CN116563461A

CN116563461A - System for quick simulation tooth alignment based on CBCT and mouth scanning data

Info

Publication number: CN116563461A
Application number: CN202310509299.2A
Authority: CN
Inventors: 刘顺堃; 王洪建; 杨俊�; 田方俊
Original assignee: Bondent Technology Co ltd
Current assignee: Bondent Technology Co ltd
Priority date: 2023-05-08
Filing date: 2023-05-08
Publication date: 2023-08-08

Abstract

The invention relates to a system for rapidly simulating tooth alignment based on CBCT and mouth sweeping data, which automatically extracts a plurality of single tooth data of CBCT and mouth sweeping segmented crowns by utilizing real CT data and mouth sweeping tooth model data, realizes automatic registration fusion of the tooth data extracted by CBCT and mouth sweeping crown data and extraction presentation of CBCT jaw bone data, and finally can automatically realize the effect of rapidly simulating tooth alignment after fusion (crown data with tooth root). The invention not only can improve the efficiency of the tooth correction, but also can reduce the pain and discomfort of patients. Meanwhile, the invention can also reduce the cost of medical institutions, improve the medical efficiency and provide better medical services for patients. Therefore, the invention has wide application prospect and market value.

Description

System for quick simulation tooth alignment based on CBCT and mouth scanning data

Technical Field

The invention relates to the field of orthodontic tooth alignment, in particular to a system for rapidly simulating tooth alignment based on CBCT and oral scanning data.

Background

Orthodontic treatment mainly adjusts coordination among facial bones, teeth and nerves and muscles of the maxillofacial region through various correction devices, namely, adjusts abnormal relations among upper and lower jawbones, between upper and lower teeth, between teeth and jawbones and between nerves and muscles connecting the upper and lower teeth, teeth and jawbones, and finally aims at achieving balance, stability and beauty of the oromandibular system. The correction of the misjaw deformity mainly depends on wearing an appliance in or outside an oral cavity, and proper 'organisms' are applied to teeth, alveolar bones and jawbones to enable the teeth, the alveolar bones and the jawbones to generate physiological movement, so that the misjaw deformity is corrected.

Traditional orthodontics requires taking X-ray films or oral cavity scans for multiple times, and then performing manual design, which is time-consuming and labor-consuming. And in the traditional orthodontic process, patients need to visit for many times, and the treatment time is long.

Disclosure of Invention

The invention aims to solve the technical problems of designing a tooth arrangement system which can quickly generate a three-dimensional tooth model and automatically design through a computer algorithm, realize quick, accurate and damage-free tooth correction and solve the existing technical problems.

In order to solve the technical problems, the system for rapidly simulating tooth alignment based on CBCT and oral scan data comprises the following steps:

step S1: submitting data by a user through a ClearFutrue Web service page, and uploading the data to an OSS server for storage, wherein the data comprises CBCT data and mouth sweeping model data;

step S2: after receiving the data, the clearFutrue Web informs an automatic processing service of the clearWeb, pulls the data set from the OSS to the local, and informs the clearFutrue Web of whether the data set is successfully downloaded or not and informs the clearFutrue Web service of the message;

step S3: the CBCT data and the mouth sweeping model data in the downloaded data set are processed through a CBCT processing module and a mouth sweeping processing module respectively to obtain a root bone STL model and a mouth sweeping model, and the model processing result is notified to ClearFutrue Web service;

step S4: automatically fusing the root bone STL model and the mouth sweeping model by the ClearShape end, establishing a model coordinate system and characteristic points, and notifying a modeling processing result message to clearFutrue Web service;

step S5: generating a new tooth arrangement coordinate system by the ClearDesign end according to the model coordinate system and the feature points, simulating tooth arrangement, generating a tooth arrangement target model, uploading the tooth arrangement target model to an OSS (open service system) for storage, and notifying a clearFutrue Web service of a tooth arrangement processing result;

step S6: the ClearFutrue Web end pulls the tooth arrangement target model from the OSS, and displays the initial data and the target tooth arrangement model simultaneously.

Further, in step S3, when CBCT data is processed, first screening DCM files, and generating root bone STL model data from qualified DCM files by CBCT automatic tooth segmentation algorithm; and the mouth scan data passes through a separate tooth dividing processing module, the redundant model is removed, and a tooth dividing model is established.

Further, in step S3, the CBCT automatic tooth segmentation algorithm flow steps are as follows:

step one: labeling the data of teeth, performing network training, and selecting conventional segmentation models of series such as unet3d, vnet, nnunet and the like by a network;

step two: using NVIDIA TensorRT to convert and accelerate the model, inputting a medical CT image, and using TensorRT to conduct sliding window segmentation reasoning on the model to generate nii data labels;

step three: nii data was converted to grid data using Marching Cubes.

Further, in step S3, the processing of the data of the mouth scan includes the following steps:

step one: converting the acquired data of the mouth scanning model into three-dimensional digital data, and storing the three-dimensional digital data in a mesh data structure, and recording the three-dimensional digital data as JawOri;

step two: converting the mesh data into volume data: filling the inside of the mesh data with points of the volume data, and recording the converted data as Jaw;

step three: dividing Jaw by using a volume data dividing network, dividing each tooth contained in Jaw, namely using U-Net, V-Net, deep LabV3, NNUNet and the like based on deep learning, meshing the surfaces of the divided volume data teeth, storing a mesh data structure, and recording as Ti, wherein i is FDI number;

step four: calculating distances diffmn between each point in Ti and all points in Jaw, 0<m < = len (Ti), 0<n < = len (Jaw), and setting a threshold Th1, if diffmn < Th1, setting the label of the Jawn point as i, and marking all labels as Segn;

step five: optimizing the result in the step four by using a shortest path method, so as to finish tooth segmentation;

step six: classifying all vertexes in the JawOri, performing secondary classification on all vertexes by using a point cloud segmentation classification network, selecting a common point cloud segmentation classification network such as PointNet, DGCNN, KPConv and the like, and marking a classification result as seg1n;

step seven: acquiring data on teeth predicted as keypoints: point= (segn= i) (seg 1 n= 1), and using a clustering algorithm to perform K clustering on non-zero values of the Point, clustering 6 types of teeth, 7 types of teeth, 9 types of teeth, and acquiring central coordinates as feature points, so as to finish tooth feature Point detection.

Further, the step S4 includes the following steps:

step one: registering the root bone with the dental crown grid, including coarse registration and fine registration, specifically comprising the following steps:

method of coarse registration: after the tooth crown is segmented, the tooth position of the tooth crown can be known, the nearest points of every two adjacent tooth crown grids are found to be used as a coarse registration point set A, the tooth root is a similar method, the nearest points of every two adjacent tooth root grids are found to be used as a coarse registration point set B of the tooth root, and the two point sets are registered;

method of fine registration: performing icp fine registration on the two data sets through the positions of the previous coarse registration;

step two: after successful registration, root crown fusion is carried out, and the fusion method is as follows:

finding out boundary point information of the dental crown grid, and constructing a segmentation grid for segmenting a part of the dental root overlapped with the dental crown by utilizing the boundary point information, and segmenting the part of the dental root overlapped with the dental crown;

and splicing the cut tooth root and the boundary of the tooth crown so that the tooth crown and the tooth root are fused.

Further, in step S5, the automatic tooth arrangement method of CBCT and crown data includes the following steps:

step one: according to the medical orthodontic, giving out the dental arch form, and fitting polynomial coefficients by using a mathematical method, wherein the coefficients are parameters of the dental arch and the width of teeth to generate an ideal dental arch curve;

step two: and generating dental arches, and hanging teeth, wherein the step is rough row operation, and when certain collision and clearance exist among the teeth in rough row, the collision clearance is eliminated by using an OBB Tree (Oriented Bounding Box Tree) hierarchical bounding box Tree detection method.

Further, the generating of the ideal dental arch curve in the first step specifically includes the following steps:

step A: according to the dental arch shape given by medicine, obtaining the pixel information of the dental arch by using a pixel point extraction method, and fitting a polynomial coefficient by using a least square method, wherein the coefficient is used as template data of the dental arch and is also a parameter of dental arch fitting; polynomial function y=ax ⁶ +bx ⁴ +cx ² +d, the fitted coefficients a, b, c, d are used as parameters of the dental arch;

and (B) step (B): generating an ideal dental arch curve in a two-dimensional space according to the template data;

step C: then, according to the data of the jaw plane, projecting an ideal dental arch curve into a three-dimensional space;

step D: generating an initial tooth rest point on the dental archwire according to the known tooth width and the FACC center point position;

step E: and taking the intersection point of the data of the hanging points of part of dental positions (such as 13, 16 and 18) of the dental arch and the connecting line of the central points of two first teeth FACC on the middle tangent plane as a control point of the dental arch.

Further, in step D, the position of the single tooth on the dental arch, namely the starting point index and the midpoint index, is obtained by a distance method; and finding out the midpoint of each tooth in the approximate range of the dental arch, wherein the midpoint is the hanging point of the tooth on the dental arch line.

Further, in the second step, the algorithm of the OBB Tree is briefly described as follows: according to the intersection detection method based on the separation axis theorem, the projections are respectively projected to 15 axes according to the separation axis theorem, and the position relation of the OBB Tree bounding box is judged according to different projection intervals.

The invention can quickly generate the three-dimensional tooth model by using CBCT and oral scanning data, and then automatically designs the three-dimensional tooth model by using a computer algorithm, thereby realizing quick, accurate and damage-free tooth correction. The CBCT technology can provide three-dimensional tooth structure information, the oral cavity scanning technology can provide high-precision three-dimensional oral cavity data, and the combination of the two data can more comprehensively know the oral cavity condition of a patient. By integrating the CBCT and the scan data, an accurate digital model can be created that simulates the tooth arrangement in the patient's mouth. From this digital model, the physician can formulate a personalized treatment regimen via Computer Aided Design (CAD) software. Meanwhile, by utilizing the 3D printing technology, the correction appliance meeting the personalized requirements of the patient can be manufactured.

According to the system for rapidly simulating tooth alignment based on CBCT and mouth scan data, disclosed by the invention, by utilizing real CT data and mouth scan dental model data, a plurality of single tooth data of CBCT and mouth scan cut dental crowns are automatically extracted, automatic registration fusion of the tooth data extracted by CBCT and mouth scan dental crown data and extraction presentation of CBCT jaw bone data are realized, and finally, the effect of rapidly simulating tooth alignment after fusion (dental crown data with tooth root) can be automatically realized. The invention not only can improve the efficiency of the tooth correction, but also can reduce the pain and discomfort of patients. Meanwhile, the invention can also reduce the cost of medical institutions, improve the medical efficiency and provide better medical services for patients. Therefore, the invention has wide application prospect and market value.

Drawings

The following description of the embodiments of the invention is further defined by reference to the accompanying drawings.

FIG. 1 is a frame diagram of a CBCT oral scanning rapid tooth arrangement process;

FIG. 2 is a display of root data after segmentation;

FIG. 3 is a diagram of the root mesh data after segmentation;

FIG. 4 is an oral scan model saved in a mesh data structure;

FIG. 5 is an oral scan model converted to volumetric data;

FIG. 6 is a schematic view of tooth surface meshing with a segmentation of volume data;

FIG. 7 is a schematic view of tooth segmentation and feature points;

FIG. 8 is an original crown mesh;

FIG. 9 is an original root grid;

FIG. 10 is a view of a root lattice after cutting;

FIG. 11 is a post-fusion root cap mesh;

FIG. 12 is a flow chart for generating a personalized dental arch;

FIG. 13 is a schematic view of initial arch and tooth position;

fig. 14 is a schematic view of the initial dental situation of a dental extraction case;

FIG. 15 is a schematic view of a target dentition with a 1-mm anchorage of 0 mm;

FIG. 16 is a scan initiation bit (left) and simulation target bit (right);

FIG. 17 is the initial (left) and simulated target (right) bits of CBCT jaw and fused root crown tooth data;

FIG. 18 is a right side view of the oral scan;

FIG. 19 is a right side view of CBCT jaw and fused root crown dental data;

FIG. 20 is a mouth sweeping web page function;

FIG. 21 is a CBCT jaw and fused root crown dental mesh page function.

Detailed Description

With reference to fig. 1, the system for rapidly simulating tooth alignment based on CBCT and oral scan data of the present embodiment includes the following steps:

in step S3, when CBCT data is processed, DCM files are screened first, and qualified DCM files are subjected to a CBCT automatic tooth dividing algorithm to generate root bone STL model data; and the mouth scan data passes through a separate tooth dividing processing module, the redundant model is removed, and a tooth dividing model is established.

In step S3, the CBCT automatic tooth-separating algorithm flow steps are as follows:

step one: labeling the data of teeth, performing network training, wherein the network in the embodiment selects conventional segmentation models of series such as unet3d, vnet, nnunet and the like;

step three: nii data are converted into grid data by using Marching Cubes, as shown in fig. 2 and 3;

in step S3, the processing of the data of the mouth scan includes the following steps:

step one: converting the acquired data of the mouth scan model into three-dimensional digital data, and storing the three-dimensional digital data in a mesh data structure, wherein the data is recorded as JawOri, as shown in figure 4;

step two: converting the mesh data into volume data: filling the inside of the mesh data with points of the volume data, setting the spatial resolution to 0.1 in the embodiment, and recording the converted data as Jaw, as shown in FIG. 5;

step three: dividing the Jaw by using a volume data dividing network, dividing each tooth contained in the Jaw, in the embodiment, using U-Net, V-Net, deep LabV3, NNUNet and the like based on deep learning, gridding the tooth surfaces of the divided volume data, storing a mesh data structure, and recording as Ti, wherein i is FDI number, as shown in fig. 6;

step seven: acquiring data on teeth predicted as keypoints: point= (segn= i) (seg 1 n= 1), and performing K clustering on non-zero values of the Point by using a clustering algorithm, wherein the clustering of teeth is 6 types, the clustering of teeth is 7 types, the clustering of teeth is 9 types, and central coordinates are obtained as feature points, so that tooth feature Point detection is completed, and the feature points are shown in fig. 7;

the step S4 includes the following steps:

the cut tooth root and the tooth crown boundary are spliced, so that the tooth crown and the tooth root are fused, as shown in fig. 8-11, and fig. 8 shows an original tooth crown grid; FIG. 9 shows an original root mesh; FIG. 10 shows the root lattice after cutting; FIG. 11 shows a post-fusion root cap mesh;

in step S5, the automatic tooth arrangement method of CBCT and crown data includes the steps of:

step one: according to the medical orthodontic, giving out the dental arch form, and fitting polynomial coefficients by using a mathematical method, wherein the coefficients are parameters of the dental arch and the width of teeth to generate an ideal dental arch curve; FIG. 12 shows a flow chart for generating a personalized dental arch; FIG. 13 is a schematic view of initial arch and tooth position; fig. 14 is a schematic view showing an initial dental state of a dental extraction case; FIG. 15 is a schematic view showing the state of a target tooth position with an anchorage of 1.0 mm; the first step specifically comprises the following steps:

step D: generating an initial tooth rest point on the dental archwire according to the known tooth width and the FACC center point position; the position of a single tooth on the dental arch, namely a starting point index and a midpoint index, is obtained by a distance method; finding out the midpoint of each tooth in the approximate range of the dental arch, wherein the midpoint is the hanging point of the tooth on the dental arch wire;

Step two: generating dental arches, and hanging teeth, wherein the step is coarse row operation, and when certain collision and clearance exist among the teeth after coarse row operation, the collision clearance is eliminated by using an OBB Tree (Oriented Bounding Box Tree) hierarchical bounding box Tree detection method; in the second step, the algorithm of the OBB Tree is briefly described as follows: according to the intersection detection method based on the separation axis theorem, the projections are respectively projected to 15 axes according to the separation axis theorem, and the position relation of the OBB Tree bounding box is judged according to different projection intervals.

Step S6: the ClearFutrue Web end pulls the tooth arrangement target model from the OSS, and displays the initial data and the target tooth arrangement model simultaneously. The clearfutureweb terminal performs the simulated tooth arrangement quick display of the original position and the target position of the tooth through the digital model based on the oral scan data, as shown in fig. 16, and simultaneously the clearfutureweb terminal supports the CBCT data (jawbone data) and the digital model after the oral scan crown and the CBCT root are fused to perform the simulated tooth arrangement quick display of the original position and the target position of the tooth, as shown in fig. 17.

The clearfutures Web side supports grid, multi-view, zoom-in, zoom-out, pan tool use and combination use. Description of grid function: the user can more conveniently and quickly look at the requirements such as tooth movement amount, rotation amount and the like before and after the simulation alignment through the grid function, and the two windows are kept in linkage; multi-view functional description: the user can more conveniently operate to observe the side view, the top view and the single jaw view of the model, and the two windows are kept linked. FIG. 18 shows the initial scan bit (left) and the simulation target bit (right); FIG. 19 shows a right side view of CBCT jawbone and fused root cap tooth data; FIG. 20 shows a Portal-sweeping web page function; fig. 21 shows CBCT jaw bone and fused root crown tooth mesh page function.

The invention reduces the condition that patients need to visit for many times in the traditional orthodontic process, shortens the treatment time and improves the accuracy and predictability of the orthodontic effect. In addition, due to the wide application of digital models and 3D printing technology, patients do not need to bear uncomfortable feeling of the traditional correction appliances, and the correction effect is more natural and attractive.

In addition, if no CBCT data is in the submitted data, the cutting process of the dental crown is performed and only the dental crown data is automatically aligned.

In the above description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. The foregoing description is only of a preferred embodiment of the invention, which can be practiced in many other ways than as described herein, so that the invention is not limited to the specific implementations disclosed above. While the foregoing disclosure has been described with reference to certain embodiments, it will be understood by those skilled in the art that various changes and modifications may be made and equivalents may be substituted for elements thereof without departing from the scope of the invention. Any simple modification, equivalent variation and modification of the above embodiments according to the technical substance of the present invention without departing from the technical solution of the present invention still falls within the scope of the technical solution of the present invention.

Claims

1. A system for rapidly simulating tooth alignment based on CBCT and oral scan data, characterized in that: the method comprises the following steps:

2. The CBCT and oral scan data based rapid simulated tooth alignment system of claim 1, wherein: in step S3, when CBCT data is processed, DCM files are screened first, and qualified DCM files are subjected to a CBCT automatic tooth dividing algorithm to generate root bone STL model data; and the mouth scan data passes through a separate tooth dividing processing module, the redundant model is removed, and a tooth dividing model is established.

3. The CBCT and oral scan data based rapid simulated tooth alignment system of claim 2, wherein: in step S3, the CBCT automatic tooth-separating algorithm flow steps are as follows:

step one: labeling the data of teeth and performing network training;

step three: nii data was converted to grid data using Marching Cubes.

4. The CBCT and oral scan data based rapid simulated tooth alignment system of claim 2, wherein: in step S3, the processing of the data of the mouth scan includes the following steps:

step three: dividing Jaw by using a volume data dividing network, dividing each tooth contained in Jaw, meshing the surfaces of the divided volume data teeth, and storing a mesh data structure which is marked as Ti;

step six: classifying all vertexes in the JawOri, performing two-class classification on all vertexes by using a point cloud segmentation classification network, and marking a classification result as seg1n;

5. The CBCT and oral scan data based rapid simulated tooth alignment system of claim 1, wherein: the step S4 includes the following steps:

6. The CBCT and oral scan data based rapid simulated tooth alignment system of claim 1, wherein: in step S5, the automatic tooth arrangement method of CBCT and crown data includes the steps of:

step two: generating dental arches and hanging teeth, wherein the step is coarse row operation, and the teeth after coarse row have a certain collision and clearance, so that the collision clearance is eliminated by using an OBB Tree hierarchical bounding box Tree detection method.

7. The CBCT and oral scan data based rapid simulated tooth alignment system of claim 6, wherein: the first step specifically comprises the following steps:

step E: and taking the intersection point of the data of the hanging point of part of the dental arch and the connecting line of the central points of two first teeth FACC on the middle tangent plane as a control point of the dental arch.

8. The CBCT and oral scan data based rapid simulated tooth alignment system of claim 7, wherein: in the step D, the position of a single tooth on the dental arch, namely a starting point index and a midpoint index, is obtained by a distance method; and finding out the midpoint of each tooth in the approximate range of the dental arch, wherein the midpoint is the hanging point of the tooth on the dental arch line.

9. The CBCT and oral scan data based rapid simulated tooth alignment system of claim 6, wherein: in the second step, the OBB Tree is projected to 15 axes respectively according to the separation axis theorem by the intersection detection method based on the separation axis theorem, and the position relation of the OBB Tree bounding box is judged according to different projection intervals.