A dental ablation navigation method, system, and storage medium

Abstract

The application provides a tooth grinding navigation method, system and storage medium, comprising the following steps: registering to obtain a coordinate conversion relationship between any two of a CT three-dimensional coordinate system, a reference plate three-dimensional coordinate system and an optical three-dimensional coordinate system; calibrating a tooth extraction handset; constructing and optimizing an oral cavity anatomical structure segmentation learning model; segmenting oral cavity CT real-time three-dimensional data; selecting a three-dimensional model of a tooth to be treated in a tooth CT three-dimensional model; obtaining a real-time relative position between the tooth to be treated and the tooth extraction handset; and updating the three-dimensional model of the tooth to be treated in real time according to a running path of a pin tip, and eliminating three-dimensional voxels on the three-dimensional model of the tooth to be treated that coincide with the running path of the pin tip. The oral cavity anatomical structure segmentation learning model can provide assistance for analysis based on oral cavity CT real-time three-dimensional data, and real-time poses of a handset tracker and an oral reference plate collected by a binocular navigator are used to realize visual tooth grinding.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, and to a method, system, and storage medium for tooth grinding navigation during complex tooth extractions. Background Technology

[0002] Tooth reshaping is a crucial step in complex tooth extractions, especially the removal of impacted wisdom teeth. Before extraction, specialized tools such as ultrasonic bone cutters are used to remove a portion of the outer layer of the tooth crown, providing a leverage point for successful tooth removal. Specifically, when the tooth to be extracted is deeply impacted below the gum line, the gum is incised before extraction, and a groove of a certain depth is created on the surface of the tooth. A tool is then inserted into the groove, and external force is applied to rotate the tool, breaking the tooth in half. The tooth is then removed, completing the extraction procedure.

[0003] When performing tooth grinding using existing techniques, operators first obtain three-dimensional oral CT data using CT equipment. This provides a general understanding of the relative positions of the affected tooth, dentition, and nerve canal within the patient's mouth. Based on personal experience, the operator then performs the tooth grinding procedure. During this process, on the one hand, the operator needs to carefully analyze the three-dimensional oral CT data, selecting boundaries layer by layer to identify the dentition, nerve canal, etc., within the data. On the other hand, due to the location of the affected tooth, the operator cannot directly observe the real-time relative position of the ultrasonic bone scalpel and the affected tooth, making it difficult to control the progress of the tooth grinding operation.

[0004] The industry urgently needs to propose a new solution to address the problems of low data processing efficiency and inability to perform visual grinding operations during complex tooth extractions. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a tooth grinding navigation method, system and storage medium that can assist in distinguishing tooth CT three-dimensional models, jawbone CT three-dimensional models and nerve tube CT three-dimensional models from real-time oral CT three-dimensional data, and can realize visualized tooth grinding.

[0006] To ensure navigation accuracy and treatment effectiveness.

[0007] The technical solution adopted by this invention to solve the technical problem is as follows:

[0008] A method for navigating tooth grinding includes the following steps:

[0009] S1. When the oral reference plate is fixed in the patient's oral cavity, the CT three-dimensional coordinate system, the reference plate three-dimensional coordinate system, and the optical three-dimensional coordinate system are registered to obtain the coordinate transformation relationship between any two of the three systems. The oral reference plate is provided with reference plate markers and ceramic balls. The CT three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the CT scanning device, the reference plate three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the oral reference plate, and the optical three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the binocular navigation device.

[0010] S2. The extraction handpiece is calibrated to obtain the end three-dimensional coordinate values, tip three-dimensional coordinate values, length three-dimensional coordinate values ​​and axial three-dimensional coordinate values ​​of the bur on the extraction handpiece in the optical three-dimensional coordinate system. The extraction handpiece is equipped with a bur fixing sleeve and a handpiece tracker. The handpiece tracker is equipped with tracker marker points.

[0011] S3. Construct an oral anatomy segmentation learning model, and optimize the oral anatomy segmentation learning model by controlling the input sample data through deep learning;

[0012] S4. Obtain the patient's real-time three-dimensional oral CT data, and control the oral anatomy structure segmentation learning model to segment the real-time three-dimensional oral CT data to obtain the patient's tooth CT three-dimensional model, jawbone CT three-dimensional model and nerve tube CT three-dimensional model.

[0013] S5. Select the three-dimensional model of the affected tooth from the three-dimensional CT model of the tooth to be processed according to the control command;

[0014] S6. Control the binocular navigator to simultaneously collect real-time optical positioning data of tracker marker points and reference board marker points, and analyze the real-time relative position between the tooth to be treated and the extraction handpiece through the analysis of the real-time optical positioning data of tracker marker points and reference board marker points;

[0015] S7. When the operator performs tooth grinding operation based on the real-time relative position between the affected tooth and the extraction handpiece, the binocular navigation system is controlled to obtain the running path of the bur tip of the extraction handpiece. The three-dimensional model of the affected tooth is updated in real time based on the running path of the bur tip, and the three-dimensional voxels on the three-dimensional model of the affected tooth that overlap with the running path of the bur tip are eliminated to obtain the updated three-dimensional model of the affected tooth.

[0016] Compared with existing technologies, the beneficial effects of this technical solution are: the oral anatomical structure segmentation learning model can provide assistance for obtaining three-dimensional models of teeth, jawbone, and nerve canals based on real-time three-dimensional data analysis of oral CT; and the real-time pose of the mobile phone tracker and oral reference plate can be collected based on the binocular navigation device, thereby realizing visualized tooth grinding.

[0017] Furthermore, step S1 specifically includes:

[0018] S101. Based on the structural parameters of the mouth reference plate, obtain the three-dimensional coordinate values ​​of the reference plate marking point and the ceramic ball in the three-dimensional coordinate system of the reference plate;

[0019] S102. With the oral reference plate fixed in the patient's oral cavity, control the CT scanning device to acquire the patient's oral and maxillofacial CBCT scan data and obtain the three-dimensional coordinate values ​​of the ceramic ball in the CT three-dimensional coordinate system.

[0020] S103. Based on the three-dimensional coordinate values ​​of the ceramic ball in the reference plate three-dimensional coordinate system and the CT three-dimensional coordinate system, establish the coordinate transformation relationship between the CT three-dimensional coordinate system and the reference plate three-dimensional coordinate system through the rigid body registration algorithm;

[0021] S104. Control the binocular navigation device to collect real-time optical positioning data of the reference plate markers on the mouth reference plate, and obtain the three-dimensional coordinates of the reference plate in the optical three-dimensional coordinate system through the real-time optical positioning data of the reference plate markers on the mouth reference plate.

[0022] S105. Based on the three-dimensional coordinate values ​​of the reference plate marking points on the mouth reference plate in the three-dimensional coordinate system of the reference plate and in the optical three-dimensional coordinate system, establish the coordinate transformation relationship between the three-dimensional coordinate system of the reference plate and the optical three-dimensional coordinate system.

[0023] S106. Based on the coordinate transformation relationship between the CT three-dimensional coordinate system and the reference plate three-dimensional coordinate system, and the coordinate transformation relationship between the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system, obtain the coordinate transformation relationship between any two of the three-dimensional coordinate system, the reference plate three-dimensional coordinate system, and the optical three-dimensional coordinate system.

[0024] The beneficial effect of adopting the above scheme is that by establishing a coordinate transformation relationship between any two of the three coordinate systems of CT, reference plate, and optical coordinate system through registration, the real-time relative position between the tooth to be treated and the extraction handpiece in the CT three-dimensional coordinate system can be determined by collecting the real-time pose of the handpiece tracker and the oral reference plate in the binocular navigation system during real-time navigation.

[0025] Furthermore, step S2 specifically includes:

[0026] S201. Obtain the three-dimensional coordinate values ​​of the calibration mark point, axial calibration rod and length calibration plate on the calibration device in the calibration three-dimensional coordinate system, wherein the calibration three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the calibration device;

[0027] S202. When the bur fixing sleeve on the tooth extraction handpiece is fitted onto the axial calibration rod of the calibration device, the binocular navigator is controlled to collect the three-dimensional coordinate values ​​of the calibration marker point and the tracker marker point in the optical three-dimensional coordinate system.

[0028] S203. Based on the three-dimensional coordinate values ​​of the calibration mark and the tracker mark in the optical three-dimensional coordinate system when the bur fixing sleeve on the calibration plate is fitted onto the axial calibration rod, and combined with the three-dimensional coordinate values ​​of the calibration mark, the axial calibration rod and the length calibration plate in the calibration three-dimensional coordinate system, obtain the end three-dimensional coordinate value, tip three-dimensional coordinate value and axial three-dimensional coordinate value of the bur on the extraction handpiece in the optical three-dimensional coordinate system;

[0029] S204. When the bur on the extraction handpiece comes into contact with the length calibration plate of the calibration device, control the binocular navigator to collect the three-dimensional coordinate values ​​of the calibration marker point and the tracker marker point in the optical three-dimensional coordinate system;

[0030] S205. Based on the three-dimensional coordinates of the calibration mark and the tracker mark in the optical three-dimensional coordinate system when the bur abuts against the length calibration plate of the calibration device, and combined with the three-dimensional coordinates of the calibration mark, the axial calibration rod and the length calibration plate in the calibration three-dimensional coordinate system, calculate and obtain the three-dimensional coordinates of the length of the bur on the extraction handpiece in the optical three-dimensional coordinate system.

[0031] The beneficial effects of adopting the above scheme are: by obtaining the end three-dimensional coordinate values, tip three-dimensional coordinate values, length three-dimensional coordinate values ​​and axial three-dimensional coordinate values ​​of the bur on the dental handpiece in the optical three-dimensional coordinate system through calibration operation, the real-time position of the end, tip, length and axial direction of the bur can be determined by the binocular navigation device when collecting real-time optical positioning data of the calibration mark point.

[0032] Furthermore, step S3 specifically includes:

[0033] S301. Acquire three-dimensional data of multiple sets of oral CT samples and perform grayscale normalization processing on the three-dimensional data of multiple sets of oral CT samples;

[0034] S302. Obtain the preset effective CT value range, and perform CT value intensity normalization processing on the three-dimensional data of the oral CT sample after grayscale normalization processing according to the effective CT value range;

[0035] S303. The three-dimensional data of the oral CT sample after CT value intensity normalization is labeled layer by layer, and each three-dimensional voxel in the three-dimensional data of the oral CT sample is assigned a category label, including tooth label, jawbone label, nerve tube label and background label;

[0036] S304. Define the combination loss function ,in, This is the total loss coefficient. The Dice loss coefficient, The cross-entropy loss coefficient is... For boundary loss coefficients, These are the parameters of the loss function;

[0037] S305. Divide the 3D data of multiple oral CT samples after assigning category labels into a dataset to obtain a training set, a validation set, and a test set;

[0038] S306. Using the U-Net++ model as the basic network architecture, input the three-dimensional data of oral CT samples in the training set into the U-Net++ model, analyze the probability that each three-dimensional voxel in the oral CBCT basic data belongs to the tooth label, jawbone label and background label, and output the one with the highest probability as the judgment result.

[0039] S307. Compare the judgment result with the tooth labels, jawbone labels, nerve tube labels and background labels obtained by annotating the basic oral CBCT data, calculate the total loss coefficient corresponding to the judgment result by combining the loss function, and update the loss function parameters according to the total loss coefficient;

[0040] S308. Iterate through the segmented training set of the oral cavity model, inputting the updated loss function parameters into the U-Net++ model, until... This yields a segmentation learning model of the oral anatomy structure.

[0041] The beneficial effects of adopting the above scheme are as follows: by unifying the grayscale scale through grayscale normalization, the generalization ability of the model of oral CBCT basic data can be effectively improved, making it easier for operators to annotate multiple sets of oral CBCT basic data; by performing CT value intensity normalization processing according to the effective CT value range, the CT values ​​obtained by different patients and different scanning equipment can be mapped to a standardized and comparable range, thereby highlighting relevant tissues and suppressing irrelevant tissues and noise, and improving the model segmentation accuracy.

[0042] Furthermore, step S7 specifically includes:

[0043] S701. Obtain the running path of the bur tip of the dental handpiece in the optical three-dimensional coordinate system;

[0044] S702. Based on the coordinate transformation relationship between the CT three-dimensional coordinate system and the optical three-dimensional coordinate system obtained through registration, obtain the three-dimensional coordinate values ​​of the bur tip of the extraction handpiece in the CT three-dimensional coordinate system. Given the cutting radius R, the running path of the bur tip of the extraction handpiece in the CT three-dimensional coordinate system is generated;

[0045] S703. Real-time collision detection is performed on the running path of the three-dimensional model of the affected tooth and the tip of the bur in the CT three-dimensional coordinate system. Boolean subtraction is performed based on the collision detection results to eliminate the three-dimensional voxels on the three-dimensional model of the affected tooth that overlap with the running path of the bur tip, and the updated three-dimensional model of the affected tooth is obtained.

[0046] The Boolean subtraction operation includes:

[0047] S7031. Traverse and calculate each three-dimensional voxel in the three-dimensional model of the affected tooth. The distance D from the tip of the needle, where, ;

[0048] S7032. Filter out all three-dimensional voxels whose distance D from the needle tip is less than the cutting radius R along the running path of the needle tip;

[0049] S7033. All three-dimensional voxels whose distance D from the tip of the bur is less than the cutting radius R are reconstructed in three dimensions to obtain the cutting voxel set;

[0050] S7034. Locate all three-dimensional voxels contained in the cutting voxel set on the three-dimensional model of the affected tooth, and assign the value of the cutting voxel set on the three-dimensional model of the affected tooth to 0, to obtain the updated three-dimensional model of the affected tooth.

[0051] The beneficial effects of adopting the above scheme are: after real-time collision detection of the three-dimensional model of the affected tooth and the running path of the bur tip, Boolean subtraction is performed based on the collision detection results to remove all three-dimensional voxels contained in the cutting voxel set. The three-dimensional model of the affected tooth can be updated in real time in the CT three-dimensional coordinate system, which makes it easier for operators to observe and grasp the real-time progress.

[0052] The technical solution adopted by this invention to solve the technical problem is as follows:

[0053] A tooth grinding navigation system includes a registration module for registering a CT three-dimensional coordinate system, a reference plate three-dimensional coordinate system, and an optical three-dimensional coordinate system when an oral reference plate is fixed in the patient's oral cavity. The module obtains the coordinate transformation relationship between any two of the three systems. The oral reference plate is equipped with reference plate markers and ceramic balls. The CT three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the CT scanning device. The reference plate three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the oral reference plate. The optical three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the binocular navigation device.

[0054] The calibration module is used to calibrate the extraction handpiece and obtain the end three-dimensional coordinate values, tip three-dimensional coordinate values, length three-dimensional coordinate values ​​and axial three-dimensional coordinate values ​​of the bur on the extraction handpiece in the optical three-dimensional coordinate system. The extraction handpiece is equipped with a bur fixing sleeve and a handpiece tracker. The handpiece tracker is equipped with tracker marker points.

[0055] The model building module is used to build a learning model for oral anatomy segmentation, and optimizes the learning model for oral anatomy segmentation by controlling the input sample data through deep learning.

[0056] The model analysis module is used to acquire real-time 3D data of the patient's oral CT scan and control the oral anatomy structure segmentation learning model to segment the real-time 3D data of the oral CT scan to obtain the patient's tooth CT 3D model, jawbone CT 3D model and nerve tube CT 3D model.

[0057] The tooth selection module is used to select the three-dimensional model of the tooth to be processed from the CT three-dimensional model of the tooth according to the control command.

[0058] The real-time navigation module is used to control the binocular navigator to simultaneously collect real-time optical positioning data of tracker marker points and reference board marker points. The real-time relative position between the tooth to be treated and the extraction handpiece is obtained by analyzing the real-time optical positioning data of tracker marker points and reference board marker points.

[0059] The grinding model update module is used to control the binocular navigation system to obtain the running path of the bur tip of the extraction handpiece when the operator performs tooth grinding operation based on the real-time relative position between the tooth and the extraction handpiece. The module updates the three-dimensional model of the tooth in real time based on the running path of the bur tip, eliminates the three-dimensional voxels on the three-dimensional model of the tooth that coincide with the running path of the bur tip, and obtains the updated three-dimensional model of the tooth.

[0060] Furthermore, the registration module includes:

[0061] The first parameter acquisition unit is used to acquire the three-dimensional coordinate values ​​of the reference plate marker point and the ceramic ball on the reference plate in the three-dimensional coordinate system of the reference plate, based on the structural parameters of the mouth reference plate.

[0062] The CT scanning unit is used to control the CT scanning device to acquire CBCT scan data of the patient's oral and maxillofacial region while the oral reference plate is fixed in the patient's oral cavity, and to obtain the three-dimensional coordinate values ​​of the ceramic ball in the CT three-dimensional coordinate system.

[0063] The first coordinate transformation unit is used to establish the coordinate transformation relationship between the CT three-dimensional coordinate system and the reference plate three-dimensional coordinate system based on the three-dimensional coordinate values ​​of the ceramic ball in the reference plate three-dimensional coordinate system and the CT three-dimensional coordinate system through the rigid body registration algorithm.

[0064] The first optical positioning unit is used to control the binocular navigator to collect real-time optical positioning data of the reference plate markers on the mouth reference plate, and to obtain the three-dimensional coordinates of the reference plate in the optical three-dimensional coordinate system through the real-time optical positioning data of the reference plate markers on the mouth reference plate.

[0065] The second coordinate transformation unit is used to establish the coordinate transformation relationship between the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system based on the three-dimensional coordinate values ​​of the reference plate marking points on the mouth reference plate in the reference plate three-dimensional coordinate system and in the optical three-dimensional coordinate system.

[0066] The third coordinate transformation unit is used to obtain the coordinate transformation relationship between any two of the three coordinate systems, namely the CT three-dimensional coordinate system, the reference plate three-dimensional coordinate system, and the optical three-dimensional coordinate system, based on the coordinate transformation relationship between the CT three-dimensional coordinate system and the reference plate three-dimensional coordinate system, and the coordinate transformation relationship between the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system.

[0067] Furthermore, the calibration module includes:

[0068] The second parameter acquisition unit is used to acquire the three-dimensional coordinate values ​​of the calibration mark point, axial calibration rod and length calibration plate on the calibration device in the calibration three-dimensional coordinate system, wherein the calibration three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the calibration device.

[0069] The second optical positioning unit is used to control the binocular navigator to collect the three-dimensional coordinate values ​​of the calibration mark point and the tracker mark point in the optical three-dimensional coordinate system when the tooth extraction handpiece bur fixing sleeve is sleeved on the axial calibration rod of the calibration device.

[0070] The first calibration unit is used to obtain the end three-dimensional coordinate value, tip three-dimensional coordinate value, and axial three-dimensional coordinate value of the bur on the extraction handpiece in the optical three-dimensional coordinate system based on the three-dimensional coordinate values ​​of the calibration mark point and the tracker mark point in the optical three-dimensional coordinate system when the bur fixing sleeve on the calibration plate is sleeved on the axial calibration rod. It combines the three-dimensional coordinate values ​​of the calibration mark point, the axial calibration rod, and the length calibration plate in the calibration three-dimensional coordinate system.

[0071] The third optical positioning unit is used to control the binocular navigator to collect the three-dimensional coordinate values ​​of the calibration mark point and the tracker mark point in the optical three-dimensional coordinate system when the bur on the extraction handpiece is in contact with the length calibration plate of the calibration device.

[0072] The second calibration unit is used to calculate the length three-dimensional coordinate value of the bur on the extraction handpiece in the optical three-dimensional coordinate system based on the three-dimensional coordinate values ​​of the calibration mark point and the tracker mark point in the optical three-dimensional coordinate system when the bur abuts against the length calibration plate of the calibration device, combined with the three-dimensional coordinate values ​​of the calibration mark point, the axial calibration rod and the length calibration plate in the calibration three-dimensional coordinate system.

[0073] Furthermore, the model building module includes:

[0074] The grayscale normalization unit is used to acquire three-dimensional data of multiple sets of oral CT samples and perform grayscale normalization processing on the three-dimensional data of multiple sets of oral CT samples.

[0075] The CT intensity normalization unit is used to obtain a preset effective CT value range, and to perform CT value intensity normalization processing on the three-dimensional data of the oral CT sample after grayscale normalization based on the effective CT value range.

[0076] The annotation unit is used to annotate the 3D data of oral CT samples after CT value intensity normalization layer by layer, assigning a category label to each 3D voxel in the oral CT sample 3D data. The category labels include tooth labels, jawbone labels, nerve canal labels, and background labels. The function definition unit is used to define the combined loss function. ,in, This is the total loss coefficient. The Dice loss coefficient, The cross-entropy loss coefficient is... For boundary loss coefficients, These are the parameters of the loss function;

[0077] The dataset partitioning unit is used to partition the 3D data of multiple oral CT samples after assigning category labels into training set, validation set and test set;

[0078] The segmentation training unit is used to input the three-dimensional data of oral CT samples from the training set into the U-Net++ model based on the U-Net++ model, analyze the probability that each three-dimensional voxel in the oral CBCT basic data belongs to the tooth label, jawbone label and background label, and output the one with the highest probability as the judgment result.

[0079] The function update unit is used to compare the judgment result with the tooth labels, jawbone labels, nerve tube labels and background labels obtained by annotating the basic oral CBCT data. It calculates the total loss coefficient corresponding to the judgment result by combining the loss function and updates the loss function parameters according to the total loss coefficient.

[0080] Traversing the training units is used to iterate through the U-Net++ model after segmenting the oral cavity model training set and updating the loss function parameters, until... This yields a segmentation learning model of the oral anatomy structure.

[0081] Furthermore, the grinding model update module includes:

[0082] The path acquisition unit is used to acquire the running path of the bur tip of the dental handpiece in the optical three-dimensional coordinate system.

[0083] The path transformation unit is used to obtain the three-dimensional coordinates of the extraction handpiece bur tip in the CT three-dimensional coordinate system based on the coordinate transformation relationship between the CT three-dimensional coordinate system and the optical three-dimensional coordinate system obtained through registration. Given the cutting radius R, the running path of the bur tip of the extraction handpiece in the CT three-dimensional coordinate system is generated;

[0084] A collision detection unit is used to perform real-time collision detection on the running path of the 3D model of the affected tooth and the tip of the bur in the CT 3D coordinate system. Based on the collision detection results, a Boolean subtraction operation is performed to eliminate 3D voxels on the 3D model of the affected tooth that overlap with the running path of the bur tip, resulting in an updated 3D model of the affected tooth. The collision detection unit includes:

[0085] Distance analysis component is used to traverse and calculate each 3D voxel in the 3D model of the affected tooth. The distance D from the tip of the needle, where, ;

[0086] A voxel screening component is used to screen out all three-dimensional voxels whose distance D from the nib tip is less than the cutting radius R along the running path of the nib tip.

[0087] A three-dimensional reconstruction component is used to reconstruct a set of three-dimensional voxels by combining all three-dimensional voxels whose distance D from the tip of the bur is less than the cutting radius R.

[0088] The grinding model update component is used to locate all three-dimensional voxels contained in the cutting voxel set on the three-dimensional model of the affected tooth, and assign the value of the cutting voxel set on the located three-dimensional model of the affected tooth to 0, so as to obtain the updated three-dimensional model of the affected tooth.

[0089] Correspondingly, a storage medium stores a computer program, the computer program including program instructions, which, when executed by a processor, execute the tooth grinding navigation method as described above. Attached Figure Description

[0090] Figure 1 This is a flowchart of the tooth grinding navigation method of the present invention.

[0091] Figure 2 This is a schematic diagram of the tooth grinding navigation system of the present invention.

[0092] The components represented by each number in the diagram are listed below:

[0093] Registration module 1, calibration module 2, model construction module 3, model analysis module 4, affected tooth selection module 5, real-time navigation module 6, grinding model update module 7. Detailed Implementation

[0094] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to 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.

[0095] In the description of this invention, it should be understood that the terms "center," "upper," "lower," "front," "rear," "left," and "right," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0096] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. When a component is referred to as being "fixed to" or "set on" another element, it can be directly on the other component or there may be an intervening component. When a component is considered to be "connected" to another element, it can be directly connected to the other element or there may be an intervening component. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0097] Tooth reshaping is a crucial step in complex tooth extractions, especially the removal of impacted wisdom teeth. Before extraction, specialized tools such as ultrasonic bone cutters are used to remove a portion of the outer layer of the tooth crown, creating a leverage point for easy removal. When the tooth to be extracted is deeply impacted below the gum line, the gum must be cut before extraction, and a groove of a certain depth is created on the tooth surface. A tool is then inserted into the groove, and external force is applied to rotate the tool, breaking the tooth in half. The tooth is then removed, completing the extraction procedure.

[0098] When performing tooth grinding using existing techniques, operators first acquire three-dimensional oral CT data using CT equipment. This provides a general understanding of the relative positions of the affected tooth, dentition, and nerve canal within the patient's mouth. Based on personal experience, the operator then performs the tooth grinding procedure. During this process, on the one hand, the operator needs to carefully analyze the three-dimensional oral CT data, selecting boundaries layer by layer to identify the dentition, nerve canal, and other structures within the data. On the other hand, due to the location of the affected tooth, the operator cannot directly observe the real-time relative position of the ultrasonic bone scalpel and the tooth, making it difficult to monitor the progress of the tooth grinding operation.

[0099] The industry urgently needs to propose a new solution to address the problems of low data processing efficiency and inability to perform visual grinding operations during complex tooth extractions.

[0100] To address the aforementioned problems, the present invention provides a method, system, and storage medium for tooth grinding navigation, the method, system, and storage medium being based on a tooth grinding navigation device.

[0101] like Figure 1 As shown, in order to solve the above problems, the present invention provides a tooth grinding navigation method, which specifically includes the following steps:

[0102] S1. When the oral reference plate is fixed inside the patient's oral cavity, the CT three-dimensional coordinate system, the reference plate three-dimensional coordinate system, and the optical three-dimensional coordinate system are registered to obtain the coordinate transformation relationship between any two of the three systems. The oral reference plate is equipped with reference plate markers and ceramic balls. The CT three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the CT scanning device, the reference plate three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the oral reference plate, and the optical three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the binocular navigation device. The purpose of step S1 is to perform registration. Through the registration operation, using the ceramic balls and reference plate markers as mediators, the coordinate transformation relationship between any two of the three systems is established, preparing data for real-time positioning in steps S6 and S7.

[0103] S2. Calibrate the extraction handpiece by acquiring the three-dimensional coordinates of the end point, tip point, length, and axial direction of the bur on the handpiece in an optical three-dimensional coordinate system. The handpiece is equipped with a bur retaining sleeve and a handpiece tracker, with tracker markers on the tracker. Step S2 aims to perform calibration. Since the bur is fixed to the handpiece by the bur retaining sleeve, there is a certain assembly error between the sleeve and the bur, and the relative positional relationship between the tracker and the handpiece may change slightly after long-term use. Calibration, in essence, determines the real-time pose of the end point, tip point, length, and axial direction of the handpiece in an optical three-dimensional coordinate system before each use. After calibration, real-time optical positioning data of the tracker is acquired using a binocular navigator. This real-time optical positioning data is a real-time image including the tracker markers, thus obtaining the real-time pose of the end point, tip point, length, and axial direction of the handpiece in an optical three-dimensional coordinate system at that moment.

[0104] S3. Construct an oral anatomy segmentation learning model and optimize it through deep learning using controlled input sample data. Unlike existing technologies, this approach constructs an oral anatomy segmentation learning model and trains it with a large amount of sample data to obtain an optimized model. Based on this model, automatic analysis of real-time 3D oral CT data can be achieved, providing assistance in determining 3D CT models of teeth, jawbone, and nerve canal.

[0105] S4. Acquire real-time 3D CT data of the patient's oral cavity, and control the oral anatomy segmentation learning model to segment the real-time 3D CT data to obtain 3D CT models of the patient's teeth, jawbone, and nerve canal. Within the human oral cavity, teeth, jawbone, nerve canal, and even blood vessels—various types of tissues—exist with different morphologies. Based on a fully trained oral anatomy segmentation learning model, 3D CT models of teeth, jawbone, and nerve canal can be obtained by analyzing real-time 3D CT data. At this point, the operator only needs to verify the analyzed structures, eliminating the tedious layer-by-layer analysis process.

[0106] S5. Select the three-dimensional model of the affected tooth from the CT three-dimensional model of the tooth according to the control command. The affected tooth is the tooth that needs to be extracted. Since its shape is very different from that of a normal tooth, this is very obvious in the CT three-dimensional model of the tooth. In step S5, the operator can input the command to select the three-dimensional model of the affected tooth from the CT three-dimensional model of the tooth, and clarify the object of the tooth grinding operation.

[0107] S6. Control the binocular navigator to simultaneously acquire real-time optical positioning data of the tracker markers and the reference board markers. Analyze the real-time pose of the mobile phone tracker and the oral reference board to obtain the real-time relative position between the tooth to be treated and the extraction handpiece. The binocular navigator can obtain the real-time position of the extraction handpiece through the real-time pose of the mobile phone tracker and the oral reference board. Therefore, during operation, it is only necessary to ensure that the binocular navigator simultaneously acquires real-time optical positioning data of the tracker markers and the reference board markers to directly obtain the real-time relative position between the tooth to be treated and the extraction handpiece.

[0108] S7. When the operator performs tooth grinding operation based on the real-time relative position between the affected tooth and the extraction handpiece, the binocular navigation system is controlled to obtain the running path of the bur tip of the extraction handpiece. The three-dimensional model of the affected tooth is updated in real time based on the running path of the bur tip, and the three-dimensional voxels on the three-dimensional model of the affected tooth that overlap with the running path of the bur tip are eliminated to obtain the updated three-dimensional model of the affected tooth.

[0109] Based on the above technical solution, the oral anatomy structure segmentation learning model can assist in obtaining three-dimensional models of teeth, jawbone, and nerve canals from real-time three-dimensional data analysis of oral CT. Furthermore, based on the real-time pose acquisition of the mobile phone tracker and oral reference plate by the binocular navigation device, visualized tooth grinding can be achieved.

[0110] Preferably, step S1 specifically includes:

[0111] S101. Based on the structural parameters of the mouth reference plate, obtain the three-dimensional coordinate values ​​of the reference plate marker points and the ceramic ball in the three-dimensional coordinate system of the reference plate. In step S101, the structural parameters of the mouth reference plate are known during design and before manufacturing; that is, the mouth reference plate is manufactured according to the known structural parameters. Therefore, the three-dimensional coordinate values ​​of the reference plate marker points and the ceramic ball in the three-dimensional coordinate system of the reference plate can be obtained through the structural parameters of the mouth reference plate.

[0112] S102. With the oral reference plate fixed inside the patient's mouth, control the CT scanning device to acquire CBCT scan data of the patient's oral and maxillofacial region, and obtain the three-dimensional coordinate values ​​of the ceramic ball in the CT three-dimensional coordinate system. By scanning the patient's oral and maxillofacial region with the CT scanning device, CBCT scan data of the oral and maxillofacial region can be acquired. Simultaneously, because the oral reference plate is fixed inside the patient's mouth and the ceramic ball of the oral reference plate has a high density, the three-dimensional coordinate values ​​of the ceramic ball in the CT three-dimensional coordinate system are also captured during the scanning process.

[0113] S103. Based on the three-dimensional coordinates of the ceramic sphere in the reference plate's three-dimensional coordinate system and the CT three-dimensional coordinate system, a coordinate transformation relationship between the CT three-dimensional coordinate system and the reference plate's three-dimensional coordinate system is established using a rigid body registration algorithm. In step S103, the three-dimensional coordinates of the ceramic sphere in the reference plate's three-dimensional coordinate system and the CT three-dimensional coordinate system are known. Using the ceramic sphere as a medium, a coordinate transformation relationship can be established between the CT three-dimensional coordinate system and the reference plate's three-dimensional coordinate system.

[0114] S104. Control the binocular navigation device to collect real-time optical positioning data of the reference board markers on the mouth reference board, and obtain the three-dimensional coordinates of the reference board in the optical three-dimensional coordinate system through the real-time optical positioning data of the reference board markers on the mouth reference board. In step S104, the binocular navigation device can directly obtain the three-dimensional coordinates of the reference board in the optical three-dimensional coordinate system, thereby obtaining the three-dimensional coordinates of the reference board markers in the optical three-dimensional coordinate system.

[0115] S105. Based on the three-dimensional coordinates of the reference plate markers on the mouth reference plate in both the reference plate's three-dimensional coordinate system and the optical three-dimensional coordinate system, establish the coordinate transformation relationship between the reference plate's three-dimensional coordinate system and the optical three-dimensional coordinate system. Since the three-dimensional coordinates of the reference plate markers in the reference plate's three-dimensional coordinate system are known, combined with the three-dimensional coordinates of the reference plate markers in the optical three-dimensional coordinate system obtained in step S104, a coordinate transformation relationship can be established between the reference plate's three-dimensional coordinate system and the optical three-dimensional coordinate system.

[0116] S106. Based on the coordinate transformation relationship between the CT three-dimensional coordinate system and the reference plate three-dimensional coordinate system, and the coordinate transformation relationship between the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system, obtain the coordinate transformation relationship between any two of the three-dimensional coordinate system, the reference plate three-dimensional coordinate system, and the optical three-dimensional coordinate system.

[0117] Based on the above data processing, a coordinate transformation relationship is established between any two of the three coordinate systems: the CT three-dimensional coordinate system, the reference plate three-dimensional coordinate system, and the optical three-dimensional coordinate system. This allows for the determination of the real-time relative position between the tooth to be treated and the extraction handpiece in the CT three-dimensional coordinate system by acquiring the real-time pose of the handpiece tracker and the oral reference plate through a binocular navigator during real-time navigation.

[0118] Preferably, step S2 specifically includes:

[0119] S201. Obtain the three-dimensional coordinate values ​​of the calibration mark point, axial calibration rod, and length calibration plate on the calibration device in the calibration three-dimensional coordinate system, wherein the calibration three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the calibration device. Since the structural parameters are consistent with those of the mouth reference plate, the relative positions of the calibration mark point, axial calibration rod, and length calibration plate on the calibration device are known during design and before manufacturing. In step S201, it is only necessary to directly obtain the above data.

[0120] S202. When the bur fixing sleeve on the extraction handpiece is fitted onto the axial calibration rod of the calibration device, the binocular navigator is controlled to collect the three-dimensional coordinate values ​​of the calibration marker point and the tracker marker point in the optical three-dimensional coordinate system.

[0121] S203. Based on the three-dimensional coordinates of the calibration mark and the tracker mark in the optical three-dimensional coordinate system when the bur fixing sleeve on the calibration plate is fitted onto the axial calibration rod, and combined with the three-dimensional coordinates of the calibration mark, the axial calibration rod, and the length calibration plate in the calibration three-dimensional coordinate system, obtain the end three-dimensional coordinates, tip three-dimensional coordinates, and axial three-dimensional coordinates of the bur on the extraction handpiece in the optical three-dimensional coordinate system.

[0122] S204. When the bur on the extraction handpiece comes into contact with the length calibration plate of the calibration device, the binocular navigator is controlled to collect the three-dimensional coordinate values ​​of the calibration marker and the tracker marker in the optical three-dimensional coordinate system.

[0123] S205. Based on the three-dimensional coordinates of the calibration mark and the tracker mark in the optical three-dimensional coordinate system when the bur abuts against the length calibration plate of the calibration device, and combined with the three-dimensional coordinates of the calibration mark, the axial calibration rod and the length calibration plate in the calibration three-dimensional coordinate system, calculate and obtain the three-dimensional coordinates of the length of the bur on the extraction handpiece in the optical three-dimensional coordinate system.

[0124] Based on the above data processing, the three-dimensional coordinates of the end, tip, length, and axis of the extraction handpiece are obtained through calibration. Thus, the real-time position of the end, tip, length, and axis of the bur can be determined by the binocular navigator when collecting real-time optical positioning data of the calibration markers.

[0125] Preferably, step S3 specifically includes:

[0126] S301. Acquire multiple sets of three-dimensional data of oral CT samples and perform grayscale normalization processing on the multiple sets of three-dimensional data of oral CT samples. In step S301, since the three-dimensional data of oral CT samples is large and comes from different sources, the grayscale value range of the three-dimensional data of oral CT samples from different devices and different patients varies greatly. By performing truncated grayscale normalization, such as retaining the grayscale values ​​of 1%-99th percentile and mapping them to 0-255, the three-dimensional data of oral CT samples are processed to uniform grayscale scale, which can enhance the generalization ability of the model.

[0127] S302. Obtain the preset effective CT value range, and perform CT value intensity normalization processing on the three-dimensional data of the oral CT sample after grayscale normalization according to the effective CT value range. In oral CT images, the fundamental purpose of intensity normalization is to map the CT values ​​(Henness units, HU) obtained from different patients and different scanning devices to a standardized and comparable range, thereby highlighting the tissues of interest (teeth, jawbone) and suppressing irrelevant tissues and noise. In step S302, the effective CT value range specifically refers to the CT value range applied to tooth grinding navigation in this technical solution. During the implementation of this technical solution, the focus is on tissues such as teeth, jawbone, and nerve canal. For example, by performing CT value intensity normalization processing on the three-dimensional data of the oral CT sample after grayscale normalization with an effective CT value range of -500HU to 1500HU, the focus can be on the jawbone and teeth, removing irrelevant information and extreme noise.

[0128] S303. The 3D data of the oral CT samples after CT value intensity normalization is labeled layer by layer. Each 3D voxel in the oral CT sample 3D data is assigned a category label, including tooth label, jawbone label, nerve canal label, and background label. In step S303, the operator labels a large number of oral CT sample 3D data layer by layer, obtaining tooth, jawbone, nerve canal, and background labels in each oral CT sample 3D data. This step is preliminary preparation; the larger the number of voxels, the better the subsequent training effect and the higher the accuracy of the oral anatomy structure segmentation learning model.

[0129] S304. Define the combination loss function ,in, This is the total loss coefficient. The Dice loss coefficient, The cross-entropy loss coefficient is... For boundary loss coefficients, These are the parameters of the loss function.

[0130] S305. Divide the 3D data of multiple oral CT samples after assigning category labels into a dataset to obtain a training set, a validation set, and a test set.

[0131] S306. Using the U-Net++ model as the basic network architecture, input the three-dimensional data of oral CT samples in the training set into the U-Net++ model, analyze the probability that each three-dimensional voxel in the oral CBCT basic data belongs to the tooth label, jawbone label and background label, and output the one with the highest probability as the judgment result.

[0132] S307. Compare the judgment result with the tooth labels, jawbone labels, nerve tube labels and background labels obtained from the basic CBCT data of the oral cavity. Calculate the total loss coefficient corresponding to the judgment result by combining the loss function, and update the loss function parameters according to the total loss coefficient.

[0133] S308. Iterate through the segmented training set of the oral cavity model, inputting the updated loss function parameters into the U-Net++ model, until... This yields a segmentation learning model of the oral anatomy structure.

[0134] Based on the above data processing, grayscale normalization can unify the grayscale scale, effectively improving the model generalization ability of oral CBCT basic data and facilitating operators to annotate multiple sets of oral CBCT basic data. CT value intensity normalization based on the effective CT value range can map CT values ​​obtained from different patients and different scanning devices to a standardized and comparable range, thereby highlighting relevant tissues and suppressing irrelevant tissues and noise, and improving the model segmentation accuracy.

[0135] Preferably, step S7 specifically includes:

[0136] S701. Obtain the running path of the bur tip of the dental extraction handpiece in the optical three-dimensional coordinate system. In step S701, during the operation, the binocular navigator continuously collects real-time optical positioning data of the calibrated marker points over a period of time, thereby obtaining the running path of the bur tip of the dental extraction handpiece in the optical three-dimensional coordinate system during that period of time.

[0137] S702. Based on the coordinate transformation relationship between the CT three-dimensional coordinate system and the optical three-dimensional coordinate system obtained through registration, obtain the three-dimensional coordinate values ​​of the bur tip of the extraction handpiece in the CT three-dimensional coordinate system. Given the cutting radius R, the running path of the bur tip of the extraction handpiece in the CT three-dimensional coordinate system is generated. In step S702, by using the coordinate transformation relationship between the CT three-dimensional coordinate system and the optical three-dimensional coordinate system, and considering the cutting radius while considering the running path of the bur tip in the optical three-dimensional coordinate system, the running path of the bur tip in the CT three-dimensional coordinate system can be calculated.

[0138] S703. Real-time collision detection is performed on the 3D model of the affected tooth and the running path of the bur tip in the CT 3D coordinate system. Based on the collision detection results, a Boolean subtraction operation is performed to eliminate 3D voxels on the 3D model of the affected tooth that overlap with the running path of the bur tip, resulting in an updated 3D model of the affected tooth. In step S703, at the voxel level, this is a voxel-by-voxel erasure operation: for all 3D voxels that "collide" with the sweep volume of the bur tip and are marked as "solid", their state is changed from "solid" to "empty", thus eliminating 3D voxels on the 3D model of the affected tooth that overlap with the running path of the bur tip.

[0139] Preferably, the Boolean subtraction operation includes:

[0140] S7031. Traverse and calculate each three-dimensional voxel in the three-dimensional model of the affected tooth. The distance D from the tip of the needle, where, In step S7031, the distance between each three-dimensional voxel in the three-dimensional model of the affected tooth and the tip of the bur can be directly calculated by traversing the three-dimensional coordinate system.

[0141] S7032. Filter out all three-dimensional voxels whose distance D from the nib tip is less than the cutting radius R along the nib tip's running path. The purpose of step S7032 is to identify all three-dimensional voxels whose distance from the nib tip is less than the cutting radius along the nib tip's running path.

[0142] S7033. All three-dimensional voxels whose distance D from the needle tip is less than the cutting radius R are reconstructed in three dimensions to obtain the cutting voxel set.

[0143] S7034. Locate all three-dimensional voxels contained in the cutting voxel set on the three-dimensional model of the affected tooth, and assign the value of the cutting voxel set on the three-dimensional model of the affected tooth to 0, to obtain the updated three-dimensional model of the affected tooth.

[0144] Based on the above data processing, after real-time collision detection of the three-dimensional model of the affected tooth and the running path of the bur tip, Boolean subtraction is performed according to the collision detection results to remove all three-dimensional voxels contained in the cutting voxel set. The three-dimensional model of the affected tooth can be updated in real time in the CT three-dimensional coordinate system, which is convenient for operators to observe and grasp the real-time progress.

[0145] like Figure 2 As shown, to solve the above problems, the present invention provides a tooth grinding navigation system, including a registration module 1, a calibration module 2, a model construction module 3, a model analysis module 4, a affected tooth selection module 5, a real-time navigation module 6, and a grinding model update module 7, wherein:

[0146] The registration module 1 is used to register the CT three-dimensional coordinate system, the reference plate three-dimensional coordinate system, and the optical three-dimensional coordinate system when the oral reference plate is fixed in the patient's oral cavity, and to obtain the coordinate transformation relationship between any two of the three systems. The oral reference plate is provided with reference plate markers and ceramic balls. The CT three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the CT scanning device, the reference plate three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the oral reference plate, and the optical three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the binocular navigation device.

[0147] The calibration module 2 is used to calibrate the tooth extraction handpiece and obtain the end three-dimensional coordinate values, tip three-dimensional coordinate values, length three-dimensional coordinate values ​​and axial three-dimensional coordinate values ​​of the bur on the tooth extraction handpiece in the optical three-dimensional coordinate system. The tooth extraction handpiece is equipped with a bur fixing sleeve and a handpiece tracker. The handpiece tracker is equipped with tracker marker points.

[0148] Model building module 3 is used to build an oral anatomy segmentation learning model. The oral anatomy segmentation learning model is optimized by controlling the input sample data and performing deep learning.

[0149] Model analysis module 4 is used to acquire real-time three-dimensional data of the patient's oral CT scan and control the oral anatomy structure segmentation learning model to segment the real-time three-dimensional data of the oral CT scan to obtain the patient's tooth CT three-dimensional model, jawbone CT three-dimensional model and nerve tube CT three-dimensional model.

[0150] The tooth selection module 5 is used to select the three-dimensional model of the tooth to be processed from the tooth CT three-dimensional model according to the control command.

[0151] Real-time navigation module 6 is used to control the binocular navigator to simultaneously collect real-time optical positioning data of tracker marker points and reference board marker points, and to analyze the real-time relative position between the tooth to be treated and the extraction handpiece through the analysis of the real-time optical positioning data of tracker marker points and reference board marker points.

[0152] The grinding model update module 7 is used to control the binocular navigation device to obtain the running path of the bur tip of the extraction handpiece when the operator performs tooth grinding operation based on the real-time relative position between the tooth and the extraction handpiece. The module updates the three-dimensional model of the tooth in real time based on the running path of the bur tip, eliminates the three-dimensional voxels on the three-dimensional model of the tooth that overlap with the running path of the bur tip, and obtains the updated three-dimensional model of the tooth.

[0153] Preferably, the registration module includes:

[0154] The first parameter acquisition unit is used to acquire the three-dimensional coordinate values ​​of the reference plate marker point and the ceramic ball on the reference plate in the three-dimensional coordinate system of the reference plate, based on the structural parameters of the mouth reference plate.

[0155] The CT scanning unit is used to control the CT scanning device to acquire CBCT scan data of the patient's oral and maxillofacial region while the oral reference plate is fixed in the patient's oral cavity, and to obtain the three-dimensional coordinate values ​​of the ceramic ball in the CT three-dimensional coordinate system.

[0156] The first coordinate transformation unit is used to establish the coordinate transformation relationship between the CT three-dimensional coordinate system and the reference plate three-dimensional coordinate system based on the three-dimensional coordinate values ​​of the ceramic ball in the reference plate three-dimensional coordinate system and the CT three-dimensional coordinate system through the rigid body registration algorithm.

[0157] The first optical positioning unit is used to control the binocular navigator to collect real-time optical positioning data of the reference plate markers on the mouth reference plate, and to obtain the three-dimensional coordinates of the reference plate in the optical three-dimensional coordinate system through the real-time optical positioning data of the reference plate markers on the mouth reference plate.

[0158] The second coordinate transformation unit is used to establish the coordinate transformation relationship between the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system based on the three-dimensional coordinate values ​​of the reference plate marking points on the mouth reference plate in the reference plate three-dimensional coordinate system and in the optical three-dimensional coordinate system.

[0159] The third coordinate transformation unit is used to obtain the coordinate transformation relationship between any two of the three coordinate systems, namely the CT three-dimensional coordinate system, the reference plate three-dimensional coordinate system, and the optical three-dimensional coordinate system, based on the coordinate transformation relationship between the CT three-dimensional coordinate system and the reference plate three-dimensional coordinate system, and the coordinate transformation relationship between the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system.

[0160] Preferably, the calibration module includes:

[0161] The second parameter acquisition unit is used to acquire the three-dimensional coordinate values ​​of the calibration mark point, axial calibration rod and length calibration plate on the calibration device in the calibration three-dimensional coordinate system, wherein the calibration three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the calibration device.

[0162] The second optical positioning unit is used to control the binocular navigator to collect the three-dimensional coordinate values ​​of the calibration mark point and the tracker mark point in the optical three-dimensional coordinate system when the tooth extraction handpiece bur fixing sleeve is sleeved on the axial calibration rod of the calibration device.

[0163] The first calibration unit is used to obtain the end three-dimensional coordinate value, tip three-dimensional coordinate value, and axial three-dimensional coordinate value of the bur on the extraction handpiece in the optical three-dimensional coordinate system based on the three-dimensional coordinate values ​​of the calibration mark point and the tracker mark point in the optical three-dimensional coordinate system when the bur fixing sleeve on the calibration plate is sleeved on the axial calibration rod. It combines the three-dimensional coordinate values ​​of the calibration mark point, the axial calibration rod, and the length calibration plate in the calibration three-dimensional coordinate system.

[0164] The third optical positioning unit is used to control the binocular navigator to collect the three-dimensional coordinate values ​​of the calibration mark point and the tracker mark point in the optical three-dimensional coordinate system when the bur on the extraction handpiece is in contact with the length calibration plate of the calibration device.

[0165] The second calibration unit is used to calculate the length three-dimensional coordinate value of the bur on the extraction handpiece in the optical three-dimensional coordinate system based on the three-dimensional coordinate values ​​of the calibration mark point and the tracker mark point in the optical three-dimensional coordinate system when the bur abuts against the length calibration plate of the calibration device, combined with the three-dimensional coordinate values ​​of the calibration mark point, the axial calibration rod and the length calibration plate in the calibration three-dimensional coordinate system.

[0166] Preferably, the model building module includes:

[0167] The grayscale normalization unit is used to acquire three-dimensional data of multiple sets of oral CT samples and perform grayscale normalization processing on the three-dimensional data of multiple sets of oral CT samples.

[0168] The CT intensity normalization unit is used to obtain a preset effective CT value range, and to perform CT value intensity normalization processing on the three-dimensional data of the oral CT sample after grayscale normalization based on the effective CT value range.

[0169] The annotation unit is used to annotate the three-dimensional data of the oral CT sample after the CT value intensity normalization process layer by layer, and to assign a category label to each three-dimensional voxel in the three-dimensional data of the oral CT sample. The category labels include tooth label, jawbone label, nerve tube label and background label.

[0170] Function definition unit, used to define the combination loss function ,in, This is the total loss coefficient. The Dice loss coefficient, The cross-entropy loss coefficient is... For boundary loss coefficients, These are the parameters of the loss function;

[0171] The dataset partitioning unit is used to partition the 3D data of multiple oral CT samples after assigning category labels into training set, validation set and test set;

[0172] The segmentation training unit is used to input the three-dimensional data of oral CT samples from the training set into the U-Net++ model based on the U-Net++ model, analyze the probability that each three-dimensional voxel in the oral CBCT basic data belongs to the tooth label, jawbone label and background label, and output the one with the highest probability as the judgment result.

[0173] The function update unit is used to compare the judgment result with the tooth labels, jawbone labels, nerve tube labels and background labels obtained by annotating the basic oral CBCT data. It calculates the total loss coefficient corresponding to the judgment result by combining the loss function and updates the loss function parameters according to the total loss coefficient.

[0174] Traversing the training units is used to iterate through the U-Net++ model after segmenting the oral cavity model training set and updating the loss function parameters, until... This yields a segmentation learning model of the oral anatomy structure.

[0175] Preferably, the grinding model update module includes:

[0176] The path acquisition unit is used to acquire the running path of the bur tip of the dental handpiece in the optical three-dimensional coordinate system.

[0177] The path transformation unit is used to obtain the three-dimensional coordinates of the extraction handpiece bur tip in the CT three-dimensional coordinate system based on the coordinate transformation relationship between the CT three-dimensional coordinate system and the optical three-dimensional coordinate system obtained through registration. Given the cutting radius R, the running path of the bur tip of the extraction handpiece in the CT three-dimensional coordinate system is generated;

[0178] The collision detection unit is used to perform real-time collision detection on the running path of the 3D model of the affected tooth and the tip of the bur in the CT 3D coordinate system. Based on the collision detection results, Boolean subtraction is performed to eliminate the 3D voxels on the 3D model of the affected tooth that overlap with the running path of the bur tip, and to obtain the updated 3D model of the affected tooth.

[0179] The collision detection unit includes:

[0180] Distance analysis component is used to traverse and calculate each 3D voxel in the 3D model of the affected tooth. The distance D from the tip of the needle, where, ;

[0181] A voxel screening component is used to screen out all three-dimensional voxels whose distance D from the nib tip is less than the cutting radius R along the running path of the nib tip.

[0182] A three-dimensional reconstruction component is used to reconstruct a set of three-dimensional voxels by combining all three-dimensional voxels whose distance D from the tip of the bur is less than the cutting radius R.

[0183] The grinding model update component is used to locate all three-dimensional voxels contained in the cutting voxel set on the three-dimensional model of the affected tooth, and assign the value of the cutting voxel set on the located three-dimensional model of the affected tooth to 0, so as to obtain the updated three-dimensional model of the affected tooth.

[0184] To address the aforementioned problems, the present invention provides a storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, execute the tooth grinding navigation method as described above.

[0185] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A dental abrasion navigation system, characterized by, The method comprises the following steps: A registration module is used to register a CT three-dimensional coordinate system, a reference plate three-dimensional coordinate system and an optical three-dimensional coordinate system when a mouth reference plate is fixed in a patient's oral cavity, to obtain a coordinate conversion relationship between any two of the CT three-dimensional coordinate system, the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system, wherein the mouth reference plate is provided with a reference plate marker and a ceramic ball, the CT three-dimensional coordinate system is a three-dimensional coordinate system corresponding to a CT scanning device, the reference plate three-dimensional coordinate system is a three-dimensional coordinate system corresponding to the mouth reference plate, and the optical three-dimensional coordinate system is a three-dimensional coordinate system corresponding to a binocular navigator; A calibration module is used to calibrate an extraction handset, and the calibration device is provided with a calibration marker, an axial calibration rod and a length calibration plate, to obtain end three-dimensional coordinate values, tip three-dimensional coordinate values, length three-dimensional coordinate values and axial three-dimensional coordinate values of the extraction handset in the optical three-dimensional coordinate system, wherein the extraction handset is provided with a burr fixing sleeve and a handset tracker, and the handset tracker is provided with a tracker marker; A model construction module is used to construct an oral anatomical structure segmentation learning model, and the oral anatomical structure segmentation learning model is optimized through deep learning of control input sample data; A model analysis module is used to obtain oral CT real-time three-dimensional data of a patient, and the oral anatomical structure segmentation learning model is used to segment the oral CT real-time three-dimensional data to obtain a tooth CT three-dimensional model, a jawbone CT three-dimensional model and a neural tube CT three-dimensional model of the patient; A tooth selection module is used to select a tooth three-dimensional model to be processed in the tooth CT three-dimensional model according to a control instruction; A real-time navigation module is used to control the binocular navigator to simultaneously collect real-time optical positioning data of the tracker marker and the reference plate marker, and the real-time relative position between the tooth to be processed and the extraction handset is obtained through real-time optical positioning data analysis of the tracker marker and the reference plate marker; An ablation model updating module is used to control the binocular navigator to obtain a running path of a burr tip of the extraction handset when an operator performs tooth ablation operation according to the real-time relative position between the tooth to be processed and the extraction handset, to update the tooth three-dimensional model in real time according to the running path of the burr tip, to eliminate three-dimensional voxels on the tooth three-dimensional model that coincide with the running path of the burr tip, and to obtain an updated tooth three-dimensional model. The ablation model updating module comprises: A path acquisition unit is used to obtain a running path of a burr tip of the extraction handset in the optical three-dimensional coordinate system; A path conversion unit is configured to obtain a three-dimensional coordinate value of the tip of the dental extraction handpiece in the CT three-dimensional coordinate system according to the coordinate conversion relationship between the CT three-dimensional coordinate system and the optical three-dimensional coordinate system obtained through the registration and a cutting radius R, to generate a running path of the tip of the dental extraction handpiece in the CT three-dimensional coordinate system; A collision detection unit is used to perform real-time collision detection on the tooth three-dimensional model and the running path of the burr tip in the CT three-dimensional coordinate system, to perform Boolean subtraction operation according to the collision detection result, to eliminate three-dimensional voxels on the tooth three-dimensional model that coincide with the running path of the burr tip, and to obtain an updated tooth three-dimensional model.

2. The dental resection navigation system of claim 1, wherein, The registration module comprises: A first parameter acquisition unit is used to obtain three-dimensional coordinate values of the reference plate marker and the ceramic ball on the reference plate in the reference plate three-dimensional coordinate system according to structural parameters of the mouth reference plate. The CT scanning unit is configured to control a CT scanning device to collect oral cavity jaw and face CBCT scanning data of a patient in a state where the oral reference plate is fixed in the oral cavity of the patient, and obtain three-dimensional coordinate values of the ceramic ball in a CT three-dimensional coordinate system; The first coordinate conversion unit is configured to establish a coordinate conversion relationship between the CT three-dimensional coordinate system and the reference plate three-dimensional coordinate system by a rigid body registration algorithm according to the three-dimensional coordinate values of the ceramic ball in the reference plate three-dimensional coordinate system and in the CT three-dimensional coordinate system; The first optical positioning unit is configured to control a binocular navigator to collect real-time optical positioning data of the reference plate marker points on the oral reference plate, and obtain three-dimensional coordinate values of the reference plate in an optical three-dimensional coordinate system through the real-time optical positioning data of the reference plate marker points on the oral reference plate; The second coordinate conversion unit is configured to establish a coordinate conversion relationship between the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system according to the three-dimensional coordinate values of the reference plate marker points on the oral reference plate in the reference plate three-dimensional coordinate system and in the optical three-dimensional coordinate system; The third coordinate conversion unit is configured to obtain a coordinate conversion relationship between any two of the CT three-dimensional coordinate system, the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system according to the coordinate conversion relationship between the CT three-dimensional coordinate system and the reference plate three-dimensional coordinate system and the coordinate conversion relationship between the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system; The calibration module comprises: The second parameter acquisition unit is configured to obtain three-dimensional coordinate values of the calibration marker points, the axial calibration rod and the length calibration plate on the calibration device in a calibration three-dimensional coordinate system, the calibration three-dimensional coordinate system being a three-dimensional coordinate system corresponding to the calibration device; The second optical positioning unit is configured to control the binocular navigator to collect three-dimensional coordinate values of the calibration marker points and the tracker marker points in the optical three-dimensional coordinate system when the tooth extraction handset needle fixing sleeve is sleeved on the axial calibration rod of the calibration device; The first calibration unit is configured to obtain end three-dimensional coordinate values, tip three-dimensional coordinate values and axial three-dimensional coordinate values of the needle of the tooth extraction handset in the optical three-dimensional coordinate system according to the three-dimensional coordinate values of the calibration marker points and the tracker marker points in the optical three-dimensional coordinate system when the tooth extraction handset needle fixing sleeve is sleeved on the axial calibration rod of the calibration device, and in combination with the three-dimensional coordinate values of the calibration marker points, the axial calibration rod and the length calibration plate in the calibration three-dimensional coordinate system; The third optical positioning unit is configured to control the binocular navigator to collect three-dimensional coordinate values of the calibration marker points and the tracker marker points in the optical three-dimensional coordinate system when the needle of the tooth extraction handset abuts against the length calibration plate of the calibration device; The second calibration unit is configured to calculate and obtain length three-dimensional coordinate values of the needle of the tooth extraction handset in the optical three-dimensional coordinate system according to the three-dimensional coordinate values of the calibration marker points and the tracker marker points in the optical three-dimensional coordinate system when the needle abuts against the length calibration plate of the calibration device, and in combination with the three-dimensional coordinate values of the calibration marker points, the axial calibration rod and the length calibration plate in the calibration three-dimensional coordinate system.

3. The dental resection navigation system of claim 1, wherein, The collision detection unit comprises: a distance analysis component for traversing each three-dimensional voxel in the three-dimensional model of the tooth under treatment from the tip of the bur, wherein ; The voxel screening component is configured to screen all three-dimensional voxels whose distance D from the needle tip is less than the cutting radius R in the running path of the needle tip. A three-dimensional reorganization component is configured to perform three-dimensional reorganization on all three-dimensional voxels with a distance D from the needle tip less than a cutting radius R to obtain a set of cutting voxels; An abrasion model updating component is configured to locate all three-dimensional voxels contained in the set of cutting voxels on the three-dimensional model of the affected tooth, and assign the set of cutting voxels on the three-dimensional model of the affected tooth as 0 to obtain an updated three-dimensional model of the affected tooth.

4. A storage medium, characterized by The storage medium stores a computer program, and the computer program includes program instructions. When the program instructions are executed by the processor, the processor executes a tooth abrasion navigation method. The tooth abrasion navigation method includes the following steps: S1. When the oral reference plate is fixed in the oral cavity of the patient, the CT three-dimensional coordinate system, the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system are registered to obtain the coordinate conversion relationship between any two of the CT three-dimensional coordinate system, the reference plate three-dimensional coordinate system and the optical three-dimensional coordinate system. The oral reference plate is provided with a reference plate identification point and a ceramic ball. The CT three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the CT scanning device. The reference plate three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the oral reference plate. The optical three-dimensional coordinate system is the three-dimensional coordinate system corresponding to the binocular navigator; S2. The tooth extraction handset is calibrated. The calibration device has a calibration identification point, an axial calibration rod and a length calibration plate. The end three-dimensional coordinate value, the tip three-dimensional coordinate value, the length three-dimensional coordinate value and the axial three-dimensional coordinate value of the needle of the tooth extraction handset in the optical three-dimensional coordinate system are obtained. The tooth extraction handset is provided with a needle fixing sleeve and a handset tracker. The handset tracker is provided with a tracker identification point; S3. An oral anatomy structure segmentation learning model is constructed. The oral anatomy structure segmentation learning model is optimized through deep learning of the control input sample data; S4. The oral CT real-time three-dimensional data of the patient is obtained. The oral anatomy structure segmentation learning model is controlled to segment the oral CT real-time three-dimensional data to obtain the tooth CT three-dimensional model, the jaw CT three-dimensional model and the nerve duct CT three-dimensional model of the patient; S5. The tooth CT three-dimensional model is selected according to the control instruction; S6. The binocular navigator is controlled to simultaneously collect real-time optical positioning data of the tracker identification point and the reference plate identification point. The real-time relative position between the tooth to be treated and the tooth extraction handset is obtained through analysis of the real-time optical positioning data of the tracker identification point and the reference plate identification point; S7. When the operator performs tooth abrasion operation according to the real-time relative position between the tooth to be treated and the tooth extraction handset, the binocular navigator is controlled to obtain the running path of the needle tip of the tooth extraction handset. The tooth three-dimensional model is updated in real time according to the running path of the needle tip to eliminate the three-dimensional voxels on the tooth three-dimensional model that coincide with the running path of the needle tip to obtain an updated tooth three-dimensional model; specifically including: S701. The running path of the needle tip of the tooth extraction handset in the optical three-dimensional coordinate system is obtained. S702. According to the coordinate conversion relationship between the CT three-dimensional coordinate system and the optical three-dimensional coordinate system obtained by registration, the three-dimensional coordinate value of the burr tip of the tooth extraction mobile phone in the CT three-dimensional coordinate system is obtained and the cutting radius R, the running path of the burr tip of the tooth extraction mobile phone in the CT three-dimensional coordinate system is generated; S703. Real-time collision detection is performed on the tooth disease three-dimensional model and the running path of the drill tip end in the CT three-dimensional coordinate system, Boolean subtraction is performed according to the collision detection result, three-dimensional voxels on the tooth disease three-dimensional model that coincide with the running path of the drill tip end are eliminated, and an updated tooth disease three-dimensional model is obtained.

5. The storage medium of claim 4, wherein, The step S1 specifically comprises: S101. Obtain three-dimensional coordinate values of reference board markers on the oral reference board and the ceramic ball in the reference board three-dimensional coordinate system according to structure parameters of the oral reference board; S102. In a state where the oral reference board is fixed in the oral cavity of the patient, control a CT scanning device to collect CBCT scanning data of the oral cavity of the patient, and obtain three-dimensional coordinate values of the ceramic ball in the CT three-dimensional coordinate system; S103. According to the three-dimensional coordinate values of the ceramic ball in the reference board three-dimensional coordinate system and in the CT three-dimensional coordinate system, establish a coordinate conversion relationship between the CT three-dimensional coordinate system and the reference board three-dimensional coordinate system through a rigid body registration algorithm; S104. Control a binocular navigator to collect real-time optical positioning data of the reference board markers on the oral reference board, and obtain three-dimensional coordinate values of the reference board in the optical three-dimensional coordinate system through the real-time optical positioning data of the reference board markers on the oral reference board; S105. According to the three-dimensional coordinate values of the reference board markers on the oral reference board in the reference board three-dimensional coordinate system and in the optical three-dimensional coordinate system, establish a coordinate conversion relationship between the reference board three-dimensional coordinate system and the optical three-dimensional coordinate system; S106. According to the coordinate conversion relationship between the CT three-dimensional coordinate system and the reference board three-dimensional coordinate system and the coordinate conversion relationship between the reference board three-dimensional coordinate system and the optical three-dimensional coordinate system, obtain a coordinate conversion relationship between any two of the CT three-dimensional coordinate system, the reference board three-dimensional coordinate system and the optical three-dimensional coordinate system.

6. The storage medium of claim 4, wherein, The step S2 specifically comprises: S201. Obtain three-dimensional coordinate values of calibration markers, an axial calibration rod and a length calibration plate on a calibration device in a calibration three-dimensional coordinate system, the calibration three-dimensional coordinate system being a three-dimensional coordinate system corresponding to the calibration device; S202. When the drill fixing sleeve on the tooth extraction mobile phone is sleeved on the axial calibration rod of the calibration device, control the binocular navigator to collect three-dimensional coordinate values of the calibration markers and the tracker markers in the optical three-dimensional coordinate system; S203. According to the three-dimensional coordinate values of the calibration markers and the tracker markers in the optical three-dimensional coordinate system when the drill fixing sleeve on the tooth extraction mobile phone is sleeved on the axial calibration rod of the calibration device, and in combination with the three-dimensional coordinate values of the calibration markers, the axial calibration rod and the length calibration plate in the calibration three-dimensional coordinate system, obtain end three-dimensional coordinate values, tip three-dimensional coordinate values and axial three-dimensional coordinate values of the drill on the tooth extraction mobile phone in the optical three-dimensional coordinate system; S204. When the drill on the tooth extraction mobile phone abuts against the length calibration plate of the calibration device, control the binocular navigator to collect three-dimensional coordinate values of the calibration markers and the tracker markers in the optical three-dimensional coordinate system; S205. According to the three-dimensional coordinate values of the calibration mark point and the tracker mark point in the optical three-dimensional coordinate system when the dental elevator abuts against the length calibration plate of the calibration device, and the three-dimensional coordinate values of the calibration mark point, the axial calibration rod and the length calibration plate in the calibration three-dimensional coordinate system, the three-dimensional coordinate values of the length on the dental elevator of the dental extraction handset in the optical three-dimensional coordinate system are calculated and obtained.

7. The storage medium of claim 4, wherein, The Boolean subtraction operation in the step S703 includes: S7031. Iterating through each three-dimensional voxel in the three-dimensional model of the tooth under consideration distance D from the tip of the stylus, wherein, ; S7032. All three-dimensional voxels with a distance D from the tip of the dental elevator less than the cutting radius R in the running path of the tip of the dental elevator are screened out; S7033. All three-dimensional voxels with a distance D from the tip of the dental elevator less than the cutting radius R are three-dimensionally recombined to obtain a cutting voxel set; S7034. All three-dimensional voxels contained in the cutting voxel set on the tooth three-dimensional model are positioned, and the cutting voxel set on the tooth three-dimensional model is valued as 0 to obtain an updated tooth three-dimensional model.

Images