Intelligent surgical planning method and system for autologous tooth transplantation

Abstract

The present invention provides an intelligent surgical planning method and system for autologous tooth transplantation, relating to the field of image processing technology. The method comprises: obtaining a CBCT image of a tooth to be operated on; segmenting each individual tooth in the CBCT image using a 3D UX‑Net network; segmenting the upper and lower jaws in the CBCT image using a SegResNet network; determining various feasible transplantation surgical planning schemes for transplanting an alternative donor tooth to a recipient tooth using principal component analysis based on the segmentation results of the individual teeth and the jaw; and determining the optimal transplantation surgical planning scheme based on the crown overlap rate, root overlap rate, and amount of bone grinding between the recipient and donor teeth in various transplantation surgical planning schemes. The method achieves precise segmentation of individual teeth and the upper and lower jaws; effectively controls the amount of bone grinding and optimizes the matching between the donor and donor teeth, thereby improving the surgical success rate and postoperative comfort.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to an intelligent surgery planning method and system for autologous tooth transplantation. Background Art

[0002] Autogenous tooth transplantation (ATT) is a dental procedure that involves transferring a patient's own teeth from their original location to another missing tooth within the same patient's mouth. This surgical technique is commonly used to treat impacted, malpositioned, or ectopically erupted teeth, restoring oral function and aesthetics by transferring these teeth to an alveolar socket in an alternative extraction site or edentulous area.

[0003] The activity of the donor tooth's periodontal ligament is a key factor affecting the success of the transplant. The activity of periodontal ligament cells will decrease significantly as the tooth is exposed to the outside of the mouth for a longer time. In order to maximize the activity of periodontal ligament cells, the recipient site needs to be shaped quickly and precisely during autologous tooth transplantation surgery to ensure that the donor tooth can be accurately embedded in the new position.

[0004] Currently, in the context of digital dentistry, many studies have used computer-assisted tooth removal (ATT) to assist with the removal of the donor tooth. Case-control studies have shown that this method effectively shortens the time the donor tooth remains in the body, promotes postoperative healing of the soft and hard tissues in the recipient area, and significantly reduces the risk of complications associated with multiple fittings using the donor tooth.

[0005] However, existing technologies still have shortcomings in the accuracy of segmenting single teeth and upper and lower jaws; at the same time, they cannot effectively control the amount of bone grinding and optimize the match between donor teeth and recipient teeth, resulting in a low surgical success rate and patients are prone to postoperative discomfort, which in turn affects their comfort and satisfaction. Summary of the Invention

[0006] In order to solve the technical problems that the existing technology is still insufficient in the accuracy of segmenting single teeth and upper and lower jaws; at the same time, it cannot effectively control the amount of bone grinding and optimize the matching between donor teeth and recipient teeth, resulting in a low surgical success rate and patients are prone to postoperative discomfort, which in turn affects the comfort and satisfaction of patients, the present invention provides an intelligent surgical planning method and system for autologous tooth transplantation.

[0007] The technical solutions provided by the embodiments of the present invention are as follows:

[0008] First aspect:

[0009] An embodiment of the present invention provides an intelligent surgical planning method for autologous tooth transplantation, comprising:

[0010] S1: Acquire CBCT images of the tooth to be operated on;

[0011] S2: Segmenting each individual tooth in the dental CBCT image using a 3D UX-Net network;

[0012] S3: Segmenting the upper and lower jaws in the dental CBCT image using a SegResNet network;

[0013] S4: Based on the segmentation results of the single tooth and the jaw bone, various feasible transplantation surgery plans for the candidate donor tooth to the recipient tooth are determined by principal component analysis;

[0014] S5: Determine the best transplant surgery planning plan based on the crown overlap rate, root overlap rate and bone grinding amount of the recipient tooth and the donor tooth in various transplant surgery planning plans.

[0015] Second aspect:

[0016] An embodiment of the present invention provides an intelligent surgical planning system for autologous tooth transplantation, comprising:

[0017] processor;

[0018] A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method for intelligent surgical planning of autologous tooth transplantation as described in the first aspect is implemented.

[0019] The third aspect:

[0020] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the method for intelligent surgical planning of autologous tooth transplantation as described in the first aspect is implemented.

[0021] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0022] (1) In the present invention, the individual teeth in the dental CBCT images are segmented by the 3D UX-Net network, and the maxillary and mandibular bones in the dental CBCT images are segmented by the SegResNet network, thereby achieving accurate segmentation of individual teeth and the maxillary and mandibular bones.

[0023] (2) In the present invention, based on the segmentation results of a single tooth and the segmentation results of the jaw bone, various feasible transplantation surgery planning schemes for transplanting alternative donor teeth to recipient teeth are determined by principal component analysis. According to the crown overlap rate, root overlap rate and bone grinding amount of the recipient tooth and the donor tooth in various transplantation surgery planning schemes, the optimal transplantation surgery planning scheme is determined, the bone grinding amount is effectively controlled and the matching between the donor tooth and the recipient tooth is optimized, the success rate of the operation and the postoperative comfort are improved, and thus the comfort and satisfaction of the patient are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0025] Figure 1 A schematic diagram of a process flow of an intelligent surgical planning method for autologous tooth transplantation provided by an embodiment of the present invention;

[0026] Figure 2 A schematic diagram of the structure of an intelligent surgical planning system for autologous tooth transplantation provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0028] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as an "exemplary" in the present invention should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of the word "exemplary" is intended to present concepts in a concrete manner. Furthermore, in the embodiments of the present invention, "and / or" can mean both or either of the two.

[0029] In the embodiments of the present invention, the terms "image" and "picture" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same. The terms "of," "corresponding," and "corresponding" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same.

[0030] In the embodiments of the present invention, sometimes a subscript such as W1 may be mistakenly written as a non-subscript form such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0031] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0032] Reference Manual Figure 1 , which shows a flow chart of an intelligent surgical planning method for autologous tooth transplantation provided by an embodiment of the present invention.

[0033] An embodiment of the present invention provides an intelligent surgical planning method for autologous tooth transplantation. The method can be implemented by an intelligent surgical planning method device for autologous tooth transplantation, which can be a terminal or a server. The processing flow of the intelligent surgical planning method for autologous tooth transplantation can include the following steps:

[0034] S1: Obtain a CBCT image of the tooth to be operated on.

[0035] S2: Segment individual teeth in dental CBCT images using the 3D UX-Net network.

[0036] In a possible implementation, S2 specifically includes:

[0037] S201: Perform binary segmentation of teeth and background on the dental CBCT image through the V-Net network to determine the tooth area image.

[0038] Among them, V-Net is a deep learning model for medical image segmentation, especially suitable for 3D image segmentation tasks. It was proposed by Fausto Milletari et al. in 2016 and is designed to handle the automatic segmentation of organs and lesions in medical images.

[0039] In a possible implementation, the loss function in the V-Net network training process is specifically: Dice loss function.

[0040] The Dice loss function is specifically:

[0041]

[0042] Among them, L dice represents the Dice loss function, X t represents the set of gold standard voxels, Y p Represents the set of predicted voxels.

[0043] Specifically, during the regional localization phase, a V-Net network is used for binary segmentation of teeth and background. To reduce computational complexity while ensuring the network learns global information about key areas, a fixed-size random region is cropped from the CBCT image and used as input to the network in this phase. This region then generates a localization box encompassing all teeth. The volume data is then cropped based on the localization box information and used as input for segmented localization.

[0044] In this paper, V-Net is used to perform preliminary tooth and background segmentation, reducing the computational burden of subsequent processing. This can significantly reduce the demand for computing resources and improve overall processing efficiency.

[0045] S202: Based on the tooth region image, the center of mass of each single tooth is predicted through an encoder-decoder network.

[0046] S203: Segment the tooth region image according to the centroid of each single tooth to obtain each single tooth block.

[0047] The Encoder-Decoder Network (ENN) is a neural network architecture widely used in fields such as image processing and natural language processing. It is mainly used to encode input data into a certain feature representation and then decode these feature representations into output data.

[0048] In one possible implementation, the loss function during the encoder-decoder network training process is specifically: L1 loss function.

[0049] The L1 loss function is specifically:

[0050]

[0051] Among them, L1 is the L1 loss function, which represents the sum of the absolute values ​​of the differences between the predicted value and the true value, y i represents the true value of the voxel, represents the predicted value of the voxel, and N represents the total number of voxels.

[0052] Specifically, the output of the segmentation phase consists of a 3D vector and a binary segmentation mask. The 3D vector represents the vector pointing from each voxel in the centroid offset map to the corresponding tooth centroid. The binary mask is used to eliminate background voxels in the offset map. A fast search clustering method is used to locate peaks in the tooth centroid density map, thereby obtaining the tooth centroid.

[0053] In this paper, the centroid is predicted by an encoder-decoder network, and 3DUX-Net is used in subsequent stages to perform finer segmentation, which can gradually improve the accuracy of segmentation. At the same time, the use of a fast search clustering method can efficiently extract the centroid from the density map, reducing computation time.

[0054] S204: Perform semantic segmentation on the single tooth block using the 3D UX-Net network to determine each single tooth in the dental CBCT image.

[0055] 3D UX-Net is a deep learning network specifically designed for 3D medical image segmentation, designed to improve segmentation accuracy and the ability to process complex 3D structures. It combines traditional convolutional neural networks (CNNs) with advanced feature extraction techniques, making it suitable for processing complex 3D medical images.

[0056] In a possible implementation, the loss function in the 3D UX-Net network training process is specifically: a binary cross entropy loss function.

[0057] The binary cross entropy loss function is specifically:

[0058]

[0059] Among them, L bce represents the binary cross entropy loss function, N represents the total number of voxels, y i represents the true value of the voxel, p(y i ) indicates that the prediction belongs to y i The probability value of .

[0060] Specifically, based on the segmentation positioning results, the image and centroid label map are cropped into multiple single tooth blocks as input, and the 3D UX-Net network is used for semantic segmentation. The framework consists of an encoder-decoder, and long skip connection operations are combined to help the network better recover the positioning of spatial details. A maximum pooling layer and three fully connected layers are added at the end of the encoder to identify the FDI number of each patch, and finally the single tooth segmentation result is obtained.

[0061] In this paper, 3D UX-Net combines long skip connections with an encoder-decoder architecture to effectively restore spatial details and provide highly accurate segmentation results. Furthermore, through skip connections, 3D UX-Net can better preserve detailed information, prevent information loss, and improve the quality of segmentation results.

[0062] S3: Segment the maxillary and mandibular bones in dental CBCT images using the SegResNet network.

[0063] SegResNet is a deep learning-based image segmentation model that combines the strengths of convolutional neural networks and residual networks (ResNet) and is specifically designed for image segmentation tasks. It is particularly effective in medical image segmentation, accurately segmenting target areas.

[0064] In a possible implementation, the SegResNet network uses an encoder-decoder convolutional neural network, and the SegResNet network includes: an encoder and a decoder.

[0065] Each ResNet block in the encoder consists of two convolutional layers. Each convolutional layer is normalized using GroupNormalization and a ReLU activation function is applied. The input is directly connected to the output of the convolutional layer via skip connections.

[0066] The decoder consists of 1x1x1 convolutional layers and uses 3D bilinear upsampling.

[0067] The SegResNet network outputs a 3-channel mask of the same size, representing the probability of each voxel belonging to the background, maxilla, and mandible, respectively.

[0068] In this paper, the SegResNet network uses ResNet blocks in the encoder. These blocks contain skip connections, which effectively address the vanishing gradient problem in deep networks. This enables the network to better train and learn complex features, thereby improving segmentation accuracy. Furthermore, by adding the input directly to the output of the convolutional layer, the residual block retains more detailed information, avoiding information loss that can occur in deep networks. This is particularly important for processing medical images with complex details, and can improve the accuracy of maxillary and mandibular segmentation.

[0069] Furthermore, the output of SegResNet is a three-channel mask of the same size as the input image, with each channel representing the probability of background, maxilla, and mandible. This allows each voxel to be accurately classified as background, maxilla, or mandible, improving the granularity of segmentation. Furthermore, providing the classification probability of each voxel facilitates subsequent processing steps, such as calculating the overlap ratio and the amount of bone grinding.

[0070] In a possible implementation, the loss function during the SegResNet network training process is specifically: a hybrid cross entropy loss function.

[0071] The hybrid cross entropy loss function is specifically:

[0072] L seg =L dice +μ1L bce

[0073] Among them, L seg represents the mixed cross entropy loss function, μ1 represents the weight of the balanced loss, L bce represents the binary cross entropy loss function, L dice represents the Dice loss function.

[0074] In the present invention, Dice loss and binary cross entropy loss are combined. This combination can better handle the problem of class imbalance and improve segmentation accuracy.

[0075] Specifically, in the encoder part, each ResNet block contains two convolutional layers. Group Normalization is used after each convolutional layer for normalization, and the ReLU activation function is applied. In order to avoid the problem of gradient disappearance in deep networks, skip connections are added to add the input directly to the output of the convolutional layer to help the network learn the residual between input and output; in the decoder part, the number of features is reduced by 1x1x1 convolution, and 3D bilinear upsampling is used to increase the spatial dimension, gradually restoring the spatial size of the original image. The network outputs a 3-channel mask of the same size as the input, indicating the probability of each voxel belonging to the background, maxilla and mandible respectively.

[0076] In summary, the benefit of using the SegResNet network for maxillary and mandibular segmentation is that it combines residual learning and encoder-decoder structure to accurately restore image details and generate high-resolution segmentation masks, thereby improving segmentation accuracy and fine-grained classification capabilities.

[0077] S4: Based on the segmentation results of the single tooth and the jaw bone, various feasible transplantation surgery planning schemes for transplanting the alternative donor tooth to the recipient tooth are determined through the principal component analysis method.

[0078] Among them, Principal Component Analysis (PCA) is a commonly used data dimensionality reduction technology, which is used to extract the main features in the data and reduce the dimension of the data while retaining the main information of the original data as much as possible.

[0079] In one possible embodiment, the transplant surgery planning scheme includes: a translation matrix for moving the center point of the donor tooth to the center point of the recipient tooth, a rotation matrix for transforming the long axis vector of the donor tooth to the axial vector of the recipient tooth, and a rotation matrix for transforming the initial direction of the donor tooth to the initial direction of the recipient tooth.

[0080] The calculation method of the translation matrix of the donor tooth center point to the recipient tooth center point includes:

[0081] Define the global coordinate system.

[0082] Calculate the translation matrix of the donor tooth center point to the recipient tooth center point in the global coordinate system.

[0083] The calculation method of the rotation matrix of the long axis vector of the donor tooth to the axial vector of the recipient tooth includes:

[0084] Calculate the covariance matrix of the decentralized original sample matrix and perform eigendecomposition.

[0085] The eigenvector corresponding to the maximum eigenvalue is selected as the tooth long axis vector.

[0086] Determine the upper and lower teeth and update the direction of the tooth long axis vector.

[0087] Taking the center point of the tooth as the rotation center, calculate the rotation matrix of the long axis vector of the donor tooth to the axial vector of the recipient tooth.

[0088] The calculation method of the rotation matrix from the initial orientation of the donor tooth to the initial orientation of the recipient tooth specifically includes:

[0089] Fit the dental arch curves of the upper and lower teeth according to the tooth center points of the upper and lower teeth.

[0090] The initial orientation of the donor tooth was calculated based on morphology.

[0091] The initial orientation of the recipient tooth was calculated based on the tangent direction of the dental arch, and the rotation matrix from the initial orientation of the donor tooth to the initial orientation of the recipient tooth was calculated based on the long axis of the tooth.

[0092] In this method, translation and rotation matrices are used to precisely align the center points, long axis vectors, and initial orientations of the donor and recipient teeth, improving transplant accuracy. This ensures that the geometry and orientation of the donor tooth match the recipient tooth, reducing the need for postoperative adjustments and improving surgical outcomes.

[0093] Furthermore, by considering the center point movement, long axis direction and initial direction, the surgical plan is comprehensively optimized to ensure that the donor tooth can be transplanted in the optimal position and direction of the recipient tooth.

[0094] S5: Determine the best transplant surgery plan based on the crown overlap rate, root overlap rate and bone grinding amount of the recipient tooth and donor tooth in various transplant surgery planning plans.

[0095] In a possible implementation, S5 specifically includes:

[0096] S501: Construct optimization function:

[0097] F operation =(1-R c )*w1+(1-R r )*w2+(G b / C)*w3

[0098] Among them, F operation represents the optimization function, R c represents the overlap ratio between the recipient tooth and the donor tooth crown, R r Indicates the root overlap rate, G b represents the amount of bone grinding, C represents a constant, ω1 represents the weight of the crown overlap rate, ω2 represents the weight of the root overlap rate, and ω3 represents the weight of the bone grinding amount.

[0099] S502: Calculate the optimization function value of each transplant surgery planning scheme.

[0100] S503: The transplant surgery planning scheme with the smallest optimization function value is used as the optimal transplant surgery planning scheme.

[0101] In this paper, an optimization function quantifies crown overlap, root overlap, and bone reduction into a comprehensive indicator, providing a systematic approach to evaluating the overall effectiveness of each surgical plan. Furthermore, by selecting the plan with the smallest optimization function value as the optimal plan, the most suitable transplant surgery plan can be objectively selected, thereby improving the success rate and effectiveness of the surgery.

[0102] Furthermore, the optimization process can identify and avoid potential problems, such as excessive bone resection or undesirable crown overlap, improving surgical safety and predictable outcomes.

[0103] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0104] (1) In the present invention, the individual teeth in the dental CBCT images are segmented by the 3D UX-Net network, and the maxillary and mandibular bones in the dental CBCT images are segmented by the SegResNet network, thereby achieving accurate segmentation of individual teeth and the maxillary and mandibular bones.

[0105] (2) In the present invention, based on the segmentation results of a single tooth and the segmentation results of the jaw bone, various feasible transplantation surgery planning schemes for transplanting alternative donor teeth to recipient teeth are determined by principal component analysis. According to the crown overlap rate, root overlap rate and bone grinding amount of the recipient tooth and the donor tooth in various transplantation surgery planning schemes, the optimal transplantation surgery planning scheme is determined, the bone grinding amount is effectively controlled and the matching between the donor tooth and the recipient tooth is optimized, the success rate of the operation and the postoperative comfort are improved, and thus the comfort and satisfaction of the patient are improved.

[0106] Reference Manual Figure 2 , showing a structural schematic diagram of an intelligent surgical planning system for autologous tooth transplantation provided by the present invention.

[0107] The present invention further provides an intelligent surgery planning system 20 for autologous tooth transplantation, which is applied to the above-mentioned intelligent surgery planning method for autologous tooth transplantation, comprising:

[0108] Processor 201;

[0109] The memory 202 stores computer-readable instructions, and when the computer-readable instructions are executed by the processor 201, the method for intelligent surgical planning of autologous tooth transplantation as described in the method embodiment is implemented.

[0110] The autologous tooth transplantation intelligent surgery planning system 20 provided by the present invention can execute the above-mentioned autologous tooth transplantation intelligent surgery planning method and achieve the same or similar technical effects. To avoid repetition, the present invention will not elaborate on it.

[0111] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0112] (1) In the present invention, the individual teeth in the dental CBCT images are segmented by the 3D UX-Net network, and the maxillary and mandibular bones in the dental CBCT images are segmented by the SegResNet network, thereby achieving accurate segmentation of individual teeth and the maxillary and mandibular bones.

[0113] (2) In the present invention, based on the segmentation results of a single tooth and the segmentation results of the jaw bone, various feasible transplantation surgery planning schemes for transplanting alternative donor teeth to recipient teeth are determined by principal component analysis. According to the crown overlap rate, root overlap rate and bone grinding amount of the recipient tooth and the donor tooth in various transplantation surgery planning schemes, the optimal transplantation surgery planning scheme is determined, the bone grinding amount is effectively controlled and the matching between the donor tooth and the recipient tooth is optimized, the success rate of the operation and the postoperative comfort are improved, and thus the comfort and satisfaction of the patient are improved.

[0114] It should be understood that the processor in the embodiments of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0115] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct rambus RAM (DR RAM).

[0116] The above embodiments can be implemented in whole or in part through software, hardware (such as circuits), firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the processes or functions described in accordance with the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired method (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, or magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0117] It should be understood that the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. A and B can be singular or plural. Furthermore, the character " / " as used herein generally indicates an "or" relationship between the associated objects, but it may also indicate an "and / or" relationship. For specific understanding, please refer to the context.

[0118] In this disclosure, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, "at least one of a, b, or c" can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or plural.

[0119] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0120] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0121] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0122] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interface, indirect coupling or communication connection of the device or unit, which can be electrical, mechanical or other forms.

[0123] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0124] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0125] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0126] An embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the intelligent surgical planning method for autologous tooth transplantation as described in the method embodiment.

[0127] The computer-readable storage medium provided by the present invention can implement the steps and effects of the autologous tooth transplantation intelligent surgery planning method of the above method embodiment. To avoid repetition, the present invention will not elaborate on them.

[0128] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0129] (1) In the present invention, the individual teeth in the dental CBCT images are segmented by the 3D UX-Net network, and the maxillary and mandibular bones in the dental CBCT images are segmented by the SegResNet network, thereby achieving accurate segmentation of individual teeth and the maxillary and mandibular bones.

[0130] (2) In the present invention, based on the segmentation results of a single tooth and the segmentation results of the jaw bone, various feasible transplantation surgery planning schemes for transplanting alternative donor teeth to recipient teeth are determined by principal component analysis. According to the crown overlap rate, root overlap rate and bone grinding amount of the recipient tooth and the donor tooth in various transplantation surgery planning schemes, the optimal transplantation surgery planning scheme is determined, the bone grinding amount is effectively controlled and the matching between the donor tooth and the recipient tooth is optimized, the success rate of the operation and the postoperative comfort are improved, and thus the comfort and satisfaction of the patient are improved.

[0131] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

[0132] There are a few points to note:

[0133] (1) The drawings of the embodiments of the present invention only relate to the structures related to the embodiments of the present invention. Other structures may refer to conventional designs.

[0134] (2) For the sake of clarity, the thickness of layers or regions in the drawings used to describe the embodiments of the present invention are exaggerated or reduced, that is, these drawings are not drawn to scale. It is understood that when an element such as a layer, film, region, or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or intervening elements may be present.

[0135] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to form new embodiments.

[0136] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. An intelligent surgical planning method for autologous tooth transplantation, characterized in that: include: S1: Acquire CBCT images of the tooth to be operated on; S2: Segmenting each individual tooth in the dental CBCT image using a 3D UX-Net network; S3: Segmenting the upper and lower jaws in the dental CBCT image using a SegResNet network; S4: Based on the segmentation results of the single tooth and the jaw bone, various feasible transplantation surgery plans for the candidate donor tooth to the recipient tooth are determined by principal component analysis; S5: determining the best transplant surgery plan according to the crown overlap rate, root overlap rate and bone grinding amount of the recipient tooth and the donor tooth in the various transplant surgery plans; Wherein, the S2 specifically includes: S201: performing binary segmentation of teeth and background on the dental CBCT image using a V-Net network to determine a dental region image; S202: Predicting the center of mass of each single tooth through an encoder-decoder network based on the tooth region image; S203: Segmenting the tooth region image according to the centroid of each single tooth to obtain each single tooth block; S204: performing semantic segmentation on the single tooth block using a 3D UX-Net network to determine each single tooth in the dental CBCT image; The transplant surgery planning scheme includes: a translation matrix for moving the center point of the donor tooth to the center point of the recipient tooth, a rotation matrix for transforming the long axis vector of the donor tooth to the axial vector of the recipient tooth, and a rotation matrix for transforming the initial direction of the donor tooth to the initial direction of the recipient tooth; The calculation method of the translation matrix for moving the center point of the donor tooth to the center point of the recipient tooth specifically includes: Define the global coordinate system; Calculate the translation matrix of the donor tooth center point to the recipient tooth center point in the global coordinate system; The calculation method of the rotation matrix of the transformation of the donor tooth long axis vector to the recipient tooth axial vector specifically includes: Calculate the covariance matrix after the original sample matrix is ​​decentralized and perform eigendecomposition; The eigenvector corresponding to the maximum eigenvalue is selected as the tooth long axis vector; Determine the upper and lower teeth and update the direction of the tooth long axis vector; Taking the center point of the tooth as the rotation center, calculate the rotation matrix of the long axis vector of the donor tooth to the axial vector of the recipient tooth; The calculation method of the rotation matrix from the initial orientation of the donor tooth to the initial orientation of the recipient tooth specifically includes: Fit the dental arch curves of the upper and lower teeth according to the tooth center points of the upper and lower teeth; The initial orientation of the donor tooth was calculated based on morphology; The initial orientation of the recipient tooth was calculated based on the tangent direction of the dental arch, and the rotation matrix from the initial orientation of the donor tooth to the initial orientation of the recipient tooth was calculated based on the long axis of the tooth.

2. The intelligent surgical planning method for autologous tooth transplantation according to claim 1, characterized in that: The loss function in the V-Net network training process is specifically: Dice loss function; The Dice loss function is specifically: ； in, represents the Dice loss function, X t represents the set of gold standard voxels, Y p Represents the set of predicted voxels.

3. The intelligent surgical planning method for autologous tooth transplantation according to claim 1, characterized in that: The loss function during the encoder-decoder network training process is specifically: Loss function; described The loss function is specifically: ； Among them, L1 is The loss function represents the sum of the absolute values ​​of the differences between the predicted values ​​and the true values, y i represents the true value of the voxel, represents the predicted value of the voxel, and N represents the total number of voxels.

4. The intelligent surgical planning method for autologous tooth transplantation according to claim 1, characterized in that: The loss function in the 3D UX-Net network training process is specifically: binary cross entropy loss function; The binary cross entropy loss function is specifically: ； in, represents the binary cross entropy loss function, N represents the total number of voxels, y i represents the true value of the voxel, Indicates that the prediction belongs to The probability value of .

5. The intelligent surgical planning method for autologous tooth transplantation according to claim 1, characterized in that: The SegResNet network adopts an encoder-decoder convolutional neural network, and the SegResNet network includes: an encoder and a decoder; Each ResNet block in the encoder includes two convolutional layers; each convolutional layer is normalized using GroupNormalization and a ReLU activation function is applied; the input is directly connected to the output of the convolutional layer via a skip connection; The decoder includes a 1x1x1 convolutional layer, and the decoder adopts 3D bilinear upsampling; The SegResNet network outputs 3-channel masks of the same size, representing the probability of each voxel belonging to the background, maxilla, and mandible respectively.

6. The method for intelligent surgical planning of autologous tooth transplantation according to claim 5, characterized in that: The loss function in the SegResNet network training process is specifically: hybrid cross entropy loss function; The hybrid cross entropy loss function is specifically: ； in, represents the mixed cross entropy loss function, μ1 represents the weight of the balanced loss, represents the binary cross entropy loss function, represents the Dice loss function.

7. The intelligent surgical planning method for autologous tooth transplantation according to claim 1, characterized in that: The S5 specifically includes: S501: Construct optimization function: ； in, represents the optimization function, R c represents the overlap ratio between the recipient tooth and the donor tooth crown, R r Indicates the root overlap rate, G b represents the amount of bone grinding, C represents a constant, ω1 represents the weight of the crown overlap rate, ω2 represents the weight of the root overlap rate, and ω3 represents the weight of the bone grinding amount; S502: Calculating the optimization function value of each transplant surgery planning scheme; S503: The transplant surgery planning scheme with the minimum optimization function value is used as the optimal transplant surgery planning scheme.

8. An intelligent surgical planning system for autologous tooth transplantation, characterized in that: include: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method for intelligent surgical planning of autologous tooth transplantation according to any one of claims 1 to 7 is implemented.

Images