spatial interference.
This method enables the objective and quantitative setting of the height of the abutment preparation, avoids crown perforation and spatial interference, ensures sufficient axial retention force, and improves the standardization of design and processing efficiency.
Smart Images

Figure CN123116376A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of digital dentistry, specifically to a method, medium, and device for intelligent generation of personalized dental implant abutments. Background Technology
[0002] Currently, in the fields of digital dental restoration and computer-aided design (CAD) of implant abutments, personalized computer-aided design of implant abutments has been gradually applied. However, existing digital design solutions still have significant limitations. Traditional abutment CAD design processes heavily rely on the technician's manual experience, typically requiring manual point marking on an intraoral scanning grid to determine edges, and repeated manual adjustment of parameters and topological suturing. This highly subjective approach not only involves complex procedures and time-consuming design of individual abutments, but also, due to differences in technician experience and skill levels, easily leads to poor consistency and standardization of design results across different batches, severely restricting the design capacity and delivery efficiency of modern dental prosthesis manufacturing plants and clinical treatment.
[0003] In terms of core spatial geometric feature extraction, existing software solutions lack sufficient anatomically adaptive adjustment mechanisms. When establishing the abutment placement path direction, conventional techniques generally employ a single general bounding box algorithm or rely entirely on subjective estimation based on visual experience. This fails to effectively account for the significant spatial differences in three-dimensional anatomical morphology among different tooth positions such as molars, anterior teeth, or canines, easily leading to placement difficulties during clinical placement of the restoration. Furthermore, existing techniques often rely on blind estimation without precise spatial exploration when setting key parameters such as the extraction height of the abutment preparation. They cannot accurately calculate collision intersections and perform boundary calibration for the actual anatomical space within the initial crown mesh, resulting in abutment preparation structures that are frequently too tall, causing them to perforate the restoration crown, or too close to the inner wall of the crown, causing severe spatial interference. However, blindly reducing the abutment height to avoid interference can lead to a serious lack of axial retention force in the abutment preparation. Summary of the Invention
[0004] In view of the above problems, the present invention provides a technical solution for intelligent generation of personalized dental implant abutments, which solves the technical problems in the design of existing abutment preparations, which are caused by the lack of a precise detection and limit calculation mechanism based on the actual anatomical space inside the crown, resulting in blind setting of the height of the abutment preparation, easy spatial interference with the restorative crown, or insufficient retention.
[0005] To achieve the above objectives, in a first aspect, the present invention provides a method for intelligent generation of personalized dental implant abutments, the method comprising:
[0006] S1: Obtain the registered intraoral scanning mesh, base mesh, and initial crown mesh;
[0007] S2: Generate the abutment gingival margin line based on the intraoral scanning grid;
[0008] S3: Based on the bottom reference ring of the base grid and the gingival margin line of the abutment, generate the transgingival contour surface;
[0009] S4: Based on the gingival contour surface and the initial crown mesh, generate the abutment preparation body;
[0010] Specifically, step S4 includes:
[0011] S41: Determine the placement path direction based on the tooth position type of the initial crown mesh, wherein the placement path direction is the axial direction when the abutment is placed into the patient's oral cavity;
[0012] S42: Starting from the point on the top boundary ring of the transgingival contour surface, generate multiple parallel probe lines along the path of insertion toward the initial crown mesh;
[0013] S43: Calculate the intersection points of each of the probe lines and the initial crown grid, and calculate the height difference between each intersection point and the center point of the top boundary ring of the transgingival contour surface in the direction of the path of insertion. Take the minimum value among all height differences as the maximum usable space height.
[0014] S44: Multiply the maximum available space height by a preset safety design factor to obtain the safety design height of the base preparation body;
[0015] S45: Generate the base preparation body according to the safety design height.
[0016] Furthermore, step S2 specifically includes:
[0017] S21: Extract multiple initial points at the edge of the gingival cuff of the intraoral scanning grid;
[0018] S22: Establish a first cylindrical coordinate system based on the plurality of initial points;
[0019] S23: Transform the plurality of initial points to the first cylindrical coordinate system, and perform interpolation based on polar angle sorting to generate the desired gingival margin trajectory that changes continuously with the polar angle;
[0020] S24: Transform all grid vertices contained in the intraoral scanning grid to the first cylindrical coordinate system, and select grid vertices that meet the preset error conditions from all grid vertices contained in the intraoral scanning grid in multiple angular neighborhoods as first-type gingival margin candidate points; if there are no grid vertices that meet the preset error conditions in the corresponding angular neighborhood, then use the point at the corresponding polar angle on the expected gingival margin trajectory as the second-type gingival margin candidate point; connect each first-type gingival margin candidate point and each second-type gingival margin candidate point in polar angle order and perform smoothing filtering to generate the abutment gingival margin line.
[0021] Furthermore, step S3 specifically includes:
[0022] S31: Extract the open edges of the base mesh and filter out the bottom reference ring;
[0023] S32: Establish a second cylindrical coordinate system based on the bottom reference ring, and transform the bottom reference ring and the gingival margin line of the abutment to the second cylindrical coordinate system;
[0024] S33: Between the bottom reference ring and the abutment gingival margin line, a plurality of transverse cross-sectional layers are divided along the height direction, and at each polar angle position, the intermediate layer vertex on each transverse cross-sectional layer is generated according to the radial coordinate values of each point on the bottom reference ring, the axial coordinate values of each point on the bottom reference ring, the radial coordinate values of each point on the abutment gingival margin line, and the axial coordinate values of each point on the abutment gingival margin line.
[0025] Among them, a neck thickness retention section is provided on the side near the gingival margin line of the abutment, and the apex of the middle layer on each transverse cross-section layer in the neck thickness retention section maintains a constant radial distance from the gingival margin line of the abutment in the radial direction.
[0026] In the region below the neck thickness holding section, the radial concavity of the intermediate layer vertex is determined based on the relative positional relationship between the height of the intermediate layer vertex on each transverse cross section and the height of the bottom reference ring. The radial concavity varies with the polar angle and with the height. The radial concavity is zero near the lower boundary of the neck thickness holding section and gradually increases with decreasing height until it reaches its maximum value at the bottom reference ring.
[0027] S34: Convert the intermediate layer vertices on each transverse section layer back to three-dimensional space coordinates, construct topological relationships to generate mesh patches, and generate the gingival contour surface.
[0028] Furthermore, after generating the transgingival contour surface in step S3, the method further includes fine-tuning the transgingival contour surface, specifically including:
[0029] Receive user input of adjustment height ratio position parameters and fine-tuning radius deformation parameters;
[0030] Extract the boundary of the transgingival contour surface and separate it into a bottom boundary feature point set and a top boundary feature point set;
[0031] Keeping the coordinate values of the bottom boundary feature point set and the top boundary feature point set unchanged, based on the adjusted height ratio position parameter, the target height value of the internal mesh vertex in the axial direction of the middle section of the transgingival contour surface is determined. The target height value is calculated based on the relative positional relationship between the height of the internal mesh vertex and the height of the bottom boundary feature point set and the ratio coefficient indicated by the adjusted height ratio position parameter.
[0032] Based on the fine-tuning radius deformation parameter, the target radial coordinate value of the inner mesh vertex in the radial direction is determined. The target radial coordinate value is calculated by weighting the original radial coordinate value of the inner mesh vertex, the deformation indicated by the fine-tuning radius deformation parameter, and the relative distance between the inner mesh vertex and the bottom boundary feature point set and the top boundary feature point set.
[0033] The internal mesh vertices are moved to the positions corresponding to the target height value and the target radial coordinate value to generate the fine-tuned transgingival contour surface.
[0034] Step S4 includes: generating an abutment preparation body based on the finely adjusted transgingival contour surface and the initial crown mesh.
[0035] Furthermore, step S41 specifically includes:
[0036] If the current tooth position is a molar, the three-dimensional convex hull of the initial crown mesh is calculated. Among the upward-facing convex hull patches, candidate convex hull patches whose angle between the normal vector of the convex hull patch and the preset reference direction is less than a preset angle threshold are retrieved. The target convex hull patch with the largest surface area is selected from the candidate convex hull patches, and the unit positive normal vector of the target convex hull patch is determined as the path of insertion.
[0037] If the current tooth position is a canine or anterior tooth, calculate the average coordinates of all grid vertices of the initial crown mesh. Establish a three-dimensional spatial coordinate system with the point corresponding to the average coordinates as the origin. Calculate the global centroid of the initial crown mesh, which is the weighted average of all grid vertices, with the weight being the area of each grid facet. Determine the vector from the origin to the global centroid as the initial axis. Extract the vertex clusters whose projection height on the initial crown mesh along the initial axis is within a preset height ratio range. Calculate the local centroid of the vertex cluster and determine the vector from the origin to the local centroid as the final placement path direction.
[0038] Furthermore, step S45 includes the generation step of the base side, specifically including:
[0039] Harmonic fitting is performed on the axial height of the intersection of the probe line and the initial crown grid to extract the overall height trend of the circumferential variation of the top boundary ring along the transgingival contour surface. The local drop positions on the fitted curve are adaptively corrected, and the fitted fluctuations are scaled to a preset range and then superimposed on the safe design height to obtain the sidewall top boundary ring. The local drop position refers to the local point that has an abnormal downward shift relative to the adjacent circumferential position or the overall low-frequency trend.
[0040] Starting from the top boundary ring of the transgingival contour surface, multiple sidewall levels are divided along the placement path to the top boundary ring of the sidewall to generate the abutment side.
[0041] Furthermore, step S45 includes the generation step of the top surface of the base, specifically including:
[0042] Using the top boundary ring of the sidewall as the outer boundary of the top surface, multiple ring-shaped sampling points are established from the outside to the inside within the area enclosed by the outer boundary of the top surface. Each sampling point is projected onto the occlusal surface of the initial crown grid along the path of placement to obtain the crown projection height value of each sampling point. For local sampling points that fail to be projected, periodic interpolation compensation is performed based on the crown projection height values of adjacent valid sampling points on the same ring.
[0043] Based on the radial position of each sampling point, a base height of the abutment that smoothly changes from the top boundary of the sidewall to the center is established. Based on the projection height of the outer region of the crown and the projection height of the multiple crown rings near the center, a reference height of the crown that smoothly changes from the periphery to the center is established.
[0044] Calculate the height difference between the crown projection height value of each sampling point and the corresponding crown reference height to obtain the local undulation characteristics of the crown occlusal surface relative to the overall height trend; retain positive height differences with a larger undulation retention ratio and limit the maximum allowable upward convexity, and retain negative height differences with a smaller undulation retention ratio and limit the maximum allowable downward concavity.
[0045] Based on the radial position of each sampling point from the outer boundary, the smoothing introduction weights of positive undulation features and negative concavity features are determined respectively, so that the crown morphological features are gradually introduced from the outer boundary to the interior of the top surface; the processed positive undulations and negative undulations are superimposed on the abutment base height according to the corresponding smoothing introduction weights to obtain the final target top surface height of each sampling point.
[0046] Using the final target top surface height of each sampling point as the fitting data, a robust low-frequency surface fitting algorithm is used to generate a continuous and smooth abutment top surface, so as to preserve the overall gentle undulation trend of the crown occlusal surface and filter out small grooves, local peaks and scanning noise.
[0047] A smooth transition region is set between the top boundary ring of the sidewall and the fitted top surface of the base. The outer periphery of the smooth transition region maintains the original height of the top boundary ring of the sidewall, and the height and radial slope of the top surface of the base are gradually matched from the outer periphery to the inner periphery, thereby obtaining a smooth top surface of the base that is continuously connected to the side of the base.
[0048] Furthermore, after generating the prepared abutment, the method further includes:
[0049] Extract the screw channel boundary ring located inside the bottom reference ring of the base mesh, and determine the central axis and channel radius of the screw channel boundary ring;
[0050] Construct a screw channel cylinder based on the central axis of the screw channel boundary ring and the channel radius;
[0051] Based on the screw channel cylinder, a Boolean subtraction operation is performed on the base preparation body to delete the target mesh facet located inside the screw channel cylinder in the base preparation body, and the cutting boundary is reconstructed at the cutting intersection position to obtain the intermediate base model with the top opening ring.
[0052] Align the screw channel boundary ring with the top opening ring on the axis so that the central axis of the screw channel boundary ring and the top opening ring coincide and the normal direction of the ring surface are consistent.
[0053] Between the screw channel boundary ring and the top opening ring, a connecting mesh is generated using a mesh reconstruction algorithm to serve as the inner wall of the screw channel, thus obtaining the final base preparation body.
[0054] In a second aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the intelligent generation method for personalized dental implant abutments as described in the first aspect of the present invention.
[0055] In a third aspect, the present invention provides an electronic device having a computer program stored thereon, including a processor and a storage medium, wherein the computer program is stored on the storage medium, and when executed by the processor, the computer program implements the intelligent generation method for personalized dental implant abutments as described in the first aspect of the present invention.
[0056] Unlike existing technologies, the above-mentioned technical solution involves a personalized intelligent generation method, medium, and device for dental implant abutments, belonging to the field of digital dental restoration and computer-aided design of implants. The method includes: acquiring the registered intraoral scanning mesh, base mesh, and initial crown mesh; generating the abutment gingival margin line based on the intraoral scanning mesh; generating the transgingival contour surface based on the bottom reference ring of the base mesh and the abutment gingival margin line; and generating the abutment preparation body based on the transgingival contour surface and the initial crown mesh. Specifically, when generating the abutment preparation body, the placement path direction is determined according to the tooth position type of the initial crown mesh. Starting from a point on the top boundary ring of the transgingival contour surface, multiple parallel probe lines are generated along the placement path direction towards the initial crown mesh. The intersection points of each probe line with the initial crown mesh are calculated, and the height difference between each intersection point and the center point of the top boundary ring in the placement path direction is calculated. The minimum value is taken as the maximum usable space height, and then multiplied by a preset safety design coefficient to obtain the safe design height. The abutment preparation body is generated accordingly. This invention achieves objective quantitative setting of the abutment preparation height by accurately probing the actual anatomical space inside the crown, effectively avoiding problems of crown perforation and spatial interference, while ensuring sufficient axial retention force.
[0057] The above description of the invention is merely an overview of the technical solution of the present invention. In order to enable those skilled in the art to better understand the technical solution of the present invention and to implement it based on the description and drawings, and to make the above-mentioned objectives and other objectives, features and advantages of the present invention easier to understand, the following description is provided in conjunction with the specific embodiments and drawings of the present invention. Attached Figure Description
[0058] The accompanying drawings are only used to illustrate the principles, implementation methods, applications, features, and effects of specific embodiments of the present invention and other related contents, and should not be considered as limitations on the present invention.
[0059] In the accompanying drawings of the instruction manual:
[0060] Figure 1 This is a flowchart of a personalized denture implant abutment intelligent generation method according to an exemplary embodiment of the present invention;
[0061] Figure 2 This is a flowchart illustrating step S4 of an exemplary embodiment of the present invention;
[0062] Figure 3 This is a schematic diagram of the generated abutment gingival margin line according to an exemplary embodiment of the present invention;
[0063] Figure 4 This is a schematic diagram of the generated transgingival contour surface according to an exemplary embodiment of the present invention;
[0064] Figure 5This is a comparison image of the generated transgingival contour surface before and after fine-tuning, which is an exemplary embodiment of the present invention.
[0065] Figure 6 This is a schematic diagram of the generated base preparation body according to an exemplary embodiment of the present invention;
[0066] Figure 7 This is a schematic diagram of an electronic device according to an exemplary embodiment of the present invention;
[0067] The reference numerals used in the above figures are explained as follows:
[0068] 10. Electronic equipment; 101. Processor; 102. Storage medium. Detailed Implementation
[0069] To explain in detail the possible application scenarios, technical principles, specific feasible solutions, and the objectives and effects that can be achieved by this invention, the following detailed description is provided in conjunction with the listed specific embodiments and accompanying drawings. The embodiments described herein are only used to more clearly illustrate the technical solutions of this invention, and are therefore only examples, and should not be used to limit the scope of protection of this invention.
[0070] In the first aspect, such as Figure 1 As shown, this application provides a method for intelligent generation of personalized dental implant abutments, the method comprising:
[0071] S1: Obtain the registered intraoral scanning mesh, base mesh, and initial crown mesh;
[0072] S2: Generate the abutment gingival margin line based on the intraoral scanning grid;
[0073] S3: Based on the bottom reference ring of the base grid and the gingival margin line of the abutment, generate the transgingival contour surface;
[0074] S4: Based on the gingival contour surface and the initial crown mesh, generate the abutment preparation body;
[0075] Among them, such as Figure 2 As shown, step S4 specifically includes:
[0076] S41: Determine the placement path direction based on the tooth position type of the initial crown mesh, wherein the placement path direction is the axial direction when the abutment is placed into the patient's oral cavity;
[0077] S42: Starting from the point on the top boundary ring of the transgingival contour surface, generate multiple parallel probe lines along the path of insertion toward the initial crown mesh;
[0078] S43: Calculate the intersection points of each of the probe lines and the initial crown grid, and calculate the height difference between each intersection point and the center point of the top boundary ring of the transgingival contour surface in the direction of the path of insertion. Take the minimum value among all height differences as the maximum usable space height.
[0079] S44: Multiply the maximum available space height by a preset safety design factor to obtain the safety design height of the base preparation body;
[0080] S45: Generate the base preparation body according to the safety design height.
[0081] In this embodiment, the intraoral scanning mesh (IOS Mesh) refers to a three-dimensional triangular mesh model obtained by directly acquiring the patient's oral gingiva, natural teeth, and soft tissue in the edentulous area through an intraoral optical scanning device. It fully carries the three-dimensional geometric information of the gingival cuff and the anatomical shape of the dental arch, and is the original data carrier for extracting the gingival margin contour and soft tissue boundary. All vertices of the scanning mesh store three-dimensional spatial coordinates, and there are scanning sampling distortions and vertex noise, which require subsequent filtering and smoothing processing.
[0082] The preferred base mesh is a titanium-based mesh (Ti-Base Mesh), which is a standardized industrial implant titanium base three-dimensional mesh model. It is spatially registered and aligned with the intraoral scanning mesh in advance, and has its own implant connection structure and screw channel annular boundary, which are used to define the reference boundary of the bottom of the abutment transgingival segment and the internal screw channel space.
[0083] The initial crown mesh refers to the three-dimensional mesh model of the pre-designed crown. After spatial registration, it serves as the spatial constraint boundary of the preparation body above the abutment, limiting the maximum space that the preparation body can accommodate and preventing spatial interference or perforation between the structure and the inner wall of the crown.
[0084] The placement path direction refers to the only axis along which the abutment is completely inserted into the implant in a straight line without jamming or interference. It is also the reference coordinate axis Z-axis for abutment preparation, height detection, and sidewall shaping. Different tooth positions and anatomical morphologies correspond to different axial calculation logics.
[0085] The probe line refers to a spatial ray that extends parallel to the path of insertion, with all the vertices of the top boundary ring of the transgingival contour surface as the origin. It is used to quantitatively detect the three-dimensional containment height inside the initial crown, and to achieve precise digital measurement of the internal space of the crown.
[0086] The maximum usable space height refers to the minimum value of the height difference along the path of décor at the intersection points of all probe rays and the initial crown mesh. It represents the limit height that the crown interior can accommodate along the path of décor, providing an objective geometric boundary for abutment height design.
[0087] The safety design factor refers to a preset constant safety margin parameter that is less than 1. By scaling the limit height, a uniform safety gap is reserved on the basis of the limit space to avoid the abutment preparation body from contacting and colliding with the inner wall of the crown.
[0088] The abutment preparation body refers to the main structure of the abutment located above the gum line and enclosed inside the restorative crown. It includes the sidewalls, top surface, rounded transition area, and through screw channel, and bears the functions of occlusal force, axial retention, and screw locking.
[0089] The working principle of this embodiment is as follows:
[0090] In step S1, the system reads three sets of three-dimensional mesh data that have been pre-registered globally. The registration operation ensures that the intraoral soft tissue, titanium base, and restoration crown are in the same unified three-dimensional coordinate system, eliminating the calculation error caused by spatial misalignment of the multi-source scanning model. The three sets of meshes provide three core geometric bases: anatomical morphology of gingival soft tissue, geometric constraints of implant base interface, and space limitations of upper restoration, which are the data foundation for all subsequent automated calculations.
[0091] In step S2, based on the geometric features of the gingival cuff from the intraoral scanning mesh, a complete set of geometric algorithms—coordinate system transformation, interpolation fitting, multi-candidate point screening, and smoothing filtering—automatically outputs a continuous, smooth, and distortion-free closed abutment gingival margin line. This abutment gingival margin line serves as the apical boundary of the abutment's transgingival segment, precisely matching the patient's actual gingival cuff anatomical contour and defining the junction between the abutment and the gingival soft tissue. The generated abutment gingival margin line is shown below. Figure 3 As shown.
[0092] In step S3, using the bottom reference ring obtained from the titanium base mesh screening as the bottom constraint and the gingival margin line generated in step S2 as the top constraint, a vase-shaped transgingival contour surface with gradient radial concavity is generated through layered interpolation between the two sets of ring boundaries. The upper and lower edges of the surface strictly conform to the titanium base and the gingival margin line, respectively, adapting to the natural morphology of the gingival soft tissue and reducing the risk of soft tissue compression after implantation. The generated transgingival contour surface is as follows: Figure 4 As shown.
[0093] In step S4, based on the transgingival contour surface and the initial crown mesh, the automated modeling of the abutment preparation inside the crown is completed. This step sets up four layers of progressive sub-logic to solve the core pain points of traditional techniques, such as highly blind estimation and spatial interference, as follows:
[0094] In step S41, the direction of the insertion path is calculated based on the anatomical differences of different tooth positions. Two independent geometric calculation logics are used to distinguish between two morphological characteristics: the wide and flat occlusal surface of molars and the narrow long axis of anterior teeth or canines. This replaces the traditional single general algorithm, ensuring that the calculated axial direction conforms to the natural growth trend of the crown and avoiding clinical insertion jamming and axial deviation defects from the source.
[0095] In step S42, each grid vertex on the top boundary ring of the transgingival contour surface is used as the starting point for ray emission. Equidistant parallel probe lines are generated in batches along the determined path of placement. The rays penetrate the internal space of the crown completely with perpendicularity to the abutment draft direction, achieving full circumferential space coverage detection inside the crown without any blind spots.
[0096] In step S43, the spatial intersection point of each probe line with the initial crown triangular mesh is solved, and the axial height difference between the intersection point and the center point of the top boundary ring is calculated. After traversing all probe line height values, the minimum value is selected as the maximum usable space height. This minimum value represents the narrowest accommodation position of the crown inner wall along the path of placement. As long as the height of the abutment preparation does not exceed this value, it can be guaranteed that the abutment will not touch or interfere with the crown inner wall at any position in the circumference.
[0097] In step S44, a safety design factor is introduced to scale the limit capacity height. The maximum available space height is multiplied by this factor to obtain the safety design height. By reserving a uniform clearance at a fixed ratio, both spatial safety and axial retention height of the base are taken into account. This avoids excessive reduction of the preparatory body height to avoid interference, which would result in insufficient retention force.
[0098] In step S45, taking the safe design height as the upper limit of the axial direction, and combining the circumferential height undulation of the inner wall of the crown and the concave and convex features of the occlusal surface, a continuous, smooth, and spatially interference-free complete abutment preparation three-dimensional mesh model is adaptively generated.
[0099] The above-described approach automates the entire implant abutment process, from soft tissue boundary extraction and transgingival surface shaping to in-coronal preparation modeling. This eliminates the reliance on traditional manual point marking and repeated parameter adjustments, significantly improving modeling standardization and processing capacity. Adaptive solution for the placement path based on tooth position accommodates various tooth anatomy, resolving difficulties in prosthesis placement. A probe ray is emitted along the entire placement path to quantify the internal space of the crown, determining the design height by combining the minimum height with a safety factor. This algorithm fundamentally eliminates abutment perforation and internal wall interference defects while maximizing the preservation of the axial height of the preparation, balancing spatial safety and retention strength, and significantly reducing the cost of manual trial and error for technicians.
[0100] In some embodiments, step S2 specifically includes:
[0101] S21: Extract multiple initial points at the edge of the gingival cuff of the intraoral scanning grid;
[0102] S22: Establish a first cylindrical coordinate system based on the plurality of initial points;
[0103] S23: Transform the plurality of initial points to the first cylindrical coordinate system, and perform interpolation based on polar angle sorting to generate the desired gingival margin trajectory that changes continuously with the polar angle;
[0104] S24: Transform all grid vertices contained in the intraoral scanning grid to the first cylindrical coordinate system, and select grid vertices that meet the preset error conditions from all grid vertices contained in the intraoral scanning grid in multiple angular neighborhoods as first-type gingival margin candidate points; if there are no grid vertices that meet the preset error conditions in the corresponding angular neighborhood, then use the point at the corresponding polar angle on the expected gingival margin trajectory as the second-type gingival margin candidate point; connect each first-type gingival margin candidate point and each second-type gingival margin candidate point in polar angle order and perform smoothing filtering to generate the abutment gingival margin line.
[0105] In this embodiment, the gingival cuff edge refers to the annular concave contour area at the junction of the gingival soft tissue and the alveolar socket at the implantation site in the intraoral scanning grid, which is the boundary reference for the contact between the abutment transgingival structure and the gingiva.
[0106] The first cylindrical coordinate system refers to a local polar coordinate system constructed based on the initial feature points extracted from the gingival cuff. The origin is the geometric mean of the initial point. The axial Z-axis is determined by PCA decomposition, and the X and Y axes are generated by planar orthogonal decomposition. The three-dimensional boundary points are reduced to a three-dimensional parameter space of polar angle-polar radius-height, simplifying the interpolation and filtering operations of the annular contour.
[0107] The expected gingival margin trajectory refers to the continuous closed-loop reference curve generated by converting the initial feature points to a cylindrical coordinate system, sorting them by polar angle, and then interpolating them using cubic splines. It represents the theoretically ideal smooth gingival margin contour and is used to compensate for point gaps in the scanning grid and noisy areas.
[0108] Angle neighborhood refers to dividing the 360° circumference of the cylindrical coordinate system into fan-shaped partitions according to a preset angle resolution, so as to realize independent retrieval and filtering of the gingival ring contour by region and ensure uniform distribution of circumferential points.
[0109] The first type of gingival margin candidate point refers to the real three-dimensional vertex in the original intraoral scanning grid within the corresponding sector angle neighborhood, whose error with the expected trajectory is within the preset tolerance range. Real scanning data is used first to ensure anatomical fitting accuracy.
[0110] The second type of gingival margin candidate point refers to the theoretical compensation point directly taken from the polar angle corresponding to the expected gingival margin trajectory when there is no qualified scanning vertex in the neighborhood of the corresponding sector angle, so as to solve the boundary breakage problem caused by scanning defects and grid noise.
[0111] Smoothing filtering (Savitzky-Golay periodic window filtering) refers to a periodic boundary smoothing algorithm for closed annular point sequences. It eliminates vertex step abrupt changes and local convex and concave noise caused by scanning grid sampling, and outputs a closed gingival margin line with continuous curvature, smoothness and no distortion.
[0112] The working principle of this embodiment is as follows:
[0113] In step S21, four initial feature points are interactively selected in the annular region of the gingival cuff of the intraoral scanning mesh. Only a small number of points need to be manually marked, and there is no need to manually outline the entire circumference, which greatly reduces the amount of manual operation. The four selected points are evenly distributed around the annular gingival cuff, completely covering the circumferential geometric features of the annular contour, providing a reference constraint for the subsequent coordinate system construction.
[0114] In step S22, the geometric mean of the four initial points is calculated as the origin of space. After the point set is centered, the PCA covariance matrix is calculated. The principal eigenvectors are extracted to determine the optimal normal vector of the plane as the Z-axis of the cylindrical coordinate system. Then, the X-axis and Y-axis are generated through orthogonal operation of plane vectors, thus completing the construction of the first cylindrical coordinate system adapted to the gingival ring contour. The cylindrical coordinate system is naturally adapted to the ring closed contour, which can simplify the solution of complex three-dimensional ring boundary to two-dimensional polar coordinate interpolation operation, reducing the computational complexity of the algorithm.
[0115] In step S23, all four initial marker points are transformed to a cylindrical coordinate system, and three sets of parameters—polar angle, polar radius, and axial height—are decomposed. The point set is sorted according to the polar angle values from smallest to largest and the first and last periodic splices are performed to simulate a closed ring boundary. Based on the ordered point set, cubic spline interpolation is used to construct the expected polar radius function and the expected height function that change continuously with the polar angle. The two functions are combined to form a continuous and uninterrupted expected gingival margin trajectory, which serves as a standard reference for subsequent point selection.
[0116] In step S24, the grid vertex partitioning retrieval, candidate point compensation, and smoothing are completed. First, all 3D vertices of the intraoral scan grid are transformed to the same first cylindrical coordinate system, and multiple fan-shaped angular neighborhoods are evenly divided along the circumference. For each fan-shaped partition, the polar angle of the partition center is substituted into two expectation functions to calculate the theoretical target point. A distance penalty function is constructed to calculate the sum of squared errors between all original grid vertices and theoretical points within the partition. Real vertices with errors within the preset tolerance are selected as first-class candidate points. If there are no qualified real vertices in the partition due to scan defects or excessive noise, the corresponding point of the expected trajectory is directly used as the second-class compensation candidate point to ensure that there are no missing points in the circumference. After all fan-shaped partitions are traversed, all candidate points are connected in ascending order of polar angle to form an original closed point sequence. For the annular point sequence, Savitzky-Golay window smoothing filtering adapted to the periodic boundary is performed to eliminate vertex step errors and local burr distortion, and finally output a smooth, continuous, and completely closed abutment gingival margin line.
[0117] The above scheme simplifies the solution logic for the annular gingival margin by using a cylindrical coordinate system. It can automatically generate a complete gingival margin contour with only a few initial points marked, significantly reducing the workload of manual point marking. Employing a dual-candidate point mechanism of "prioritizing real scan points and compensating for theoretical trajectories," it can both reconstruct the true gingival anatomy based on the original scan data and compensate for scan defects, avoiding gingival margin line breaks. Periodic SG filtering eliminates contour burrs caused by mesh noise, resulting in a smooth and continuous gingival margin curvature, improving the fit between the abutment edge and gingival soft tissue, enhancing post-implantation soft tissue biocompatibility, and significantly improving the standardization of gingival margin contours output by different operators.
[0118] In some embodiments, step S3 specifically includes:
[0119] S31: Extract the open edges of the base mesh and filter out the bottom reference ring;
[0120] S32: Establish a second cylindrical coordinate system based on the bottom reference ring, and transform the bottom reference ring and the gingival margin line of the abutment to the second cylindrical coordinate system;
[0121] S33: Between the bottom reference ring and the abutment gingival margin line, a plurality of transverse cross-sectional layers are divided along the height direction, and at each polar angle position, the intermediate layer vertex on each transverse cross-sectional layer is generated according to the radial coordinate values of each point on the bottom reference ring, the axial coordinate values of each point on the bottom reference ring, the radial coordinate values of each point on the abutment gingival margin line, and the axial coordinate values of each point on the abutment gingival margin line.
[0122] Among them, a neck thickness retention section is provided on the side near the gingival margin line of the abutment, and the apex of the middle layer on each transverse cross-section layer in the neck thickness retention section maintains a constant radial distance from the gingival margin line of the abutment in the radial direction.
[0123] In the region below the neck thickness holding section, the radial concavity of the intermediate layer vertex is determined based on the relative positional relationship between the height of the intermediate layer vertex on each transverse cross section and the height of the bottom reference ring. The radial concavity varies with the polar angle and with the height. The radial concavity is zero near the lower boundary of the neck thickness holding section and gradually increases with decreasing height until it reaches its maximum value at the bottom reference ring.
[0124] S34: Convert the intermediate layer vertices on each transverse section layer back to three-dimensional space coordinates, construct topological relationships to generate mesh patches, and generate the gingival contour surface.
[0125] In this embodiment, the open edge of the base mesh refers to all the annular topological boundaries extending outward from the top of the titanium base mesh. All closed annular rings are extracted by traversing the mesh topological adjacency relationship, including multiple sets of annular boundaries such as the assembly limiting ring, the inner ring of the screw channel, and the outer reference ring.
[0126] The bottom reference ring refers to the outermost annular boundary selected from all closed and open edges of the base mesh through diameter threshold and axial height constraints. It corresponds to the assembly interface connecting the titanium base and the transgingival surface and serves as the bottom fixed constraint boundary of the transgingival surface.
[0127] The second cylindrical coordinate system refers to a local cylindrical coordinate system constructed by taking the geometric centroid of the bottom reference ring as the origin, extracting the height Z-axis by SVD singular value decomposition after decentering the vertices of the reference ring, and generating the X and Y axes by orthogonal operation. This system uniformly transforms the two sets of annular boundaries, the bottom reference ring and the gingival margin, to achieve parametric layered modeling of the upper and lower annular contours.
[0128] The transverse section layer refers to the parallel annular layers that are evenly divided along the Z-axis height direction between the bottom reference ring (bottom) and the abutment gingival margin line (top). Each layer corresponds to a fixed axial height. After generating the annular vertices layer by layer, they are stitched together to form a complete curved surface.
[0129] The neck thickness maintenance section refers to the fixed height range of the transgingival contour surface near the top of the gingival margin. The radial distance between all the layer vertices in this range and the gingival margin line remains constant, ensuring that the neck of the abutment has sufficient solid thickness, improving the mechanical strength of the structure, and preventing neck breakage during processing and use.
[0130] Radial concavity refers to the radial offset value of the layer apex relative to the straight-line layout position of the upper and lower boundaries. This value changes simultaneously with both axial height and circumferential polar angle. The concavity of the lower boundary of the cervical margin retaining segment is 0. The concavity increases linearly as it gets closer to the bottom reference ring, adapting to the differences in soft tissue thickness at different heights and orientations of the gingiva.
[0131] Mesh patch topology refers to the connection rules between triangular patches of adjacent vertices at the same height and between corresponding vertices in upper and lower layers. A continuous, crack-free, and non-overlapping closed surface mesh is constructed through ordered vertex indexing.
[0132] This embodiment generates a biomimetic vase-shaped transgingival surface through layered interpolation and segmented differential radial constraints, solving the technical problems of rigid morphology and easy compression of gingival soft tissue caused by traditional linear lofted surfaces. The specific implementation process is as follows:
[0133] In step S31, the topological adjacency boundaries of all triangular facets of the titanium substrate mesh are traversed, all closed ring open edges of the mesh are tracked, and the upper and lower limits of the ring diameter and the threshold of the axial height are set to complete the boundary screening. The inner ring of the screw channel and the small limiting ring are removed, and the outermost complete ring contour is retained as the bottom reference ring. This reference ring perfectly matches the upper assembly shape of the titanium substrate, ensuring that the bottom of the through-gum curved surface fits seamlessly with the titanium substrate without assembly gaps.
[0134] In step S32, the geometric centroids of all vertices of the bottom reference ring are extracted and used as the origin of the coordinate system. After the vertices are decentered, singular value decomposition (SVD) is performed to extract feature vectors to determine the Z-axis of the height direction of the transgingival structure. Vector orthogonal operation is performed to generate the plane X and Y axes, thus completing the construction of the second cylindrical coordinate system. The two sets of annular boundaries, namely the bottom reference ring and the abutment gingival margin line generated in step S2, are all transformed to this coordinate system and uniformly decomposed into polar angle, polar radius, and axial height parameters to achieve parameterized alignment of the upper and lower sets of annular boundaries, providing a unified calculation benchmark for layered interpolation.
[0135] In step S33, multiple transverse cross-sectional layers are uniformly divided along the Z-axis height between the bottom reference ring and the gingival margin. For each circumferential polar angle position, the radial and axial coordinates of the bottom reference ring and the corresponding radial and axial coordinates of the gingival margin are read. Layered interpolation is used to calculate all the intermediate vertices on each cross-sectional ring. The layered interpolation distinguishes two segments of differentiated radial constraint logic. The upper segment is set as a cervical margin thickness maintenance segment, in which the radial distance of all layered vertices within the interval is fixed to the gingival margin, maintaining a constant wall thickness of the abutment neck and avoiding the risk of fracture due to excessive thinness of the neck structure. The interval below the cervical margin maintenance segment is set with a gradient of radial indentation. The indentation gradually increases as the axial height decreases, reaching its maximum value at the bottom reference ring. At the same time, the indentation is synchronously fine-tuned with the circumferential polar angle to adapt to the differences in soft tissue thickness in different gingival directions, so that the curved surface naturally tapers inward to form a vase-like bionic shape, which is different from the traditional flat curved surface with direct linear stretching of the upper and lower boundaries.
[0136] In step S34, all the middle vertices of each transverse cross section are converted from cylindrical coordinates back to three-dimensional Cartesian space coordinates. Based on the polar angle sorting and the correspondence between the upper and lower layers, the topological connection rules of the triangular facets are established. All vertices are stitched together layer by layer to generate a complete continuous mesh facet. Finally, a transgingival contour surface with the upper and lower edges respectively attached to the titanium substrate and the gingival margin line, and the middle section smoothly retracting inward is obtained.
[0137] The above scheme achieves precise positioning of the bottom of the transgingival curved surface by selecting the outer ring reference ring of the titanium base, and completes parametric interpolation of the upper and lower ring boundaries using a unified cylindrical coordinate system. The cervical margin thickness preservation segment ensures the mechanical strength of the abutment neck and avoids fracture in weak areas; the gradient-variable radial concave structure in the lower region conforms to the natural gingival anatomy, eliminating the problems of gingival compression, gingival inflammation, and recession caused by excessive outward expansion of traditional linear lofted curved surfaces. The circumferential concavity of the curved surface adaptively adjusts with the polar angle, greatly improving soft tissue fit and long-term aesthetic effects. The overall grid of the curved surface is continuous and gapless, and can be directly used for 3D printing and machining.
[0138] In some embodiments, after generating the transgingival contour surface in step S3, the method further includes fine-tuning the transgingival contour surface, specifically including:
[0139] Receive user input of adjustment height ratio position parameters and fine-tuning radius deformation parameters;
[0140] Extract the boundary of the transgingival contour surface and separate it into a bottom boundary feature point set and a top boundary feature point set;
[0141] Keeping the coordinate values of the bottom boundary feature point set and the top boundary feature point set unchanged, based on the adjusted height ratio position parameter, the target height value of the internal mesh vertex in the axial direction of the middle section of the transgingival contour surface is determined. The target height value is calculated based on the relative positional relationship between the height of the internal mesh vertex and the height of the bottom boundary feature point set and the ratio coefficient indicated by the adjusted height ratio position parameter.
[0142] Based on the fine-tuning radius deformation parameter, the target radial coordinate value of the inner mesh vertex in the radial direction is determined. The target radial coordinate value is calculated by weighting the original radial coordinate value of the inner mesh vertex, the deformation indicated by the fine-tuning radius deformation parameter, and the relative distance between the inner mesh vertex and the bottom boundary feature point set and the top boundary feature point set.
[0143] The internal mesh vertices are moved to the positions corresponding to the target height value and the target radial coordinate value to generate the fine-tuned transgingival contour surface.
[0144] Step S4 includes: generating an abutment preparation body based on the finely adjusted transgingival contour surface and the initial crown mesh.
[0145] In this embodiment, the height ratio position parameter refers to the axial ratio coefficient input by the user, which is used to control the distribution position of the middle section region with the most significant surface deformation in the overall height of the transgingival area. The value range is 0-1, where 0 represents near the bottom base and 1 represents near the top gingival margin.
[0146] The fine-tuning radius deformation parameter refers to the radial offset control value input by the user. A positive number represents the overall outward bulge of the middle section of the surface, while a negative number represents further inward contraction. This is used to match the personalized needs of patients with local gingival thickening or thinning.
[0147] The bottom boundary feature point set refers to the set of annular vertices connecting the bottom of the transgingival contour surface and the titanium substrate. The coordinates are locked and fixed throughout the fine-tuning process to ensure that the assembly accuracy of the abutment and the titanium substrate is not affected by deformation.
[0148] The top boundary feature point set refers to the set of ring-shaped vertices that fit the top of the transgingival contour surface with the gingival margin. Fine-tuning the coordinates throughout the process maintains a perfect fit between the top boundary and the gingival cuff, preventing boundary misalignment and gaps.
[0149] The cKD-Tree nearest neighbor retrieval algorithm is a fast spatial nearest neighbor retrieval algorithm that quickly distinguishes between the vertices of the surface boundary and the vertices of the internal middle section mesh, efficiently separates the bottom and top sets of boundary feature points, and reduces the computation time of spatial retrieval.
[0150] The height remapping mechanism refers to the mapping operation that recalculates the target axial height of the vertex based on the original axial relative height of the vertex and the user-input ratio parameter, so as to accurately locate the deformation center to the height range specified by the user.
[0151] Relative distance weighting refers to using the spatial distance between the internal grid vertices and the bottom and top boundaries as weights. The closer to the two boundaries, the greater the attenuation of radial deformation. Only the middle section of the surface achieves complete deformation, and the transition to the upper and lower boundaries is smooth without abrupt faults.
[0152] This embodiment provides a boundary-locked interactive fine-tuning function after the basic transgingival surface is generated. This solves the problem of synchronous distortion of the upper and lower boundaries during traditional surface adjustments, which compromises the precision of assembly and soft tissue adhesion. The fine-tuning operation only changes the shape of the middle section of the surface, while the upper and lower connecting boundaries maintain their original geometric coordinates throughout. The transgingival contour surface before and after fine-tuning is as follows: Figure 5 As shown.
[0153] The system receives two sets of intuitive control parameters input by the operator: the height ratio position parameter and the fine-tuning radius deformation parameter. Then, it calls the cKD-Tree spatial nearest neighbor search algorithm to traverse all the grid vertices of the transgingival contour surface. By using the spatial distance threshold between the vertex and the ring boundary, it separates the bottom boundary feature point set and the top boundary feature point set. The coordinate values of the two boundary point sets are completely locked during the entire fine-tuning operation and do not participate in any displacement adjustment, thus eliminating boundary misalignment, gaps, and distortion from the underlying operation logic.
[0154] The internal vertex axial height remapping calculation is performed based on the second cylindrical coordinate system. For each unlocked internal mesh vertex in the middle section of the surface, the original ratio of the current axial height of the vertex to the total height of the bottom boundary is calculated. Combined with the user-input adjustment height ratio position parameter, adaptive remapping is performed to solve the target axial height of the vertex after deformation. This accurately controls the surface concave and convex deformation to be concentrated in the user-specified height range, without problems such as deformation area offset or overall surface stretching.
[0155] Radial deformation employs a dual-boundary distance weighted calculation logic. It reads the polar radius value of the original cylindrical coordinate system of the vertex, combines it with the fine-tuning radius deformation parameter input by the user, and then allocates deformation weights based on the spatial distance of the vertex from the bottom and top boundaries. The weight of vertices near the top and bottom boundaries approaches 0, with minimal radial offset, while the weight of the middle region is the largest. It fully executes the set radial inward or outward offset, ensuring that the deformation smoothly transitions from the central region to both boundary ends without obvious bends or breaks.
[0156] After all internal mesh vertices have completed the axial height and radial polar radius coordinate updates, the triangular facet normals and vertex index topology are uniformly and synchronously corrected, and the fine-tuned through-gum contour surface with the upper and lower boundaries completely retaining the original fitting state and the middle section shape being shaped as needed is output. The subsequent abutment preparation body generation process in step S4 is all based on this optimized surface.
[0157] The above-described scheme enables independent fine-tuning of the mid-section of the transgingival surface, with the bottom and top boundaries locked throughout. The adjustment process does not compromise the assembly precision of the titanium base or the soft tissue adhesion boundary at the gingival margin, overcoming the shortcomings of traditional surface deformation methods that involve boundary distortion. Only two simple numerical values are required to control the deformation position and the degree of concavity / convexity; the operation is simple and intuitive, adapting to the varying thickness of the gingiva in different patients for highly personalized shaping. Deformation uses a distance-weighted smooth transition, resulting in no sharp angles or steps after surface adjustment, leading to a more uniform distribution of soft tissue stress and further reducing the risk of gingival compression. After fine-tuning, the model requires no additional boundary repair and can be directly integrated into the subsequent automated modeling process for abutment preparation.
[0158] After generating the transgingival contour surface in step S3 and fine-tuning in step S4, an automated finite element iterative optimization sub-process can be added, as follows: The fine-tuned transgingival contour surface is imported into the soft tissue finite element simulation module. Using the overall gingival stress uniformity as the optimization objective function, the radial concavity of each layer of the surface is iteratively adjusted using gradient descent or a genetic algorithm. The simulation identifies high-stress compression areas, automatically expanding the surface vertices outward and appropriately retracting low-stress support areas. After iterating until the stress uniformity reaches a preset uniformity threshold, the optimal surface is output. This eliminates the need for repeated manual adjustments and can uniformly distribute gingival implantation stress at the algorithmic level, reducing the risk of postoperative swelling, pain, and recession.
[0159] In some embodiments, step S41 specifically includes:
[0160] If the current tooth position is a molar, the three-dimensional convex hull of the initial crown mesh is calculated. Among the upward-facing convex hull patches, candidate convex hull patches whose angle between the normal vector of the convex hull patch and the preset reference direction is less than a preset angle threshold are retrieved. The target convex hull patch with the largest surface area is selected from the candidate convex hull patches, and the unit positive normal vector of the target convex hull patch is determined as the path of insertion.
[0161] If the current tooth position is a canine or anterior tooth, calculate the average coordinates of all grid vertices of the initial crown mesh. Establish a three-dimensional spatial coordinate system with the point corresponding to the average coordinates as the origin. Calculate the global centroid of the initial crown mesh, which is the weighted average of all grid vertices, with the weight being the area of each grid facet. Determine the vector from the origin to the global centroid as the initial axis. Extract the vertex clusters whose projection height on the initial crown mesh along the initial axis is within a preset height ratio range. Calculate the local centroid of the vertex cluster and determine the vector from the origin to the local centroid as the final placement path direction.
[0162] In this embodiment, the three-dimensional convex hull refers to the smallest convex polyhedron that encloses all vertices of the initial crown mesh. It can completely extract the geometric features of the outer surface of the crown without depressions, strip away local concave structures such as pits and grooves, and retain only the macroscopic outer contour patch of the crown. It is suitable for extracting features of the flat occlusal surface of molars.
[0163] The preset reference direction refers to the vertical reference vector of oral anatomy, which is parallel to the vertical growth direction of the alveolar bone and serves as a unified reference standard for determining the orientation of the occlusal surface.
[0164] The preset angle threshold is a value used to filter out approximately horizontal occlusal surfaces. Only when the angle between the normal vector of the convex hull surface and the vertical reference direction is less than this threshold is it determined to be an approximately horizontal occlusal bearing surface.
[0165] The global centroid refers to the overall centroid of the crown calculated using the area of the mesh facets as the weight. Unlike the average value of ordinary vertices, it can avoid interference from local dense vertices and accurately reflect the position of the overall geometric centroid of the crown.
[0166] The preset height ratio range refers to the upper height interval of the crown divided along the initial axis. It is generally taken as the first 30% of the total axial height, corresponding to the occlusal surface of the crown and the top area where the cusp is located, to filter the interference of the cervical structure of the lower half of the crown on the calculation of the path of insertion.
[0167] The local centroid refers to the geometric center point of the vertex set of the crown top region. It is used to finely calibrate the initial axis of the global centroid output to conform to the long axis anatomical trend of the anterior teeth and canines, which narrows from the neck to the incisal edge.
[0168] The unit positive normal vector refers to the standardized unit vector of the convex hull plate that is perpendicular to the surface of the plate and faces outward from the crown. It represents the optimal bearing and assembly direction of the occlusal surface and serves as the reference for the molar's dedicated insertion path.
[0169] This embodiment designs two independent placement path solution logics to address the anatomical differences between molars and anterior or canines, solving the shortcomings of traditional general algorithms that use a one-size-fits-all approach, leading to mismatches between the placement path and crown morphology and clinical insertion difficulties. The specific process is as follows: When the system identifies the current restoration tooth as a molar, it executes the convex hull screening logic. First, it solves the three-dimensional convex hull for all vertices of the initial crown mesh, removing pits and grooves and local depressions on the crown surface, retaining only the complete external convex hull. Then, it traverses all upward-facing convex hull hull hulls, calculating the angle between the unit normal vector of each hull and the preset anatomical vertical reference direction, and screening candidate hull ...
[0170] When the system identifies the current restoration tooth as a canine or anterior tooth, it executes a two-layer centroid calibration axial logic. The first step calculates the average coordinates of all vertices of the initial crown mesh as the spatial origin, thus establishing a dedicated 3D computational coordinate system. The second step calculates the global centroid of the crown using a facet area weighting method; the larger the area of a single facet, the higher the weight of the corresponding vertex, avoiding the result of dense vertices at the cervical region shifting the centroid. The spatial vector pointing from the origin to the global centroid is normalized to obtain the initial axial direction. The third step sorts all mesh vertices along the initial axial direction by projection height, extracting the top vertices whose projection height is within the top 30% of the total axial height. This top vertex set contains the cusps, incisal edges, and core occlusal structures. The fourth step calculates the local centroid of the top vertex set, correcting the initial axial direction with the vector pointing from the origin to this local centroid, obtaining the final placement path direction that conforms to the narrowing shape of the anterior or canine from the cervical to the incisal edge.
[0171] The entire logic automatically switches the calculation path according to the tooth position, eliminating the need for manual dragging and adjustment of the axis. It relies on the mesh geometry features to adaptively output a unique assembly axis that matches the anatomical morphology, providing a precise reference coordinate axis for subsequent detection ray emission and sidewall modeling.
[0172] The above scheme differentiates the anatomical features of molars, anterior teeth, and canines to calculate the placement path, abandoning the single, universal bounding box estimation method. For molars, the assembly axis perpendicular to the occlusal surface is locked using the largest area horizontal convex facet, adapting to wide occlusal structures; for anterior teeth and canines, a dual-layer calibration using global centroid and top local centroid is employed to fit the long axis of the narrow tooth. Compared to manual visual judgment of the axis, the algorithm outputs standardized results without subjective errors, fundamentally avoiding problems such as difficulty in clinical placement of restorations, axial displacement, and jamming, significantly reducing the clinical rework rate.
[0173] In some embodiments, step S45 includes a step of generating the side of the base, specifically including:
[0174] Harmonic fitting is performed on the axial height of the intersection of the probe line and the initial crown grid to extract the overall height trend of the circumferential variation of the top boundary ring along the transgingival contour surface. The local drop positions on the fitted curve are adaptively corrected, and the fitted fluctuations are scaled to a preset range and then superimposed on the safe design height to obtain the sidewall top boundary ring. The local drop position refers to the local point that has an abnormal downward shift relative to the adjacent circumferential position or the overall low-frequency trend.
[0175] Starting from the top boundary ring of the transgingival contour surface, multiple sidewall levels are divided along the placement path to the top boundary ring of the sidewall to generate the abutment side.
[0176] In this embodiment, harmonic fitting (periodic low-frequency Fourier fitting) refers to a periodic curve fitting algorithm for the circumferential height sequence of the ring. It filters out high-frequency abrupt changes caused by dental grooves and local peaks, extracts the low-frequency macroscopic height trend that changes uniformly along the ring, preserves the overall smooth undulation of the inner wall of the crown, and eliminates local abnormal low points.
[0177] In the height trend obtained by fitting the periodic low-frequency harmonics of the local sudden drop position, there are local points that are abnormally offset downward relative to the adjacent circumferential position or the overall low-frequency trend. These points usually correspond to the probe rays projected onto the deep pits and grooves of the crown, local depressions, or abnormal scanning areas.
[0178] Axial convergence degree refers to the overall taper parameter that controls the upward convergence of the base sidewall along the positioning path. The larger the value, the greater the inward contraction of the sidewall. It can be flexibly adjusted according to the retention requirements.
[0179] Marginal retraction refers to the radial overall shrinkage dimension of the top boundary of the lateral wall relative to the bottom transgingival ring, used to reserve a basic safety gap for the inner wall of the crown.
[0180] The initial transition region of the abutment sidewall near the top boundary ring of the transgingival contour surface accounts for the proportion of the overall sidewall level. This is used to control the process of the edge retraction gradually increasing from zero to a preset edge retraction amount. After the transition region, the edge retraction amount remains at a preset value, while radial contraction continues to occur as the axial height increases according to the axial convergence degree, thereby allowing the abutment sidewall to smoothly transition from the top of the transgingival contour surface to the coronal sidewall with a preset convergence degree.
[0181] The top boundary ring of the sidewall refers to the closed ring profile after adaptive correction and amplitude limiting, which serves as the top boundary of the base sidewall and also as the outer constraint boundary for subsequent top surface modeling.
[0182] The adaptive correction process is as follows: First, the top boundary loop of the sidewall is formed by Fourier curves fitted through multiple iterations. In each iteration, the weight of local low points in the next round of fitting is reduced, thereby gradually extracting the initial low-frequency height trend. After obtaining this low-frequency trend, the average height of two adjacent points is used as a local reference benchmark. If the depression of the midpoint relative to this reference height exceeds the preset maximum allowable depression, the point is corrected upwards to within the allowable range, thus completing the curve fitting.
[0183] This embodiment uses the full-circumferential intersection height sequence obtained by ray detection in step S43 to eliminate local depression interference through periodic harmonic fitting, generating a sidewall profile with a stable shape and no local collapse. The specific process is as follows:
[0184] First, the axial height data of all intersections between the probe rays and the initial crown mesh in steps S42 and S43 are read. This set of height data is distributed in a ring-shaped periodic pattern along the transgingival top boundary ring. Periodic low-frequency harmonic fitting is performed on the sequence to separate high-frequency noise from low-frequency macroscopic fluctuations, extracting the overall height trend of the crown inner wall that changes slowly along the ring, and filtering out abrupt height changes caused by small pits and fissures and scanning distortion. Then, all points on the fitted curve are traversed, and adaptive correction is performed on local sudden drop points to eliminate extreme low points caused by ray projection onto deep dental grooves, preventing large local depressions and structural weakness of the lateral wall. After adaptive correction, the overall fluctuation amplitude of the curve is scaled to a preset range to avoid excessive circumferential height differences that could lead to drastic changes in lateral wall taper. Finally, the processed ring-shaped height curve is superimposed on the safe design height calculated in step S44 to generate a completely closed lateral wall top boundary ring. This contour takes into account the macroscopic spatial fluctuations of the crown inner wall without any local abnormal collapse.
[0185] Then, using the top boundary ring of the transgingival contour surface as the starting point of the sidewall bottom, and combining three adjustable parameters preset by the user—axial convergence degree, edge retraction amount, and transition ratio—multiple sidewall layers are evenly divided between the bottom starting point and the top boundary ring of the sidewall along the axial direction of the placement path. The radial shrinkage amount of the annular apex of each layer is calculated layer by layer, and the taper variation amplitude is smoothly distributed according to the transition ratio. After generating the annular apex layer by layer, the triangular facets are sutured to form an overall smooth conical abutment sidewall structure. The number and position of the bottom of the sidewall correspond one-to-one with the top of the transgingival contour surface, with seamless connection and no gaps.
[0186] The above scheme extracts the macroscopic trend of the crown's internal height through harmonic fitting, adaptively corrects the low points of the fissures, and avoids local collapse defects of the abutment sidewalls. It adjusts the sidewall taper and shrinkage amplitude using three sets of controllable parameters, balancing axial retention force with a safe clearance in the crown's inner wall. The bottom of the sidewall perfectly matches the apex of the transgingival surface, resulting in a smooth and continuous model transition. This filters out scanning noise and morphological distortions caused by changes in the dental grooves, improving the abutment's structural stability and fabrication quality.
[0187] In some embodiments, step S45 includes a step of generating the top surface of the base, specifically including:
[0188] Using the top boundary ring of the sidewall as the outer boundary of the top surface, multiple ring-shaped sampling points are established from the outside to the inside within the area enclosed by the outer boundary of the top surface. Each sampling point is projected onto the occlusal surface of the initial crown grid along the path of placement to obtain the crown projection height value of each sampling point. For local sampling points that fail to be projected, periodic interpolation compensation is performed based on the crown projection height values of adjacent valid sampling points on the same ring.
[0189] Based on the radial position of each sampling point, a base height of the abutment that smoothly changes from the top boundary of the sidewall to the center is established. Based on the projection height of the outer region of the crown and the projection height of the multiple crown rings near the center, a reference height of the crown that smoothly changes from the periphery to the center is established.
[0190] Calculate the height difference between the crown projection height value of each sampling point and the corresponding crown reference height to obtain the local undulation characteristics of the crown occlusal surface relative to the overall height trend; retain positive height differences with a larger undulation retention ratio and limit the maximum allowable upward convexity, and retain negative height differences with a smaller undulation retention ratio and limit the maximum allowable downward concavity.
[0191] Based on the radial position of each sampling point from the outer boundary, the smoothing introduction weights of positive undulation features and negative concavity features are determined respectively, so that the crown morphological features are gradually introduced from the outer boundary to the interior of the top surface; the processed positive undulations and negative undulations are superimposed on the abutment base height according to the corresponding smoothing introduction weights to obtain the final target top surface height of each sampling point.
[0192] Using the final target top surface height of each sampling point as the fitting data, a robust low-frequency surface fitting algorithm is used to generate a continuous and smooth abutment top surface, so as to preserve the overall gentle undulation trend of the crown occlusal surface and filter out small grooves, local peaks and scanning noise.
[0193] A smooth transition region is set between the top boundary ring of the sidewall and the fitted top surface of the base. The outer periphery of the smooth transition region maintains the original height of the top boundary ring of the sidewall, and the height and radial slope of the top surface of the base are gradually matched from the outer periphery to the inner periphery, thereby obtaining a smooth top surface of the base that is continuously connected to the side of the base.
[0194] In this embodiment, the multi-ring sampling points refer to multiple sets of concentric ring discrete sampling points distributed from the outside to the inside with the geometric center of the top surface region as the center. These points are used to perform discrete sampling of the top surface space and collect the height information of the crown and occlusal surface. The number of sampling rings and the number of sampling points per ring can be adaptively adjusted according to the model accuracy requirements.
[0195] Periodic interpolation compensation refers to an interpolation method for a closed ring set of points. Since the sampling points are arranged in a closed ring, the first and last points are connected during interpolation calculation, avoiding the jumps and breaks that occur at the beginning and end of the ring in ordinary linear interpolation.
[0196] The abutment base height refers to the reference height surface where the top surface smoothly transitions from the periphery to the center without superimposing the occlusal undulation features of the crown, ensuring that the top surface has a basic geometric shape and solid thickness.
[0197] The crown reference height refers to the overall macroscopic height trend surface obtained by fitting the crown occlusal surface sampling height, filtering out local abrupt changes, and is used to separate the overall undulations and local minor concavities and convexities of the crown.
[0198] The height difference refers to the difference between the actual projected height of the tooth crown at the sampling point and the reference height of the tooth crown; a positive difference indicates that the tooth crown at that location tends to bulge upwards relative to the overall trend (positive height difference); a negative difference indicates that the tooth crown at that location tends to bulge downwards relative to the overall trend (negative height difference).
[0199] The undulation retention ratio, maximum allowable convexity, and maximum allowable concavity are configurable parameters. The undulation retention ratio is used to control the degree of inheritance of local concave and convex features. The maximum allowable convexity and maximum allowable concavity are used to cut off extreme concavities and convexities that exceed the safe range to prevent interference.
[0200] The smoothing introduction weight refers to a weighting coefficient related to the radial distance of the sampling point from the outer boundary. In the region near the top boundary ring of the sidewall, the weight is smaller to maintain the stability of the outer boundary; as it moves from the outer edge to the interior of the top surface, the weight gradually increases, so that the local undulations of the crown after scaling and amplitude limiting are gradually introduced into the top surface of the abutment.
[0201] Robust low-frequency surface fitting algorithm refers to a surface fitting algorithm that resists the interference of outliers. It prioritizes fitting the low-frequency macroscopic change trend of the data and suppresses the disturbance of individual outlier sampling points (noise, spikes, small grooves) on the overall surface. In other embodiments, other surface fitting methods that can achieve low-frequency trend fitting and outlier suppression can be used. In this embodiment, a two-dimensional polynomial is preferably used for robust low-frequency fitting.
[0202] The smooth transition zone refers to the annular transition zone set between the top boundary ring of the sidewall and the top surface of the internal base, which eliminates the sharp corners and steps at the splicing of the sidewall and the top surface, and ensures the geometric continuity of the mesh.
[0203] The working principle of this embodiment is as follows:
[0204] The closed region enclosed by the top boundary ring of the sidewall is used as the top surface generation domain. Multiple concentric ring sampling points are generated from the outside to the inside within this region. All sampling points are projected parallel to the path of insertion onto the occlusal triangular facet of the initial crown mesh, and the crown projection height value corresponding to each sampling point is read. Due to objective factors such as scanning defects and mesh voids, some sampling points may fail to project and thus cannot obtain a valid height. For these failed sampling points, periodic interpolation compensation is performed using adjacent valid sampling points on the same ring to calculate the corresponding height. Closed-loop interpolation logic ensures the continuous and complete sampling height of the entire ring. The number of sampling rings can be determined based on the radial dimension of the top surface and the preset lattice density. The number of circumferential points in each sampling ring is consistent with the number of circumferential points in the top boundary ring of the sidewall.
[0205] After completing the sampling point height acquisition and completion, based on the radial position of each sampling point, a base height of the abutment is constructed that smoothly changes from the top boundary of the sidewall towards the center, ensuring that the top surface of the abutment has a basic solid thickness. At the same time, combined with the projected height of the outer region of the crown and the projected height of multiple crown rings near the center, a crown reference height that smoothly changes from the periphery to the center is fitted to establish a reference height of the crown, thereby distinguishing the overall macroscopic trend of the crown occlusal surface and local fragmented abrupt changes.
[0206] Subsequently, the height difference between the crown projection height at each sampling point and the corresponding crown reference height was calculated to obtain the local undulation characteristics of the crown occlusal surface relative to the overall height trend. For positive height differences, a larger undulation retention ratio was used for feature preservation, and a maximum allowable upward convexity was set as an upper limit truncation to prevent the abutment top surface from hitting the crown excessively high. For negative height differences, a smaller undulation retention ratio was used for feature preservation, and a maximum allowable downward concavity was set as a lower limit truncation to avoid excessive downward depression of the top surface causing local structural weakness. The degree of preservation of positive undulations was greater than that of negative concavities, and only a small proportion of negative dental grooves were preserved; independent maximum allowable variations were set for the positive and negative directions respectively.
[0207] Then, based on the radial position of each sampling point from the outer boundary, the smoothing introduction weights corresponding to the positive undulation features and negative concavity features are determined respectively, so that the crown morphological features are gradually enhanced from the outer boundary to the interior of the top surface. Then, the positive and negative undulations after amplitude limiting and scaling are superimposed on the abutment base height according to the corresponding smoothing introduction weights to obtain the final target top surface height of each sampling point. By relying on the radially gradual weight setting, a smooth transition of the top surface morphology from the outer circle to the center is achieved, avoiding abrupt morphological breaks.
[0208] After obtaining the target height of all sampling points, the final target top surface height of each sampling point is used as the fitting control point. A robust low-frequency surface fitting algorithm is employed to generate a continuous and smooth abutment top surface. During the fitting process, the overall undulation trend of the crown occlusal surface is preserved, while abnormal control points caused by small grooves, local peaks, and scanning noise are filtered out to avoid the risk of point contact interference caused by directly replicating fragmented anatomical structures. Finally, a smooth transition region is set between the top boundary ring of the sidewall and the fitted abutment top surface. The outer periphery of the smooth transition region maintains the original height of the top boundary ring of the sidewall, and the height and radial slope of the abutment top surface are gradually matched from the outer periphery to the inner periphery, ultimately resulting in a smooth abutment top surface that is continuously connected to the side surface of the abutment, achieving an integrated connection between the side and top surface meshes.
[0209] This approach effectively replicates the anatomical features of the dental crown and maxillofacial region while protecting against interference. It avoids simply copying all the convex and concave details of the crown and maxillofacial region. By distinguishing between macroscopic trends and minor local undulations, it differentiates the preservation and limitation of protrusions and depressions, balancing anatomical fit with interference prevention. Utilizing a smoothing introduction weight tied to radial distance, it achieves a gradual transition of the top surface morphology from the outer edge to the center. The outer region prioritizes maintaining the stability of the lateral wall top boundaries, gradually introducing smoothed, scaled, and limited crown morphology information into the interior of the top surface. Simultaneously, robust center reference and low-frequency surface fitting suppress the influence of local depressions such as central fossae and fine grooves on the final top surface. This effectively avoids excessive central concavity of the top surface, ensuring the overall mechanical strength of the abutment top surface.
[0210] To address the issue of sampling point projection failure caused by scanning defects, a circular periodic interpolation compensation method is used to complete the data. Even if the input clinical scan mesh has local defects, it can still output complete and usable top surface sampling data, improving the algorithm's fault tolerance for real clinical scan data. A specially designed smooth transition area between the sidewalls and the top surface eliminates right angles and sharp corners at the splicing interface, alleviates stress concentration under occlusal conditions, improves the mechanical performance of the abutment, and generates a smooth and continuous overall mesh that can be directly adapted to machining or 3D printing processes without the need for manual grinding of seams.
[0211] The entire top surface generation process is fully automated, eliminating the need for technicians to manually outline the top surface contour. This reduces reliance on individual operator experience, minimizes design fluctuations caused by human error, and effectively improves the standardization and output efficiency of personalized base designs. The robust low-frequency surface fitting algorithm used in this step can be replaced with robust thin-plate spline fitting, anti-outlier B-spline surface fitting, or other equivalent algorithms. The number of ring sampling points and the number of points per ring can also be flexibly adjusted according to actual processing accuracy requirements. All of the above replacement methods can achieve the corresponding technical effects of this invention.
[0212] Based on the abutment top surface generation logic, tooth position type and local crown curvature can be introduced as correction factors to further improve accuracy. Specifically, this includes: when calculating the height ratio coefficient, a larger neighborhood radius is used for molars to capture the overall shape of a flat occlusal surface, while a smaller neighborhood radius is used for anterior teeth / canines to preserve local details of the incisal edge; simultaneously, the local curvature of the crown mesh at the sampling point is read, and when the curvature exceeds a preset curvature threshold, the target height ratio is reduced to avoid spatial interference caused by abrupt changes in the curvature of cusps and fissures. This optimization scheme can serve as an advanced adaptive implementation of this invention to further improve the model adaptation accuracy of complex malformed crowns.
[0213] In this embodiment, based on the geometric calculation of the tooth placement path, a 3D point cloud deep learning model can be introduced to assist in calibration: a large number of clinically validated standard abutment design samples are collected, and the placement path prediction model is trained using PointNet++ and DGCNN dynamic graph convolutional networks; in actual use, the initial crown mesh is input into the model to obtain the predicted axis, which is then weighted and fused with the placement path vector obtained from the geometric calculation to obtain the final assembly axis. This compensates for the calculation errors of a single geometric algorithm for malformed or irregular crowns and improves the robustness of the placement path in extreme cases.
[0214] In some embodiments, after generating the abutment preparatory body, the method further includes:
[0215] Extract the screw channel boundary ring located inside the bottom reference ring of the base mesh, and determine the central axis and channel radius of the screw channel boundary ring;
[0216] Construct a screw channel cylinder based on the central axis of the screw channel boundary ring and the channel radius;
[0217] Based on the screw channel cylinder, a Boolean subtraction operation is performed on the base preparation body to delete the target mesh facet located inside the screw channel cylinder in the base preparation body, and the cutting boundary is reconstructed at the cutting intersection position to obtain the intermediate base model with the top opening ring.
[0218] Align the screw channel boundary ring with the top opening ring on the axis so that the central axis of the screw channel boundary ring and the top opening ring coincide and the normal direction of the ring surface are consistent.
[0219] Between the screw channel boundary ring and the top opening ring, a connecting mesh is generated using a mesh reconstruction algorithm to serve as the inner wall of the screw channel, thus obtaining the final base preparation body.
[0220] In this embodiment, the screw channel boundary ring refers to the closed annular contour inside the reference ring at the bottom end of the base grid, which is the standard through-channel for implanting locking screws. The central axis of the screw channel boundary ring is coaxial with the implant interface.
[0221] A screw channel cylinder is a virtual cylindrical geometry constructed with the central axis of the channel ring as its center and the inner diameter of the channel ring as its radius. It is used for Boolean subtraction operations to remove excess mesh inside the base and form a screw through space.
[0222] Boolean subtraction refers to the operation of three-dimensional mesh entities, which removes triangular faces that are completely within the virtual cylinder inside the base preparation body, while calculating the intersection profile of the cylinder and the preparation body surface, and reconstructing the continuous cutting boundary.
[0223] The top opening ring refers to the boundary of the annular cutting hole formed on the top surface of the base preparation body after Boolean subtraction, which is the opening at the upper end of the screw channel.
[0224] Axis alignment constraint refers to synchronously matching the center axis and ring surface normal vector of the screw channel boundary ring and the top opening ring to ensure that the upper and lower holes are completely coaxial and without angular offset.
[0225] The mesh reconstruction algorithm refers to generating ordered triangular patches between two sets of coaxial annular boundaries to construct a smooth and continuous inner wall of a cylindrical channel, extending from the bottom to the top of the platform.
[0226] This embodiment defines an automated machining process for the integrated screw channel after the base preform is formed. It utilizes the original channel geometry parameters of the titanium substrate for precise hole drilling, ensuring the screw runs coaxially from top to bottom, eliminating the need for manual hole trimming. The specific process is as follows:
[0227] First, read the complete pre-built model of the base after the sidewalls and top surface are spliced. Then, backtrack and call the titanium base mesh data. Extract the closed boundary ring of the screw channel from the topological boundary inside the bottom reference ring. Solve the channel center axis by the geometric mean of the ring vertices. Take the maximum radial distance of the ring vertices to determine the standard radius of the channel. Construct an infinitely extending virtual screw channel cylinder with the axis as the center and the channel radius as the cross-sectional size, as the calculation reference for mesh cutting.
[0228] Next, perform a three-dimensional Boolean subtraction operation, traverse all triangular faces of the base preparation body, and delete all invalid faces that are completely enclosed inside the virtual cylinder; at the same time, solve the intersection contour line of the cylindrical surface and the outer surface of the base preparation body, and reconstruct a continuous and closed top opening ring along the contour line to obtain the middle base model with the screw hole reserved at the top.
[0229] Subsequently, a dual-ring coaxial alignment calibration was performed. The central axis and ring surface normal vector of the bottom titanium substrate screw channel boundary ring and the top open ring were extracted respectively. The spatial attitude of the top open ring was adjusted by rigid coordinate transformation so that the two sets of ring central axes were completely coincident and the ring surface normal vectors were in the same direction, thus eliminating the problems of misalignment of the upper and lower channels and screw insertion jamming.
[0230] Finally, a mesh reconstruction algorithm was used to match the number of vertices and the vertex arrangement order of the upper and lower sets of annular boundaries. Ordered triangular facets were generated layer by layer between the two rings to construct a smooth and continuous cylindrical screw channel inner wall, connecting the bottom titanium base channel with the opening on the top surface of the abutment. The inner wall has no steps or protrusions obstructing the view, ultimately resulting in a pre-built abutment model with an integrated coaxial screw channel, which can be directly output for denture fabrication. The generated abutment pre-built model is shown below. Figure 6 As shown.
[0231] The above solution directly reuses the original screw channel geometry parameters of the titanium substrate for hole opening, and uses Boolean subtraction to precisely remove the internal mesh, ensuring that the channel size perfectly matches the implanted screw. The upper and lower annular boundaries are forced to be coaxially aligned to avoid screw assembly misalignment and jamming; the continuous and smooth inner wall of the channel is automatically reconstructed, eliminating the need for technicians to manually grind holes, simplifying the processing procedure, ensuring smooth internal channel flow of the abutment, and adapting to clinical screw tightening operations.
[0232] After generating the abutment preparation body with screw channels, this invention can further add a gap closure correction step, as follows: The abutment preparation body is superimposed on the initial crown mesh; the shortest distance from all surface vertices of the abutment to the inner wall of the crown is calculated, generating a global gap distribution map; interference areas with gaps smaller than a preset safety threshold are identified; a gradual local deformation algorithm is used to shrink the abutment surface inwards, with the deformation smoothly decreasing from the correction area to the periphery, ensuring that the corrected abutment surface is smooth and continuous overall. This achieves fully automated interference detection and repair, eliminating the need for manual trial-and-error adjustments by technicians.
[0233] This invention's single-abutment generation process can be extended to the intelligent design of bridge abutments for multiple consecutive missing teeth: it unifies and globally registers multiple sets of registered intraoral, abutment, and crown meshes, and solves the common placement path of all abutments through global three-dimensional geometry; it synchronously calculates the spatial interference between the transgingival contour surfaces of adjacent abutments, automatically performs local contour setback correction, coordinates the spatial position between multiple abutments, and realizes fully automatic integrated modeling of the entire set of abutments for implant bridges, which is suitable for long-span continuous missing tooth restoration scenarios.
[0234] While maintaining the core process logic of this invention, each key step can be replaced with an equivalent algorithm to achieve the same technical effect, specifically including:
[0235] (1) In the step of generating the abutment gingival margin line, the PCA principal component analysis method used to construct the first cylindrical coordinate system can be replaced by the least squares spatial plane fitting and the general point cloud normal estimation algorithm; the Savitzky-Golay periodic smoothing filter used to eliminate vertex step noise can be equivalently replaced by Gaussian smoothing filter and Laplace smoothing filter, both of which can achieve the second-order curvature continuity of the closed gingival margin curve.
[0236] (2) The ray detection method for calculating the safety design height of the abutment preparation body can be equivalently replaced by the spatial voxelized bounding box interference detection algorithm and the symbolic distance field interlacing calculation algorithm. It can also traverse and obtain the total internal height of the initial crown, solve the maximum usable space height, and realize the interference-proof height constraint.
[0237] (3) When generating the transgingival contour surface, the sine function gradient concave generation logic of layered interpolation can be replaced with piecewise linear concave and quadratic curve concave functions. Only the concave amount change curve is changed, without changing the core constraints of the neck edge thickness maintenance segment and the upper and lower boundaries fitting.
[0238] (4) The abutment top surface fitting and side wall height fitting can replace B-spline surface and thin plate spline interpolation, and can achieve the effect of filtering scanning noise and preserving the macroscopic undulation of the crown.
[0239] All adjustable design parameters of this invention (safety design coefficient, axial convergence degree, marginal retraction, Bézier curve order, radial concavity extreme value, etc.) can be connected to a multi-objective optimization algorithm module (NSGA-II, MOEA / D); the optimization objectives simultaneously set the following: maximizing the minimum gap between the abutment and the crown, maximizing the axial retention force of the abutment, minimizing the average stress of the gingival soft tissue, and minimizing the overall volume of the abutment; the algorithm automatically generates a Pareto optimal solution set, and doctors can directly select the optimal parameter combination according to the patient's occlusal force, gingival thickness, and aesthetic requirements, without the need for manual repeated adjustment of multiple sets of parameters.
[0240] In a second aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the intelligent generation method for personalized dental implant abutments as described in the first aspect of the present invention.
[0241] The computer-readable storage medium may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory.
[0242] The non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD ROM); the magnetic surface memory may be a disk storage device or a magnetic tape storage device.
[0243] The volatile memory may be random access memory (RAM), which serves as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM), synchronous static random access memory (SSRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synclink dynamic random access memory (SLDRAM), and direct memory bus random access memory (DRRAM). The computer-readable storage media described in the embodiments of the present invention are intended to include these and any other suitable types of memory.
[0244] like Figure 7 As shown, in a third aspect, the present invention provides an electronic device 10, including a processor 101 and a storage medium 102, wherein a computer program is stored on the storage medium, and the computer program, when executed by the processor, implements the intelligent generation method for personalized denture implant abutments as described in the first aspect of the present invention.
[0245] In some embodiments, the processor may be implemented by software, hardware, firmware, or a combination thereof, and may use at least one of the following: circuit, single or multiple application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), central processing units (CPUs), controllers, microcontrollers, and microprocessors, thereby enabling the processor to execute some or all of the steps or any combination of the steps in the intelligent generation method for personalized denture implant abutments described in various embodiments of the present invention.
[0246] Finally, it should be noted that although the above embodiments have been described in the description and drawings of this invention, this should not limit the scope of patent protection of this invention. Any technical solutions that are based on the essential concept of this invention, utilize the content described in the description and drawings of this invention to make equivalent structural or procedural substitutions or modifications, as well as the direct or indirect application of the technical solutions of the above embodiments to other related technical fields, are all included within the scope of patent protection of this invention.
Claims
1. A method for intelligently generating personalized dental implant abutments, characterized in that, The method includes: S1: Obtain the registered intraoral scanning mesh, base mesh, and initial crown mesh; S2: Generate the abutment gingival margin line based on the intraoral scanning grid; S3: Based on the bottom reference ring of the base grid and the gingival margin line of the abutment, generate the transgingival contour surface; S4: Based on the gingival contour surface and the initial crown mesh, generate the abutment preparation body; Specifically, step S4 includes: S41: Determine the placement path direction based on the tooth position type of the initial crown mesh, wherein the placement path direction is the axial direction when the abutment is placed into the patient's oral cavity; S42: Starting from the point on the top boundary ring of the transgingival contour surface, generate multiple parallel probe lines along the path of insertion toward the initial crown mesh; S43: Calculate the intersection points of each of the probe lines and the initial crown grid, and calculate the height difference between each intersection point and the center point of the top boundary ring of the transgingival contour surface in the direction of the path of insertion. Take the minimum value among all height differences as the maximum usable space height. S44: Multiply the maximum available space height by a preset safety design factor to obtain the safety design height of the base preparation body; S45: Generate the base preparation body according to the safety design height.
2. The intelligent generation method for personalized dental implant abutments as described in claim 1, characterized in that, Step S2 specifically includes: S21: Extract multiple initial points at the edge of the gingival cuff of the intraoral scanning grid; S22: Establish a first cylindrical coordinate system based on the plurality of initial points; S23: Transform the plurality of initial points to the first cylindrical coordinate system, and perform interpolation based on polar angle sorting to generate the desired gingival margin trajectory that changes continuously with the polar angle; S24: Transform all grid vertices contained in the intraoral scanning grid to the first cylindrical coordinate system, and select grid vertices that meet the preset error conditions from all grid vertices contained in the intraoral scanning grid in multiple angular neighborhoods as first-type gingival margin candidate points; if there are no grid vertices that meet the preset error conditions in the corresponding angular neighborhood, then use the point at the corresponding polar angle on the expected gingival margin trajectory as the second-type gingival margin candidate point; connect each first-type gingival margin candidate point and each second-type gingival margin candidate point in polar angle order and perform smoothing filtering to generate the abutment gingival margin line.
3. The intelligent generation method for personalized dental implant abutments as described in claim 1, characterized in that, Step S3 specifically includes: S31: Extract the open edges of the base mesh and filter out the bottom reference ring; S32: Establish a second cylindrical coordinate system based on the bottom reference ring, and transform the bottom reference ring and the gingival margin line of the abutment to the second cylindrical coordinate system; S33: Between the bottom reference ring and the abutment gingival margin line, a plurality of transverse cross-sectional layers are divided along the height direction, and at each polar angle position, the intermediate layer vertex on each transverse cross-sectional layer is generated according to the radial coordinate values of each point on the bottom reference ring, the axial coordinate values of each point on the bottom reference ring, the radial coordinate values of each point on the abutment gingival margin line, and the axial coordinate values of each point on the abutment gingival margin line. Among them, a neck thickness retention section is provided on the side near the gingival margin line of the abutment, and the apex of the middle layer on each transverse cross-section layer in the neck thickness retention section maintains a constant radial distance from the gingival margin line of the abutment in the radial direction. In the region below the neck thickness holding section, the radial concavity of the intermediate layer vertex is determined based on the relative positional relationship between the height of the intermediate layer vertex on each transverse cross section and the height of the bottom reference ring. The radial concavity varies with the polar angle and with the height. The radial concavity is zero near the lower boundary of the neck thickness holding section and gradually increases with decreasing height until it reaches its maximum value at the bottom reference ring. S34: Convert the intermediate layer vertices on each transverse section layer back to three-dimensional space coordinates, construct topological relationships to generate mesh patches, and generate the gingival contour surface.
4. The intelligent generation method for personalized dental implant abutments as described in claim 1, characterized in that, After generating the transgingival contour surface in step S3, the method further includes fine-tuning the transgingival contour surface, specifically including: Receive user input of adjustment height ratio position parameters and fine-tuning radius deformation parameters; Extract the boundary of the transgingival contour surface and separate it into a bottom boundary feature point set and a top boundary feature point set; Keeping the coordinate values of the bottom boundary feature point set and the top boundary feature point set unchanged, based on the adjusted height ratio position parameter, the target height value of the internal mesh vertex in the axial direction of the middle section of the transgingival contour surface is determined. The target height value is calculated based on the relative positional relationship between the height of the internal mesh vertex and the height of the bottom boundary feature point set and the ratio coefficient indicated by the adjusted height ratio position parameter. Based on the fine-tuning radius deformation parameter, the target radial coordinate value of the inner mesh vertex in the radial direction is determined. The target radial coordinate value is calculated by weighting the original radial coordinate value of the inner mesh vertex, the deformation indicated by the fine-tuning radius deformation parameter, and the relative distance between the inner mesh vertex and the bottom boundary feature point set and the top boundary feature point set. The internal mesh vertices are moved to the positions corresponding to the target height value and the target radial coordinate value to generate a fine-tuned transgingival contour surface; Step S4 includes: generating an abutment preparation body based on the finely adjusted transgingival contour surface and the initial crown mesh.
5. The intelligent generation method for personalized denture implant abutments as described in claim 1, characterized in that, Step S41 specifically includes: If the current tooth position is a molar, the three-dimensional convex hull of the initial crown mesh is calculated. Among the upward-facing convex hull patches, candidate convex hull patches whose angle between the normal vector of the convex hull patch and the preset reference direction is less than a preset angle threshold are retrieved. The target convex hull patch with the largest surface area is selected from the candidate convex hull patches, and the unit positive normal vector of the target convex hull patch is determined as the path of insertion. If the current tooth position is a canine or anterior tooth, calculate the average coordinates of all grid vertices of the initial crown mesh. Establish a three-dimensional spatial coordinate system with the point corresponding to the average coordinates as the origin. Calculate the global centroid of the initial crown mesh, which is the weighted average of all grid vertices, with the weight being the area of each grid facet. Determine the vector from the origin to the global centroid as the initial axis. Extract the vertex clusters whose projection height on the initial crown mesh along the initial axis is within a preset height ratio range. Calculate the local centroid of the vertex cluster and determine the vector from the origin to the local centroid as the final placement path direction.
6. The intelligent generation method for personalized denture implant abutments as described in claim 1, characterized in that, Step S45 includes the generation of the base side, specifically including: Harmonic fitting is performed on the axial height of the intersection of the probe line and the initial crown grid to extract the overall height trend of the circumferential variation of the top boundary ring along the transgingival contour surface. The local drop positions on the fitted curve are adaptively corrected, and the fitted fluctuations are scaled to a preset range and then superimposed on the safe design height to obtain the sidewall top boundary ring. The local drop position refers to the local point that has an abnormal downward shift relative to the adjacent circumferential position or the overall low-frequency trend. Starting from the top boundary ring of the transgingival contour surface, multiple sidewall levels are divided along the placement path to the top boundary ring of the sidewall to generate the abutment side.
7. The intelligent generation method for personalized denture implant abutments as described in claim 6, characterized in that, Step S45 includes the generation step of the top surface of the base, specifically including: Using the top boundary ring of the sidewall as the outer boundary of the top surface, multiple ring-shaped sampling points are established from the outside to the inside within the area enclosed by the outer boundary of the top surface. Each sampling point is projected onto the occlusal surface of the initial crown grid along the path of placement to obtain the crown projection height value of each sampling point. For local sampling points that fail to be projected, periodic interpolation compensation is performed based on the crown projection height values of adjacent valid sampling points on the same ring. Based on the radial position of each sampling point, a base height of the abutment that smoothly changes from the top boundary of the sidewall to the center is established. Based on the projection height of the outer region of the crown and the projection height of the multiple crown rings near the center, a reference height of the crown that smoothly changes from the periphery to the center is established. Calculate the height difference between the crown projection height value of each sampling point and the corresponding crown reference height to obtain the local undulation characteristics of the crown occlusal surface relative to the overall height trend; retain positive height differences with a larger undulation retention ratio and limit the maximum allowable upward convexity, and retain negative height differences with a smaller undulation retention ratio and limit the maximum allowable downward concavity. Based on the radial position of each sampling point from the outer boundary, the smoothing introduction weights of positive undulation features and negative concavity features are determined respectively, so that the crown morphological features are gradually introduced from the outer boundary to the interior of the top surface; the processed positive undulations and negative undulations are superimposed on the abutment base height according to the corresponding smoothing introduction weights to obtain the final target top surface height of each sampling point. Using the final target top surface height of each sampling point as the fitting data, a robust low-frequency surface fitting algorithm is used to generate a continuous and smooth abutment top surface, so as to preserve the overall gentle undulation trend of the crown occlusal surface and filter out small grooves, local peaks and scanning noise. A smooth transition region is set between the top boundary ring of the sidewall and the fitted top surface of the base. The outer periphery of the smooth transition region maintains the original height of the top boundary ring of the sidewall, and the height and radial slope of the top surface of the base are gradually matched from the outer periphery to the inner periphery, thereby obtaining a smooth top surface of the base that is continuously connected to the side of the base.
8. The method for intelligent generation of personalized denture implant abutments as described in claim 1, characterized in that, After generating the prepared abutment, the method further includes: Extract the screw channel boundary ring located inside the bottom reference ring of the base mesh, and determine the central axis and channel radius of the screw channel boundary ring; Construct a screw channel cylinder based on the central axis of the screw channel boundary ring and the channel radius; Based on the screw channel cylinder, a Boolean subtraction operation is performed on the base preparation body to delete the target mesh facet located inside the screw channel cylinder in the base preparation body, and the cutting boundary is reconstructed at the cutting intersection position to obtain the intermediate base model with the top opening ring. Align the screw channel boundary ring with the top opening ring on the axis so that the central axis of the screw channel boundary ring and the top opening ring coincide and the normal direction of the ring surface are consistent. Between the screw channel boundary ring and the top opening ring, a connecting mesh is generated using a mesh reconstruction algorithm to serve as the inner wall of the screw channel, thus obtaining the final base preparation body.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the intelligent generation method for personalized dental implant abutments as described in any one of claims 1 to 8.
10. An electronic device having a computer program stored thereon, characterized in that, The device includes a processor and a storage medium, wherein a computer program is stored on the storage medium, and the computer program, when executed by the processor, implements the intelligent generation method for personalized dental implant abutments as described in any one of claims 1 to 8.
