Method and system for tooth segmentation and tooth position identification of intraoral digital model

Abstract

The invention provides a method and a system for tooth segmentation and tooth position identification of an intraoral digital model. The method comprises the following steps: acquiring an intraoral three-dimensional model; generating a two-dimensional occlusion view perpendicular to an occlusion plane based on the intraoral three-dimensional model; segmenting the teeth in the two-dimensional occlusion view through an SAM model to obtain an occlusion segmentation mask; performing regression on a dental arch curve according to the occlusion segmentation mask; generating a plurality of two-dimensional dental crown views along the dental arch curve based on the intraoral three-dimensional model; segmenting teeth in each two-dimensional dental crown view through an SAM model to obtain a plurality of dental crown segmentation masks; and obtaining a segmentation result on the intraoral three-dimensional model based on the occlusion segmentation mask and the plurality of dental crown segmentation masks.

Description

Technical Field

[0001] The present invention relates to the field of oral medicine, and in particular to a method and system for tooth segmentation and tooth position recognition of an intraoral digital model. Background Art

[0002] Tooth segmentation and FDI numbering based on 3D optical intraoral scan (IOS) model data is an important step in the process of digital oral diagnosis and treatment. Due to the large variation in the number and shape of teeth, IOS tooth segmentation and tooth position identification pose great challenges. With the development of deep learning in the field of 3D image segmentation, automatic IOS segmentation methods mainly include two categories: (1) processing 2D rendered images; (2) directly processing 3D point cloud or mesh data, but both require detailed annotation by experienced professional doctors and belong to fully supervised learning models. The basic idea of ​​​​currently using weakly supervised learning to attempt IOS segmentation is to calibrate the centroid of all teeth and the masks of a few teeth to perform instance segmentation, but the centroid calibration process is complicated and cannot achieve tooth position identification and FDI numbering. Summary of the invention

[0003] In view of the problems in the prior art, combined with the SAM model (Segment Anything model) proposed in recent years for suggestible and category-independent segmentation, the SAM model pre-trained on more than 1 billion masks shows strong object understanding ability and achieves accurate two-dimensional scene segmentation, showing the potential of using weak annotation for tooth segmentation. The present invention provides a method for tooth segmentation and tooth position recognition of an intraoral digital model, the method comprising:

[0004] Obtaining intraoral 3D models;

[0005] generating a two-dimensional occlusal view perpendicular to the occlusal plane based on the intraoral three-dimensional model;

[0006] Segmenting the teeth in the two-dimensional occlusal view using a SAM model to obtain an occlusal segmentation mask;

[0007] regressing the dental arch curve according to the occlusal segmentation mask;

[0008] Generate a plurality of two-dimensional crown views along the dental arch curve based on the intraoral three-dimensional model;

[0009] The teeth in each two-dimensional crown view are segmented by the SAM model to obtain a number of crown segmentation masks;

[0010] A segmentation result is obtained on the intraoral three-dimensional model based on the occlusal segmentation mask and the plurality of tooth crown segmentation masks.

[0011] Furthermore, segmenting the teeth in the two-dimensional occlusal view by using the SAM model includes:

[0012] The SAM model is provided with class-related bounding box cues for each tooth, as well as foreground cues and background cues generated based on the bounding box cues.

[0013] Furthermore, regressing the dental arch curve according to the occlusal segmentation mask includes:

[0014] Based on the cusps found in the occlusal segmentation mask, the dental arch curve is regressed using cubic Bezier curves.

[0015] Furthermore, segmenting the teeth in each two-dimensional crown view by using the SAM model includes:

[0016] For any of the two-dimensional crown views, the overlapping portion of the segmentation result of the occlusal view and the current two-dimensional crown view on the intraoral three-dimensional model is projected into the current two-dimensional crown view as an initial prompt to the SAM model.

[0017] Furthermore, segmenting the teeth in each two-dimensional crown view by using the SAM model further includes:

[0018] The crown center of each tooth and the crown centers of its adjacent teeth in the two-dimensional crown view are provided to the SAM model as foreground cues and background cues, respectively.

[0019] Furthermore, obtaining a segmentation result on the intraoral three-dimensional model based on the occlusal segmentation mask and the plurality of tooth crown segmentation masks includes:

[0020] The view-aware label diffusion strategy is adopted to promote the segmentation results of each two-dimensional view to the three-dimensional domain and merge them, so as to obtain the segmentation result of the teeth in the intra-oral three-dimensional model.

[0021] Furthermore, the segmentation results obtained on the intraoral 3D model were evaluated by the average Dice score, overall accuracy, average sensitivity and average positive predictive value.

[0022] The present invention also provides a system for tooth segmentation and tooth position recognition of an intraoral digital model, the system comprising:

[0023] An occlusal prompter, which is used to provide a category-related bounding box prompt of each tooth to the SAM model according to a two-dimensional occlusal view, as well as a foreground prompt and a background prompt generated based on the bounding box prompt, so as to segment the teeth in the two-dimensional occlusal view, wherein the two-dimensional occlusal view is a view perpendicular to the occlusal plane generated based on the intraoral three-dimensional model;

[0024] A crown prompter, which is used for projecting the segmentation result of the occlusal view and the overlapping part of the current two-dimensional crown view on the intraoral three-dimensional model into the current two-dimensional crown view as an initial prompt to the SAM model, so as to segment the teeth in the two-dimensional crown views, wherein the two-dimensional crown views are views generated along the dental arch curve;

[0025] The view-aware label diffusion module is used to elevate the segmentation results of the two-dimensional occlusal view and the plurality of two-dimensional crown views to a three-dimensional domain and merge them, thereby obtaining a three-dimensional tooth segmentation result.

[0026] The present invention also provides a device for tooth segmentation and tooth position recognition of an intraoral digital model, the device comprising:

[0027] Processor; and

[0028] A memory arranged to store computer executable instructions which, when executed, cause the processor to perform the operations of the above-described method.

[0029] The present invention also provides a computer readable medium storing instructions, which, when executed, cause the system to perform the operations of the above method.

[0030] The method and system for tooth segmentation and tooth position recognition of an intraoral digital model of the present invention have the following effective effects:

[0031] 1. From the quantitative evaluation results, the method and system of the present invention outperform the previous weakly supervised methods in both global mean metric and label metric, and can achieve comparable performance to the fully supervised method when trained on low-cost weak annotations. These results show that using pre-trained shape knowledge in SAM can promote weakly annotated IOS segmentation, and the segmentation effect of posterior teeth is better than that of anterior teeth, and even better than the fully supervised learning model.

[0032] 2. From the qualitative evaluation results, the method and system of the present invention can visualize complex cases and have advantages in the segmentation and marking of posterior teeth and edge segmentation. It has the ability to correctly identify missing teeth and generate unified internal segmentation with clearer boundaries. The pre-trained shape knowledge embedded in SAM can promote the consistency of internal segmentation and boundary delineation.

[0033] 3. The system and method of the present invention greatly improve the efficiency and accuracy of IOS segmentation, greatly save the manpower and time costs of labeling, complete the task of automatic FDI tooth position identification and numbering, improve the key steps in the intelligent digital diagnosis and treatment process, and facilitate doctors to read films, diagnose and formulate treatment plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:

[0035] Figure 1 A schematic flow chart showing a method for teeth segmentation and tooth position recognition of an intraoral digital model according to an embodiment of the present invention;

[0036] Figure 2 An architectural diagram showing a system for teeth segmentation and tooth position recognition of an intraoral digital model according to an embodiment of the present invention;

[0037] Figure 3 Show Figure 2 A partial enlarged view of the mid-bite indicator;

[0038] Figure 4 Show Figure 2 A partial enlarged view of the central crown indicator;

[0039] Figure 5 Show Figure 2 Schematic diagram of the working of the view-aware label diffusion module;

[0040] Figure 6 The functional modules of an exemplary system that can be used in various embodiments of the present invention are shown.

[0041] The same or similar reference numerals in the drawings represent the same or similar components. DETAILED DESCRIPTION

[0042] The present application is described in further detail below in conjunction with the accompanying drawings.

[0043] In a typical configuration of the present invention, the terminal, the device of the service network and the trusted party each include one or more processors (eg, a central processing unit (CPU)), an input / output interface, a network interface and a memory.

[0044] Memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash memory. Memory is an example of a computer-readable medium.

[0045] Computer readable media include permanent and non-permanent, removable and non-removable media and can store information by any method or technology. The information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, Phase-Change Memory (PCM), Programmable Random Access Memory (PRAM), Static Random-Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read-Only Memory (ROM), Electrically-Erasable Programmable Read-Only Memory (EEPROM), Flash Memory or other memory technology, Compact Disc Read-Only Memory (CD-ROM), Digital Versatile Disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0046] The device referred to in the present invention includes but is not limited to user equipment, network equipment, or equipment formed by integrating user equipment and network equipment through a network. The user equipment includes but is not limited to any mobile electronic product that can interact with a user (for example, interact with a user through a touchpad), such as a smart phone, a tablet computer, etc. The mobile electronic product can use any operating system, such as an Android operating system, an iOS operating system, etc. Among them, the network device includes an electronic device that can automatically perform numerical calculations and information processing according to pre-set or stored instructions, and its hardware includes but is not limited to a microprocessor, an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a programmable logic device (Programmable Logic Device, PLD), a field programmable gate array (Field Programmable Gate Array, FPGA), a digital signal processor (Digital Signal Processor, DSP), an embedded device, etc. The network device includes but is not limited to a computer, a network host, a single network server, a plurality of network server sets or a cloud composed of a plurality of servers; here, the cloud is composed of a large number of computers or network servers based on cloud computing (Cloud Computing), wherein cloud computing is a kind of distributed computing, a virtual supercomputer composed of a group of loosely coupled computer sets. The network includes but is not limited to the Internet, a wide area network, a metropolitan area network, a local area network, a VPN network, a wireless ad hoc network, etc. Preferably, the device may also be a program running on the user device, the network device, or a device formed by integrating the user device and the network device, the network device, the touch terminal, or the network device and the touch terminal through a network.

[0047] Of course, those skilled in the art should understand that the above-mentioned devices are only examples, and other existing or future devices that are applicable to the present invention should also be included in the scope of protection of the present invention and are included here by reference.

[0048] In the description of the embodiments of the present invention, “plurality” means two or more, unless otherwise clearly and specifically defined.

[0049] The training of existing IOS automatic segmentation algorithm models still requires professional doctors to annotate point by point, which consumes a lot of time and manpower, and hinders the large-scale expansion of data volume and the promotion of learning-based algorithms. Models that can achieve IOS automatic segmentation under weak annotation need to calibrate the centroid of all teeth, which is a complicated process and cannot directly achieve tooth position recognition, that is, accurate tooth FDI numbering. Although the SAM model has powerful object understanding and two-dimensional segmentation capabilities, there are still some problems: (1) The efficiency and accuracy of SAM segmentation are highly dependent on the generated prompts; (2) SAM can only complete segmentation tasks but cannot actively complete classification tasks, such as FDI tooth position numbering; (3) SAM focuses on two-dimensional image segmentation, while IOS segmentation requires three-dimensional masks.

[0050] Figure 1 A schematic flow chart of a method for tooth segmentation and tooth position recognition of an intraoral digital model according to an embodiment of the present invention is shown, comprising the following steps:

[0051] Step 1: Obtain the intraoral 3D model.

[0052] In this embodiment, an intraoral scanner is used to obtain the intraoral three-dimensional model. The intraoral three-dimensional model can also be obtained by other methods such as CBCT.

[0053] Step 2: Generate a two-dimensional occlusal view perpendicular to the occlusal plane based on the intraoral three-dimensional model.

[0054] In this embodiment, the two-dimensional occlusal view is a rendered RGB and depth image perpendicular to the occlusal plane obtained based on the intraoral three-dimensional model.

[0055] Step 3: Segment the teeth in the two-dimensional occlusal view using the SAM model to obtain an occlusal segmentation mask.

[0056] In this embodiment, the generated two-dimensional occlusal view (specifically, its RGB image) is processed and a supplementary prompt related to the category is provided to the SAM model, so that the SAM model can segment the two-dimensional occlusal view and identify the tooth position FDI number, specifically including:

[0057] Provide category-related bounding box hints for each tooth in the two-dimensional occlusal view. The generation of tooth bounding boxes is the only model that needs to be trained in this method. Low-cost two-dimensional boundary annotations are manually performed on the two-dimensional occlusal view. Specifically, two points (a set of diagonal vertices of a rectangular box) and the corresponding FDI number are annotated for each tooth, that is, the bounding box annotation carries both the boundary information and FDI number information of each tooth. This is used to train the model, thereby generating a predicted bounding box containing the FDI number for each tooth in the two-dimensional occlusal view, and providing it to the SAM model as a bounding box hint.

[0058] Preferably, on this basis, the prompt is further enhanced, and a foreground prompt and a background prompt are generated according to the border prompt. Specifically, the center of the predicted bounding box is used as the foreground prompt, and a group of diagonal vertices of the predicted bounding box is used as the background prompt.

[0059] According to the various tips provided above, the SAM model is used to segment the teeth in the two-dimensional occlusal view to obtain an occlusal segmentation mask.

[0060] Step 4: regress the dental arch curve according to the occlusal segmentation mask.

[0061] In this embodiment, the dental arch curve is regressed using a cubic Bezier curve according to the tooth cusps found in the occlusal segmentation mask. The tooth cusps can be determined by the depth image of the two-dimensional occlusal view, that is, the point with the smallest depth in the depth image of each tooth segmented in the occlusal segmentation mask is the tooth cusp of the tooth.

[0062] Step 5: Generate a number of two-dimensional crown views along the dental arch curve based on the intraoral three-dimensional model.

[0063] In this embodiment, the two-dimensional crown view is a rendered RGB and depth image generated along the dental arch curve based on the intraoral three-dimensional model.

[0064] Step 6: Segment the teeth in each two-dimensional crown view using the SAM model to obtain a number of crown segmentation masks.

[0065] In this embodiment, for any of the two-dimensional crown views, the segmentation result of the occlusal view and the overlapping part of the current two-dimensional crown view on the intra-oral three-dimensional model are projected into the current two-dimensional crown view as an initial prompt to the SAM model. Specifically, the segmentation result of the two-dimensional occlusal view is first back-projected into the three-dimensional space, and the overlapping part of the current two-dimensional crown view and the two-dimensional occlusal view on the intra-oral three-dimensional model is calculated, and the overlapping part is then projected into the current two-dimensional crown view. At this time, the overlapping part also exists in the two-dimensional occlusal view containing semantic labels (i.e., in step three, the SAM model performs semantic segmentation on the two-dimensional occlusal view according to the various prompts provided, thereby generating semantic labels). Therefore, the semantic label contained in the overlapping part is inherited by the current two-dimensional crown view to form an initial segmentation mask, so that the mask centroid of the visible tooth is used as the initial prompt for the SAM model.

[0066] Preferably, on this basis, the prompt is further enhanced, and the crown center of each tooth in the two-dimensional crown view is used as a foreground prompt. The crown center of each tooth is calculated specifically by first obtaining a predicted bounding box of the tooth according to the generation model of the tooth border in step 2, and then approximating the crown width by the width of the predicted bounding box. t ) and length (CL t ) Calculate the crown length, where the r of the central incisor, lateral incisor and canine is preset t is 0.75, and the r of the premolar t Set to 0.8, the r of the molar t Set to 1.7; move the tooth tip of each tooth (the point with the smallest depth in the occlusal segmentation mask in the depth image) downward to the approximate center of the tooth crown, and use this point as the foreground cue. t , then the Z-axis coordinate of the corresponding crown center can be approximated as

[0067] Preferably, on the basis of the above, the prompt is further enhanced, and for each tooth in the two-dimensional crown view, the crown center of the adjacent teeth is used as a background prompt.

[0068] According to the various tips provided above, the SAM model is used to segment the teeth in each two-dimensional crown view to obtain a number of crown segmentation masks.

[0069] Step seven: obtaining a segmentation result on the intraoral three-dimensional model based on the occlusal segmentation mask and the plurality of tooth crown segmentation masks.

[0070] There are many methods in the prior art for converting multi-view two-dimensional views into three-dimensional views. In this embodiment, for the IOS model, a view-aware label diffusion strategy is adopted to promote the segmentation results of each two-dimensional view to the three-dimensional field and merge them, thereby obtaining the segmentation results of the teeth in the intra-oral three-dimensional model.

[0071] Specifically, in order to assign point-wise labels to the entire IOS model, all multi-view Figure 2 The dimensional segmentation results should be propagated to the 3D field and merged. First, the symbols used in the following are explained: the input IOS model has N points points, represented by By N views The rendered images form a two-dimensional view set represented as Where v1 is a 2D occlusion view. Each view v k Contains a set of pixel coordinates and the predicted one-hot labels Aggregate the 2D view pixel labels to form a label matrix Also initialize a zero matrix As the IOS model 3D label matrix, the goal is to Calculate and update L χ , this task is formulated as a graph-based semi-supervised label diffusion problem.

[0072] A graph is constructed using all 2D pixels and 3D points as nodes and camera projection relationships as edges. The graph consists of multiple subgraphs from different views. The adjacency matrix of The definition is as follows:

[0073]

[0074] in, is the diffusion intensity, It's pixels The three-dimensional coordinates obtained by back projection, the back projection formula is:

[0075]

[0076] in are homogeneous pixel coordinates, and from the extrinsic camera matrix, is the internal camera matrix. The diffusion strength is view-aware rather than fixed, and segmentation reliability increases when teeth are closer to the view center due to less morphological distortion.

[0077] For the crown view, the “comfort zone” (Z t ). Then, according to Z t The horizontal distance from the center to the center of the view is assigned decreasing diffusion intensity. The calculation formula is:

[0078]

[0079] where μ mid and μ t is the view and Z t The x-axis coordinate of the center, is the width of the view image, α and β are coefficients. For the 2D occlusal view v1 (all teeth are visible), set its After getting all subgraphs After that, the calculation formula of the IOS model label matrix is ​​as follows:

[0080]

[0081] Finally, the point-by-point label of the entire IOS model is obtained according to the following formula:

[0082]

[0083] Step 8: Segmentation result evaluation.

[0084] In this embodiment, the segmentation results obtained on the intraoral three-dimensional model were evaluated by the average Dice score, overall accuracy, average sensitivity and average positive predictive value.

[0085] Figure 2 The diagram shows the architecture of a system for tooth segmentation and tooth position recognition of an intraoral digital model according to an embodiment of the present invention. The system includes an occlusal prompter, a dental crown prompter, and a view-aware label diffusion module.

[0086] like Figure 3 As shown, the occlusal prompter includes a teeth detection module (TeethDetector Module), an occlusal prompt augmentation module (Occlusal Prompt Augmentation Module) and a dental arch curve regression module (Dental Arch Curve Regression Module).

[0087] For the intraoral three-dimensional model that requires tooth segmentation and tooth position identification, a two-dimensional occlusal view perpendicular to the occlusal plane is first generated based on the intraoral three-dimensional model. In this embodiment, the two-dimensional occlusal view is a rendered RGB and depth image perpendicular to the occlusal plane obtained based on the intraoral three-dimensional model.

[0088] The Teeth Detector Module is used to provide the SAM model with category-related bounding box hints for each tooth based on the input two-dimensional occlusal view (specifically, its RGB image). Specifically, the Teeth Detector Module is the only model that needs to be trained in this system. It trains the model by manually performing low-cost two-dimensional boundary annotations on the two-dimensional occlusal view. Specifically, two points (a set of diagonal vertices of a rectangular box) and the corresponding FDI number are annotated for each tooth, that is, the bounding box annotation carries both the boundary information and FDI number information of each tooth, thereby generating a predicted bounding box containing the FDI number for each tooth in the two-dimensional occlusal view, and providing it to the SAM model as a bounding box hint.

[0089] The Occlusal Prompt Augmentation Module is used to provide the predicted bounding box generated by the tooth detection module with the center of the predicted bounding box as the foreground prompt and a set of diagonal vertices of the predicted bounding box as the background prompt to the SAM model.

[0090] According to various prompts provided by the tooth detection module and the occlusion prompt enhancement module, the SAM model is used to segment the teeth in the two-dimensional occlusion view to obtain an occlusion segmentation mask.

[0091] The Dental Arch Curve Regression Module is used to regress the dental arch curve using a cubic Bezier curve according to the cusps found in the occlusal segmentation mask. The cusps of the teeth can be determined by the depth image of the two-dimensional occlusal view, that is, the point with the smallest depth in the depth image of each tooth segmented in the occlusal segmentation mask is the cusp of the tooth.

[0092] like Figure 4 As shown, the dental crown prompter includes a cover prompt module, a crown prior prompt module, and a crown prompt augmentation module.

[0093] For the above-mentioned intraoral three-dimensional model that requires tooth segmentation and tooth position identification, several two-dimensional crown views are generated based on the intraoral three-dimensional model along the dental arch curve obtained by the dental arch curve regression module. In this embodiment, the two-dimensional crown view is a rendered RGB and depth image generated based on the intraoral three-dimensional model along the dental arch curve.

[0094] The Cover Prompt Module is used to project the segmentation result of the occlusal view and the overlapping part of the current two-dimensional crown view on the intraoral three-dimensional model into the current two-dimensional crown view for any of the two-dimensional crown views as an initial prompt to the SAM model. Specifically, the segmentation result of the two-dimensional occlusal view is first back-projected into the three-dimensional space, and the overlapping part of the current two-dimensional crown view and the two-dimensional occlusal view on the intraoral three-dimensional model is calculated, and the overlapping part is then projected onto the current two-dimensional crown view. At this time, the overlapping part also exists in the two-dimensional occlusal view containing semantic labels (i.e., the SAM model performs semantic segmentation on the two-dimensional occlusal view according to various prompts provided by the occlusal prompter, thereby generating semantic labels). Therefore, the semantic label contained in the overlapping part is inherited by the current two-dimensional crown view to form an initial segmentation mask, so that the mask centroid of the visible tooth is used as the initial prompt for the SAM model.

[0095] The crown prior prompt module is used to provide the crown center of each tooth in the two-dimensional crown view as a foreground prompt to the SAM model. The crown center of each tooth is calculated by first obtaining the predicted bounding box of the tooth according to the tooth border generation model in the tooth detection module, and then approximating the crown width by the width of the predicted bounding box. t ) and length (CL t ) Calculate the crown length, where the r of the central incisor, lateral incisor and canine is preset t is 0.75, and the r of the premolar t Set to 0.8, the r of the molar t Set to 1.7; move the tooth tip of each tooth (the point with the smallest depth in the occlusal segmentation mask in the depth image) downward to the approximate center of the tooth crown, and use this point as the foreground cue. t , then the Z-axis coordinate of the corresponding crown center can be approximated as

[0096] The crown prompt augmentation module is used to provide the crown center of the adjacent teeth to the SAM model as a background prompt for each tooth in the two-dimensional crown view.

[0097] like Figure 5As shown, the view-aware label diffusion module adopts a view-aware label diffusion strategy to promote the segmentation results of each two-dimensional view to the three-dimensional domain and merge them, thereby obtaining the segmentation results of the teeth in the intraoral three-dimensional model. For details, please refer to the specific implementation of step 7 in the method embodiment, which will not be repeated here.

[0098] This embodiment further provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer code. When the computer code is executed, the method described in any of the preceding items is executed.

[0099] This embodiment further provides a computer program product. When the computer program product is executed by a computer device, the method described in any of the preceding items is executed.

[0100] This embodiment further provides a computer device, the computer device comprising:

[0101] one or more processors;

[0102] a memory for storing one or more computer programs;

[0103] When the one or more computer programs are executed by the one or more processors, the one or more processors are caused to implement the method as described in any of the preceding items.

[0104] Figure 6 An exemplary system is shown that can be used to implement the various embodiments described herein.

[0105] like Figure 6 As shown, in some embodiments, the system 1000 can be used as any user terminal device in each of the embodiments. In some embodiments, the system 1000 may include one or more computer-readable media (e.g., system memory or NVM / storage device 1020) having instructions and one or more processors (e.g., (one or more) processors 1005) coupled to the one or more computer-readable media and configured to execute instructions to implement modules to perform the actions described in the present invention.

[0106] For one embodiment, the system control module 1010 may include any suitable interface controller to provide any suitable interface to at least one of the processor(s) 1005 and / or any suitable device or component in communication with the system control module 1010 .

[0107] The system control module 1010 may include a memory controller module 1030 to provide an interface to the system memory 1015. The memory controller module 1030 may be a hardware module, a software module, and / or a firmware module.

[0108] The system memory 1015 may be used, for example, to load and store data and / or instructions for the system 1000. For one embodiment, the system memory 1015 may include any suitable volatile memory, such as a suitable DRAM. In some embodiments, the system memory 1015 may include double data rate type four synchronous dynamic random access memory (DDR4 SDRAM).

[0109] For one embodiment, system control module 1010 may include one or more input / output (I / O) controllers to provide interfaces to NVM / storage device 1020 and communication interface(s) 1025 .

[0110] For example, NVM / storage device 1020 may be used to store data and / or instructions. NVM / storage device 1020 may include any suitable non-volatile memory (e.g., flash memory) and / or may include any suitable non-volatile storage device(s) (e.g., one or more hard disk drives (HDDs), one or more compact disk (CD) drives, and / or one or more digital versatile disk (DVD) drives).

[0111] NVM / storage device 1020 may include storage resources that are physically part of the device on which system 1000 is installed, or it may be accessible to the device without being part of the device. For example, NVM / storage device 1020 may be accessed via communication interface(s) 1025 over a network.

[0112] Communication interface(s) 1025 may provide an interface for system 1000 to communicate over one or more networks and / or with any other suitable devices. System 1000 may wirelessly communicate with one or more components of a wireless network in accordance with any of one or more wireless network standards and / or protocols.

[0113] For one embodiment, at least one of the processor(s) 1005 may be packaged together with the logic of one or more controllers (e.g., memory controller module 1030) of the system control module 1010. For one embodiment, at least one of the processor(s) 1005 may be packaged together with the logic of one or more controllers of the system control module 1010 to form a system-in-package (SiP). For one embodiment, at least one of the processor(s) 1005 may be integrated on the same die with the logic of one or more controllers of the system control module 1010. For one embodiment, at least one of the processor(s) 1005 may be integrated on the same die with the logic of one or more controllers of the system control module 1010 to form a system on chip (SoC).

[0114] In various embodiments, the system 1000 may be, but is not limited to: a server, a workstation, a desktop computing device, or a mobile computing device (e.g., a laptop computing device, a handheld computing device, a tablet computer, a netbook, etc.). In various embodiments, the system 1000 may have more or fewer components and / or a different architecture. For example, in some embodiments, the system 1000 includes one or more cameras, a keyboard, a liquid crystal display (LCD) screen (including a touch screen display), a non-volatile memory port, multiple antennas, a graphics chip, an application specific integrated circuit (ASIC), and a speaker.

[0115] It should be noted that the present invention can be implemented in software and / or a combination of software and hardware, for example, can be implemented using an application specific integrated circuit (ASIC), a general purpose computer or any other similar hardware device. In one embodiment, the software program of the present invention can be executed by a processor to implement the steps or functions described above. Similarly, the software program of the present invention (including related data structures) can be stored in a computer readable recording medium, for example, a RAM memory, a magnetic or optical drive or a floppy disk and the like. In addition, some steps or functions of the present invention can be implemented using hardware, for example, as a circuit that cooperates with a processor to perform each step or function.

[0116] In addition, a part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of computer program instructions in computer-readable media includes, but is not limited to, source files, executable files, installation package files, etc., and accordingly, the way in which computer program instructions are executed by a computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0117] Communication media include media by which communication signals containing, for example, computer readable instructions, data structures, program modules, or other data are transmitted from one system to another. Communication media may include guided transmission media such as cables and wires (e.g., fiber optic, coaxial, etc.) and wireless (unguided transmission) media that can propagate energy waves, such as acoustic, electromagnetic, RF, microwave, and infrared. Computer readable instructions, data structures, program modules, or other data may be embodied as a modulated data signal in, for example, a wireless medium such as a carrier wave or similar mechanism such as embodied as part of spread spectrum technology. The term "modulated data signal" refers to a signal whose one or more characteristics are changed or set in such a manner as to encode information in the signal. Modulation may be analog, digital, or a hybrid modulation technique.

[0118] By way of example and not limitation, computer-readable storage media may include volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules or other data. For example, computer-readable storage media include, but are not limited to, volatile memory, such as random access memory (RAM, DRAM, SRAM); and non-volatile memory, such as flash memory, various read-only memories (ROM, PROM, EPROM, EEPROM), magnetic and ferromagnetic / ferroelectric memories (MRAM, FeRAM); and magnetic and optical storage devices (hard disks, magnetic tapes, CDs, DVDs); or other media now known or later developed that can store computer-readable information / data for use by a computer system.

[0119] Here, according to one embodiment of the present invention, a device is included, which includes a memory for storing computer program instructions and a processor for executing the program instructions, wherein, when the computer program instructions are executed by the processor, the device is triggered to run the methods and / or technical solutions based on the aforementioned multiple embodiments of the present invention.

[0120] It is obvious to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved. In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in the device claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

Claims

1. A method for tooth segmentation and tooth position recognition of an intraoral digital model, characterized in that: include: Obtaining intraoral 3D models; generating a two-dimensional occlusal view perpendicular to the occlusal plane based on the intraoral three-dimensional model; Segmenting the teeth in the two-dimensional occlusal view using a SAM model to obtain an occlusal segmentation mask; regressing the dental arch curve according to the occlusal segmentation mask; Generate a plurality of two-dimensional crown views along the dental arch curve based on the intraoral three-dimensional model; The teeth in each two-dimensional crown view are segmented by the SAM model to obtain a number of crown segmentation masks; A segmentation result is obtained on the intraoral three-dimensional model based on the occlusal segmentation mask and the plurality of tooth crown segmentation masks.

2. The method according to claim 1, characterized in that The segmenting of the teeth in the two-dimensional occlusal view by using the SAM model includes: The SAM model is provided with class-related bounding box cues for each tooth, as well as foreground cues and background cues generated based on the bounding box cues.

3. The method according to claim 1, characterized in that The regressing the dental arch curve according to the occlusal segmentation mask comprises: Based on the cusps found in the occlusal segmentation mask, the dental arch curve is regressed using cubic Bezier curves.

4. The method according to claim 1, characterized in that The segmentation of teeth in each two-dimensional crown view by using the SAM model includes: For any of the two-dimensional crown views, the overlapping portion of the segmentation result of the occlusal view and the current two-dimensional crown view on the intraoral three-dimensional model is projected into the current two-dimensional crown view as an initial prompt to the SAM model.

5. The method according to claim 4, characterized in that The segmenting of teeth in each two-dimensional crown view by using the SAM model also includes: The crown center of each tooth and the crown centers of its adjacent teeth in the two-dimensional crown view are provided to the SAM model as foreground cues and background cues, respectively.

6. The method according to claim 1, characterized in that The step of obtaining a segmentation result on the intraoral three-dimensional model based on the occlusal segmentation mask and the plurality of tooth crown segmentation masks comprises: The view-aware label diffusion strategy is adopted to promote the segmentation results of each two-dimensional view to the three-dimensional domain and merge them, so as to obtain the segmentation result of the teeth in the intra-oral three-dimensional model.

7. The method according to claim 1, characterized in that The segmentation results obtained on the intraoral 3D model were evaluated by the mean Dice score, overall accuracy, mean sensitivity and mean positive predictive value.

8. A system for tooth segmentation and tooth position recognition of an intraoral digital model, characterized in that: include: An occlusal prompter, which is used to provide a category-related bounding box prompt of each tooth to the SAM model according to a two-dimensional occlusal view, as well as a foreground prompt and a background prompt generated based on the bounding box prompt, so as to segment the teeth in the two-dimensional occlusal view, wherein the two-dimensional occlusal view is a view perpendicular to the occlusal plane generated based on the intraoral three-dimensional model; A crown prompter, which is used for projecting the segmentation result of the occlusal view and the overlapping part of the current two-dimensional crown view on the intraoral three-dimensional model into the current two-dimensional crown view as an initial prompt to the SAM model, so as to segment the teeth in the two-dimensional crown views, wherein the two-dimensional crown views are views generated along the dental arch curve; The view-aware label diffusion module is used to elevate the segmentation results of the two-dimensional occlusal view and the plurality of two-dimensional crown views to a three-dimensional domain and merge them, thereby obtaining a three-dimensional tooth segmentation result.

9. A device for tooth segmentation and tooth position recognition of an intraoral digital model, wherein: The device comprises: Processor; and A memory arranged to store computer executable instructions which, when executed, cause the processor to perform the operations of the method according to any one of claims 1 to 7.

10. A computer readable medium storing instructions which, when executed, cause a system to perform the operations of the method according to any one of claims 1 to 7.

Images