A method, system, device and storage medium for designing a universal model of a donor tooth in autologous tooth transplantation

Abstract

This invention belongs to the field of oral medicine technology and relates to a method, system, device, and storage medium for designing a universal donor tooth model for autologous tooth transplantation. The method includes: acquiring multiple images of wisdom teeth as raw data; extracting vertex information from the raw data and removing outliers to obtain point cloud data; classifying the point cloud data according to the growth location and root morphology of the wisdom teeth to obtain various types of datasets; registering any two point cloud data sets in the same type of dataset to obtain a reference coordinate system for multi-point cloud registration; during the registration process, removing unregisterable wisdom tooth point cloud data to obtain filtered point cloud data; fusing the registered and filtered point cloud data to obtain a fused point cloud, calculating the local density of the fused point cloud and filtering it, retaining the dense part of the point cloud as a preliminary model; extracting the point cloud boundaries of the preliminary model and then smoothing it to obtain universal donor tooth models of different sizes for autologous tooth transplantation.

Description

Technical Field

[0001] This invention belongs to the field of oral medicine technology, specifically relating to a method, system, device and storage medium for designing a universal donor tooth model in autologous tooth transplantation. Background Technology

[0002] Autologous tooth transplantation is one of the treatment options for single tooth loss. By using the patient's own teeth to restore missing teeth, maintain alveolar bone volume, restore normal periodontal tissue and tooth proprioception in the recipient area, and achieve a certain aesthetic effect, it is gradually being accepted by dental professionals and patients with missing teeth.

[0003] In the process of preparing the alveolar socket for autologous tooth transplantation, teeth or models are needed for trial implantation until the tooth or model can be fully accommodated in the alveolar socket. Using solid natural teeth would inevitably increase the donor tooth's detachment time and damage the healthy periodontal ligament. Therefore, most transplants now use 3D-printed personalized donor tooth models to replace donor teeth for repeated trial implantation, shortening the donor tooth's detachment time, reducing periodontal ligament damage, and improving the survival and success rate of the transplanted tooth. However, the creation of personalized 3D-printed models is characterized by long preparation cycles, high costs, extensive professional knowledge requirements, and a lack of health economics considerations. Summary of the Invention

[0004] To address the above problems, the present invention aims to provide a method, system, device, and storage medium for designing a universal donor tooth model in autologous tooth transplantation, thereby solving the problems of long preparation cycles, high costs, extensive professional knowledge requirements, and lack of health economics in the production of personalized 3D printed models.

[0005] This invention is achieved through the following technical solution:

[0006] This invention discloses a universal model design method for donor teeth in autologous tooth transplantation, comprising the following steps:

[0007] S1. Acquire multiple images of the wisdom teeth as raw data;

[0008] S2. Extract vertex information from the raw data, remove outliers from the vertex information, and obtain point cloud data;

[0009] S3. Based on the growth location and root morphology of the wisdom teeth, the point cloud data obtained in S2 is classified to obtain various types of datasets;

[0010] S4. Register any two point cloud data in the same type of dataset to obtain the reference coordinate system for multi-point cloud registration;

[0011] During the registration process, point cloud data of wisdom teeth that cannot be registered are removed to obtain filtered point cloud data;

[0012] S5. Merge the registered and filtered point cloud data to obtain a fused point cloud, calculate the local density of the fused point cloud and filter it, and retain the dense part of the point cloud as a preliminary model.

[0013] S6. Extract the point cloud boundaries from the preliminary model to obtain a general model of donor teeth in autologous tooth transplantation.

[0014] S7. The surface of the obtained general model for donor teeth in autologous tooth transplantation is smoothed, and then scaled proportionally with different coefficients to obtain general models for donor teeth in autologous tooth transplantation of different sizes.

[0015] Furthermore, in S2, outliers in the vertex information are removed according to the Laida criterion.

[0016] Furthermore, in S3, there are multiple types of datasets, including upper single-rooted wisdom tooth datasets, upper multi-rooted wisdom tooth datasets, lower single-rooted wisdom tooth datasets, and lower multi-rooted wisdom tooth datasets.

[0017] Furthermore, in S4, any two point cloud data points in the same type of dataset are registered to obtain a reference coordinate system for multi-point cloud registration, specifically as follows:

[0018] By introducing scale transformation into the ICP algorithm for registration, the point cloud undergoes both rigid transformation and proportional scaling transformation during the registration process. Through continuous iteration, the degree of point cloud matching tends to the local optimum.

[0019] Furthermore, the error function constructed in the ICP algorithm is as follows:

[0020] E = ∑ i (R·s·p i +tq i ) 2

[0021] Where R is the rotation transformation, t is the translation transformation, s is the scaling transformation; E is the error value; p i and q i For any two point cloud data;

[0022] After iterating through R,t,s, as E→min, p i Transformed point cloud Get as close to q as possible i To achieve registration;

[0023]

[0024] error = min E

[0025] After registering all point clouds, calculate the root mean square error between the registered point clouds:

[0026]

[0027] N represents the number of point cloud data points; i = 1, 2, ..., N; RMSE represents the root mean square error.

[0028] Based on the point cloud q corresponding to the minimum root mean square error aim The spatial coordinate system is the reference coordinate system for multi-point cloud registration, so that all point clouds can find the optimal transformation R, t, s in a unified reference coordinate system.

[0029] Furthermore, S5 specifically refers to:

[0030] Multiple registered and filtered point clouds are fused to obtain a fused point cloud Q. According to the distribution characteristics of the fused point cloud, it presents a hollow structure that gradually becomes denser and then sparser from the outside to the inside.

[0031] Q = {q aim ,q1,q2,…,q aim-1 ,q aim+1 ,…,q n}

[0032] n is the number of point cloud data points after filtering;

[0033] The local density of the fused point cloud is calculated and filtered to retain the denser parts of the point cloud as a preliminary model. Specifically:

[0034] Construct a KD tree based on the point cloud Q, and for any point (x) l ,y l ,z l Based on the constructed KD tree, its local density ρ is calculated. l Let the density threshold be φ, if ρ l If the value is greater than φ, then that point is retained, resulting in the filtered point cloud. The filtered point cloud is denoted as...

[0035] Furthermore, in S6, the point cloud boundaries are extracted, specifically as follows:

[0036] For point clouds any point in The neighborhood relationships are calculated using the K-nearest neighbor algorithm. For each point in the neighborhood, the distances from all points to the nearest neighbor are calculated. The distance d between the planes formed j The normal vector is found by minimizing the sum of squares of the distances, and the tangent plane is determined.

[0037]

[0038] Project all points in the neighborhood onto the tangent plane, and use the angle rule to determine whether a point is a contour point; for a point Choose any point in the neighborhood as the starting point, and calculate the angle between each of the other points and the two points in a clockwise or counterclockwise direction. Let θ be the angle between the two points and the angle between them. Let Ω be the set of the two points.

[0039] Ω=(θ1,θ2,…,θ n …,θ k )

[0040] Then every two adjacent points and points The angle formed is denoted as θ′, and the set formed is denoted as Ω′;

[0041] Ω′=(θ1′,θ2′,…θ′ n-1 …θ k ′); where θ′ n-1 =θ n -θ n-1 ;

[0042] Let the threshold of the contour points be σ, when θ′ max When σ > 0, the point is considered to be... For the contour points;

[0043] Where, θ′ max This represents the maximum value of Ω′.

[0044] This invention also discloses a universal donor tooth model design system for autologous tooth transplantation, comprising:

[0045] The data acquisition module is used to acquire multiple images of the wisdom tooth as raw data.

[0046] The raw data includes vertex information and normal information of the wisdom tooth;

[0047] The point cloud data extraction module is used to extract vertex information from the raw data, remove outliers from the vertex information, and obtain point cloud data.

[0048] The data classification module is used to classify point cloud data based on the growth location and root morphology of the wisdom teeth, resulting in various types of datasets.

[0049] The registration module is used to register any two point cloud data in the same type of dataset to obtain a reference coordinate system for multi-point cloud registration.

[0050] During the registration process, point cloud data of wisdom teeth that cannot be registered are removed to obtain filtered point cloud data;

[0051] The fusion module is used to fuse multiple registered and filtered point cloud data to obtain a fused point cloud, calculate the local density of the fused point cloud and filter it, and retain the dense part of the point cloud as a preliminary model.

[0052] The boundary extraction module is used to extract the point cloud boundaries of the preliminary model, thus obtaining a general model of the donor tooth in autologous tooth transplantation.

[0053] The smoothing module is used to smooth the surface of the general donor tooth model in autologous tooth transplantation, and scales it proportionally with different coefficients to obtain general donor tooth models of different sizes in autologous tooth transplantation.

[0054] The present invention also discloses a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the general model design method for donor teeth in autologous tooth transplantation.

[0055] The present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the general model design method for donor teeth in autologous tooth transplantation.

[0056] Compared with the prior art, the present invention has the following beneficial technical effects:

[0057] This invention discloses a method for designing a universal donor tooth model in autologous tooth transplantation. The method uses an image of a wisdom tooth as the raw data, preprocesses it to remove outliers, and obtains point cloud data. The point cloud data is pre-classified according to the different positions and shapes of the wisdom tooth. All point clouds in each dataset are paired and registered, with the point cloud having the smallest RMSE (Real Mean Squared Error) used as the registration reference, ensuring that all other point clouds find their optimal transformation within the coordinate system of that point cloud. The registered point clouds are then fused to obtain the point cloud distribution under a unified reference, and filtered to obtain a preliminary model. Finally, point cloud boundaries are extracted from the preliminary model to obtain a universal donor tooth model for autologous tooth transplantation.

[0058] After designing a universal model, it can replace the individualized model for alveolar socket preparation in autogenous tooth transplantation. Specifically, the universal model is divided into three sizes according to the crown (large, medium, and small) and two sizes according to the root (single-root and multi-root). The universal model of the donor tooth is made using 3D printing technology with a reusable sterilizable cobalt-chromium alloy. When surgery is required clinically, the patient's donor tooth imaging images can be matched with the universal model, and then the appropriate size universal model can be selected and used immediately during alveolar socket preparation in autogenous tooth transplantation. This eliminates the need for individualized models, simplifies the treatment process, shortens the treatment cycle, and has good health economics characteristics. Attached Figure Description

[0059] Figure 1 This is a flowchart of a method for designing a universal donor tooth model in autologous tooth transplantation according to the present invention.

[0060] Figure 2A visualization of the raw wisdom tooth data;

[0061] Figure 3 This is an illustration of outliers in the original data;

[0062] Figure 4 Schematic diagrams of the pre-classification results for four different types of wisdom teeth;

[0063] Figure 5 Comparison images of point cloud pairwise registration before and after;

[0064] Figure 6 A schematic diagram illustrating the principle of multi-point cloud registration;

[0065] Figure 7 To integrate the distribution characteristics of point clouds, a hollow structure is presented that gradually becomes denser and then sparser from the outside to the inside.

[0066] Figure 8 This is a general model of donor teeth in autologous tooth transplantation obtained after fusion filtering;

[0067] Figure 9 These are the large, medium, and small models obtained after smoothing with different coefficients;

[0068] Figure 10 This is a schematic diagram of a universal donor tooth model design system for autologous tooth transplantation according to the present invention. Detailed Implementation

[0069] To make the objectives, technical solutions, and advantages of the present invention clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention; that is, the described embodiments are only a part of the embodiments of the present invention, and not all of them.

[0070] The components described and illustrated in the accompanying drawings and embodiments of this invention can be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of the invention provided in the following drawings is not intended to limit the scope of the claimed invention, but merely to illustrate one selected embodiment of the invention. All other embodiments obtained by those skilled in the art based on the accompanying drawings and embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0071] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0072] Based on the vertex information and physiological characteristics of the original wisdom tooth data, this invention designs a general three-dimensional wisdom tooth model to replace the individualized model for alveolar socket preparation in autologous tooth transplantation through data preprocessing, pre-classification, multi-point cloud registration, fused point cloud filtering, and point cloud boundary extraction.

[0073] like Figure 1 As shown, this invention discloses a universal model design method for donor teeth in autologous tooth transplantation, comprising the following steps:

[0074] S1. Obtain an image of the wisdom tooth as raw data;

[0075] Specifically, DICOM format data containing wisdom tooth images and related information is obtained based on medical imaging. This data is then converted to STL format for storage and is treated as raw data. This raw data includes vertex and normal information, and its visualization is as follows: Figure 2 As shown.

[0076] Vertex information can determine the shape and outline of an object. By using the coordinates of these vertices, the external shape of the wisdom tooth can be delineated.

[0077] Normal information helps determine the lighting effects and visual appearance of an object's surface during rendering or simulation. For example, when visualizing a wisdom tooth image in 3D, normal information can make the light reflection on the wisdom tooth's surface look more realistic and natural.

[0078] S2, Data Preprocessing

[0079] The vertex information is extracted from the raw data to obtain point cloud data;

[0080] The obtained point cloud data contains noisy points and sparse outliers, such as Figure 3 As shown, outliers in the original data are removed according to the Laida criterion.

[0081] The Laida criterion is a statistical method for identifying and removing outliers (also known as gross errors or bad values) in data. Also called the standard deviation method or 3σ method, it calculates the arithmetic mean and standard deviation of a set of data, and defines a confidence interval based on fluctuations of three standard deviations above and below the mean. Data exceeding this interval are considered outliers. For data that approximates a normal distribution and is sufficiently large, this method can effectively find outliers. For all points in a point cloud {P1, P2, ..., P...} i …,P n For each point, perform statistical analysis on its neighborhood, calculate the average distance d from that point to all its neighbors, and then calculate the arithmetic mean of the distances. And the standard deviation σ, for a point P i The average distance d to all neighboring points i If the following conditions are met:

[0082]

[0083] If it contains gross errors, it is considered an outlier and should be removed.

[0084] S3, Wisdom Tooth Pre-classification

[0085] During the actual formation of wisdom teeth, their morphology varies greatly depending on whether they are single-rooted, multi-rooted, and located. It is unreasonable to use a uniform model for wisdom teeth with such great differences. Therefore, we pre-classify them according to their location and root morphology to obtain a dataset.

[0086] Generally, there are four types: upper single-rooted wisdom tooth dataset, upper multi-rooted wisdom tooth dataset, lower single-rooted wisdom tooth dataset, and lower multi-rooted wisdom tooth dataset.

[0087] Specifically, based on the location of wisdom teeth growth, wisdom teeth located in the maxillary positions 18 and 28 are distinguished from those located in the mandibular positions 38 and 48; based on the morphology of the root, wisdom teeth are distinguished as single-rooted and multi-rooted, such as... Figure 4 As shown.

[0088] After pre-classification, the point cloud data was divided into four datasets: single-rooted wisdom teeth datasets for 18 and 28 tooth positions, multi-rooted wisdom teeth datasets for 18 and 28 tooth positions, single-rooted wisdom teeth datasets for 38 and 48 tooth positions, and multi-rooted wisdom teeth datasets for 38 and 48 tooth positions.

[0089] This invention obtained wisdom tooth data from 1003 patients, including 269 single-rooted wisdom teeth in positions 18 and 28, 215 multi-rooted wisdom teeth in positions 18 and 28, 169 single-rooted wisdom teeth in positions 38 and 48, and 350 multi-rooted wisdom teeth in positions 38 and 48. The data was saved in STL format and contains a lot of information; only a portion of the data is shown in the table below for illustration.

[0090]

[0091]

[0092]

[0093] S4, Multi-point Cloud Registration

[0094] In order to obtain a universal model, this invention needs to be universal. However, the obtained tooth data comes from different individuals and is unique in size, position, and orientation (for example, some people's wisdom teeth grow horizontally or even backward). Therefore, these teeth need to be registered to unify the data into a coordinate system with the same scale and reference.

[0095] The ICP algorithm is an algorithm that finds the optimal coordinate transformation through iterative optimization. Its basic idea is to minimize the error between point clouds so that the two point clouds are aligned in the same coordinate system. The algorithm is based on the nearest point search method and uses an iterative process to gradually approach the optimal solution.

[0096] For any point cloud p in the same dataset i and q i By introducing scaling transformation into the ICP algorithm for registration, the point cloud undergoes both rigid and proportional scaling transformations simultaneously during the registration process. Through continuous iteration, the point cloud fit gradually approaches a local optimum, such as... Figure 5 As shown. The constructed error function is:

[0097] E = ∑ i (R·s·p i +tq i ) 2

[0098] Where R is the rotation transformation, t is the translation transformation, and s is the scaling transformation.

[0099] After iterating through R,t,s, as E→min, p i Transformed point cloud Get as close to q as possible i To achieve registration.

[0100] in,

[0101] After registering all point clouds, calculate the root mean square error between the registered point clouds:

[0102]

[0103] Where error = minE, and N represents the number of point cloud data points in the dataset. Error is the registration error, which is the minimum value of E after continuous learning iterations.

[0104] Based on the point cloud q corresponding to the minimum root mean square error aim The spatial coordinate system is the reference coordinate system for multi-point cloud registration, and the point cloud q aim Using a reference point cloud, the optimal transformation R, t, s of all point clouds is found in a unified reference coordinate system, achieving the transformation with q. aim Multi-point cloud registration based on the baseline. For example... Figure 6 As shown.

[0105] S5, Fusion Point Cloud Filtering

[0106] Because some individual teeth have a significantly different shape from all the other teeth, the error between them and other teeth remains large after registration using the ICP algorithm.

[0107] An error threshold α is set. When the error between the point cloud to be registered and the reference point cloud satisfies error > α, the corresponding tooth in that point cloud is considered an error tooth, and this type of special tooth is discarded. The point cloud p retained after discarding is... jAfter registration to the reference coordinate system, the corresponding point cloud q is obtained. j Where j = 1, 2, ..., n, and n is the number of filtered point cloud data; the registered and filtered point clouds are fused to obtain the fused point cloud Q. According to the distribution characteristics of the fused point cloud, it presents a hollow structure that gradually becomes denser and then sparser from the outside to the inside, such as... Figure 7 As shown.

[0108] Q = {q aim ,q1,q2,…,q aim-1 ,q aim+1 ,…,q n}

[0109] The local density of the fused point cloud is calculated and filtered to retain the denser parts of the point cloud as a preliminary model. Specifically:

[0110] Construct the corresponding KD tree based on the point cloud Q, and for any point (x) l ,y l ,z l Based on the constructed KD tree, its local density ρ is calculated. l Set a density threshold φ, if ρ l If the value is greater than φ, then that point is retained, resulting in the filtered point cloud. The filtered point cloud is denoted as...

[0111] S6, Point Cloud Boundary Extraction

[0112] The fused and filtered point cloud already possesses the relevant features of a general model. Therefore, by extracting the point cloud boundaries, an STL model suitable for 3D printing is generated, resulting in a general model of the donor tooth for autologous tooth transplantation, such as... Figure 8 As shown.

[0113] The point cloud boundary is extracted by analyzing the neighborhood relationship of each point in the point cloud, calculating the curvature and normal vector of each point in the point cloud based on the neighborhood relationship, and extracting the outline of the point cloud by analyzing these features.

[0114] Specifically, for point clouds any point in The neighborhood relationships are calculated using the K-nearest neighbor algorithm. For each point in the neighborhood, the distances from all points to the nearest neighbor are calculated. The distance d between the planes formed j The normal vector is found by minimizing the sum of the squares of the distances, thus determining the tangent plane.

[0115]

[0116] Project all points in the neighborhood onto the tangent plane, and determine whether a point is a contour point using the angle rule. For a point... Choose any point in the neighborhood as the starting point, and calculate the angle between each of the other points and the two points in a clockwise or counterclockwise direction, denoted as θ.

[0117] Ω=(θ1,θ2,…,θ n …,θ k )

[0118] Then every two adjacent points and points The angle formed is denoted as θ′, and the set formed is denoted as Ω′;

[0119] Ω′=(θ1′,θ2′,…θ′ n-1 …θ k ′); where θ′ n-1 =θ n -θ n-1 ;

[0120] Let the threshold of the contour points be σ, when θ′ max When σ > 0, the point is considered to be... For the contour points;

[0121] Where, θ′ max This represents the maximum value of Ω′.

[0122] S7, Smoothing

[0123] The surface of the obtained STL model was smoothed, and then scaled proportionally by coefficients of 0.98, 1.03, and 1.08 respectively to obtain the following results: Figure 9 The large, medium, and small models shown are designed to fit different client groups. When surgery is required clinically, the patient's donor tooth imaging images can be matched with the universal models. Then, the appropriate universal model can be selected and used immediately during alveolar socket preparation in autologous tooth transplantation. This eliminates the need for individualized models, simplifies the diagnostic and treatment process, shortens the treatment cycle, and offers excellent health economics.

[0124] like Figure 10 As shown, this invention also discloses a universal donor tooth model design system for autologous tooth transplantation, comprising:

[0125] The data acquisition module is used to acquire multiple images of the wisdom tooth as raw data.

[0126] The raw data includes vertex information and normal information of the wisdom tooth;

[0127] The point cloud data extraction module is used to extract vertex information from the raw data, remove outliers from the vertex information, and obtain point cloud data.

[0128] The data classification module is used to classify point cloud data based on the growth location and root morphology of the wisdom teeth, resulting in various types of datasets.

[0129] The registration module is used to register any two point cloud data in the same type of dataset to obtain a reference coordinate system for multi-point cloud registration.

[0130] During the registration process, point cloud data of wisdom teeth that cannot be registered are removed to obtain filtered point cloud data;

[0131] The fusion module is used to fuse multiple registered and filtered point cloud data to obtain a fused point cloud, calculate the local density of the fused point cloud and filter it, and retain the dense part of the point cloud as a preliminary model.

[0132] The boundary extraction module is used to extract the point cloud boundaries of the preliminary model, thus obtaining a general model of the donor tooth in autologous tooth transplantation.

[0133] The smoothing module is used to smooth the surface of the general donor tooth model in autologous tooth transplantation, and scales it proportionally with different coefficients to obtain general donor tooth models of different sizes in autologous tooth transplantation.

[0134] The system can also continuously add wisdom tooth model data, making the general model more accurate.

[0135] The donor tooth universal model design method for autologous tooth transplantation of the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0136] The universal donor tooth model design method for autologous tooth transplantation of the present invention, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable storage medium includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. It should be noted that the content contained in the computer-readable medium can be appropriately added or subtracted according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals. The computer storage medium can be any available medium or data storage device that a computer can access, including but not limited to magnetic storage (e.g., floppy disk, hard disk, magnetic tape, magneto-optical disk (MO)), optical storage (e.g., CD, DVD, BD, HVD), and semiconductor storage (e.g., ROM, EPROM, EEPROM, non-volatile memory (NAND FLASH), solid-state drive (SSD)).

[0137] In an exemplary embodiment, a computer device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the method for designing a universal donor tooth model in autologous tooth transplantation. The processor may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for designing a universal donor tooth model in autologous tooth transplantation, characterized in that, Includes the following steps: S1. Acquire multiple images of the wisdom teeth as raw data; S2. Extract vertex information from the raw data, remove outliers from the vertex information, and obtain point cloud data; S3. Based on the growth location and root morphology of the wisdom teeth, the point cloud data obtained in S2 is classified to obtain various types of datasets; S4. Register any two point cloud data in the same type of dataset to obtain the reference coordinate system for multi-point cloud registration; During the registration process, point cloud data of wisdom teeth that cannot be registered are removed to obtain filtered point cloud data; S5. Merge the registered and filtered point cloud data to obtain a fused point cloud, calculate the local density of the fused point cloud and filter it, and retain the dense part of the point cloud as a preliminary model. S6. Extract the point cloud boundaries from the preliminary model to obtain a general model of donor teeth in autologous tooth transplantation. S7. Smooth the surface of the obtained general model of donor tooth in autologous tooth transplantation, and scale it proportionally with different coefficients to obtain general models of donor tooth in autologous tooth transplantation of different sizes. In S3, there are multiple types of datasets, including upper single-rooted wisdom tooth dataset, upper multi-rooted wisdom tooth dataset, lower single-rooted wisdom tooth dataset, and lower multi-rooted wisdom tooth dataset. The universal model is divided into three sizes according to the crown size: large, medium, and small, to adapt to different customer groups.

2. The method for designing a universal donor tooth model in autologous tooth transplantation according to claim 1, characterized in that, In S2, outliers in the vertex information are removed according to the Laida criterion.

3. The method for designing a universal donor tooth model in autologous tooth transplantation according to claim 1, characterized in that, In S4, any two point cloud data points in the same type of dataset are registered to obtain a reference coordinate system for multi-point cloud registration, specifically as follows: By introducing scale transformation into the ICP algorithm for registration, the point cloud undergoes both rigid transformation and proportional scaling transformation during the registration process. Through continuous iteration, the degree of point cloud matching tends to the local optimum.

4. The method for designing a universal donor tooth model in autologous tooth transplantation according to claim 1, characterized in that, The error function constructed in the ICP algorithm is as follows: in, For rotational transformation, For translation transformation, For scaling transformation; This is the error value; and For any two point cloud data; Through continuous iteration ,when hour, Transformed point cloud as close as possible To achieve registration; After registering all point clouds, calculate the root mean square error between the registered point clouds: N represents the number of point cloud data; Represents the root mean square error; Based on the point cloud corresponding to the minimum root mean square error The spatial coordinate system in which it is located serves as the reference coordinate system for multi-point cloud registration, enabling all point clouds to find the optimal transformation. , , In a unified reference coordinate system.

5. The method for designing a universal donor tooth model in autologous tooth transplantation according to claim 1, characterized in that, S5 specifically refers to: The registered and filtered point clouds are then fused to obtain a merged point cloud. Based on the distribution characteristics of the fused point cloud, it exhibits a hollow structure that gradually becomes denser and then sparser from the outside to the inside. The number of point cloud data points after filtering; The local density of the fused point cloud is calculated and filtered to retain the denser parts of the point cloud as a preliminary model. Specifically: Based on point cloud Construct a KD-tree, and for any point in it... The local density is calculated based on the constructed KD tree. Let the density threshold be ,like If the point is not found, then retain that point to obtain the filtered point cloud. The filtered point cloud is denoted as... .

6. The method for designing a universal donor tooth model in autologous tooth transplantation according to claim 5, characterized in that, In S6, the point cloud boundaries are extracted, specifically as follows: For point clouds any point in The neighborhood relationship is calculated using the K-nearest neighbor algorithm. For each point in the neighborhood, the distances from all points to the nearest neighbor are calculated. Distance of the planes formed The normal vector is found by minimizing the sum of squares of the distances, and the tangent plane is determined. Project all points in the neighborhood onto the tangent plane, and use the angle rule to determine whether a point is a contour point; for a point Choose any point in the neighborhood as the starting point, and calculate the angle between each of the remaining points and these two points in a clockwise or counterclockwise direction, denoted as . The set formed is denoted as ; Then every two adjacent points and points The angle of composition is denoted as The set formed is denoted as ; ;in ; Let the threshold of the contour points be ,when At that time, it was considered that the point For the contour points; in, represent The maximum value.

7. A system for designing a universal donor tooth model for autogenous tooth transplantation, implementing the universal donor tooth model design method as described in any one of claims 1 to 6, characterized in that, include: The data acquisition module is used to acquire multiple images of the wisdom tooth as raw data. The raw data includes vertex information and normal information of the wisdom tooth; The point cloud data extraction module is used to extract vertex information from the raw data, remove outliers from the vertex information, and obtain point cloud data. The data classification module is used to classify point cloud data based on the growth location and root morphology of the wisdom teeth, resulting in various types of datasets. The registration module is used to register any two point cloud data in the same type of dataset to obtain a reference coordinate system for multi-point cloud registration. During the registration process, point cloud data of wisdom teeth that cannot be registered are removed to obtain filtered point cloud data; The fusion module is used to fuse multiple registered and filtered point cloud data to obtain a fused point cloud, calculate the local density of the fused point cloud and filter it, and retain the dense part of the point cloud as a preliminary model. The boundary extraction module is used to extract the point cloud boundaries of the preliminary model, thus obtaining a general model of the donor tooth in autologous tooth transplantation. The smoothing module is used to smooth the surface of the general donor tooth model in autologous tooth transplantation, and scales it proportionally with different coefficients to obtain general donor tooth models of different sizes in autologous tooth transplantation.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the donor tooth general model design method in autologous tooth transplantation as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the general model design method for donor teeth in autologous tooth transplantation as described in any one of claims 1 to 6.

Images