Tooth model adjusting method and device, storage medium and medical system

Abstract

The invention discloses a tooth model adjusting method and device, a storage medium and a medical system. The method comprises the following steps: acquiring point cloud data of a tooth model in a three-dimensional space; moving the tooth model to a preset position of a three-dimensional space based on the point cloud data; and adjusting the tooth model at the preset position to a first preset orientation based on the neural network model. According to the invention, the technical problem of low tooth model adjustment efficiency in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of tooth model adjustment, and in particular to a tooth model adjustment method, device, storage medium and medical system. Background Art

[0002] At present, in order to improve the display effect of the tooth model in the user interface, the subsequent automatic recognition effect of the tooth number in the tooth model and the segmentation effect of the tooth model, the axial direction of the tooth model is generally adjusted manually during the pre-processing of the tooth model. However, in the process of implementing the present invention, the inventors found that the manual adjustment of the tooth model requires repeated calibration, resulting in low efficiency of the axial adjustment.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present invention provide a method, device, storage medium and medical system for adjusting a tooth model, so as to at least solve the technical problem of low adjustment efficiency of a tooth model in the related art.

[0005] According to one aspect of an embodiment of the present invention, a method for adjusting a tooth model is provided, comprising: acquiring point cloud data of the tooth model in three-dimensional space; moving the tooth model to a preset position in the three-dimensional space based on the point cloud data; and adjusting the tooth model at the preset position to a first preset orientation based on a neural network model.

[0006] Optionally, the first preset orientation is used to represent at least one of the following: the orientation of the target tooth in the dental model coincides with the orientation of the first axis in the three-dimensional space, the orientation of the occlusal surface of the tooth in the dental model coincides with the orientation of the second axis in the three-dimensional space, and the orientation of the wide surface of the jaw in the dental model coincides with the orientation of the third axis in the three-dimensional space.

[0007] Optionally, adjusting the tooth model in a preset position to a first preset orientation based on a neural network model includes: determining a feature vector of point cloud data; rotating the tooth model based on the feature vector to obtain a rotated tooth model; and adjusting the rotated tooth model to the first preset orientation based on the neural network model.

[0008] Optionally, the rotated tooth model is adjusted to a first preset orientation based on a neural network model, including: using a point cloud dynamic graph convolutional neural network model to adjust the rotated tooth model so that the rotated tooth model is adjusted to a second preset orientation; using a residual neural network model to adjust the second preset orientation so that the rotated tooth model is adjusted to the first preset orientation.

[0009] Optionally, the rotated tooth model is adjusted using a point cloud dynamic graph convolutional neural network model so that the rotated tooth model is adjusted to a second preset orientation, including: using a point cloud dynamic convolutional neural network model to determine a first preset direction of the rotated tooth model in three-dimensional space; using the first preset direction to adjust the rotated tooth model so that the rotated tooth model is adjusted to a second preset orientation.

[0010] Optionally, a point cloud dynamic convolutional neural network model is used to determine a first preset direction of the rotated tooth model in three-dimensional space, including: using a point cloud dynamic convolutional neural network model to identify a first display direction of the rotated tooth model in three-dimensional space; classifying the first display direction to obtain a first category to which the first display direction belongs; and using a first preset correspondence relationship to determine a first preset direction corresponding to the first category, wherein the first preset correspondence relationship is used to represent a correspondence between the category and the first preset direction.

[0011] Optionally, the second preset orientation is used to represent at least one of the following: the orientation of the target tooth in the rotated tooth model coincides with the orientation of the first axis, the tooth occlusal surface in the rotated tooth model coincides with the target plane, and the target plane is constructed by the first axis and the third axis.

[0012] Optionally, the second preset orientation is adjusted using the residual neural network model so that the rotated tooth model is adjusted to the first preset orientation, including: using the residual neural network model to determine the second preset direction of the rotated tooth model in three-dimensional space; using the second preset direction to adjust the second preset orientation so that the rotated tooth model is adjusted to the first preset orientation.

[0013] Optionally, a residual neural network model is used to determine a second preset direction of the rotated tooth model in three-dimensional space, including: using a residual neural network model to identify a second display direction of the rotated tooth model in three-dimensional space; classifying the second display direction to obtain a second category to which the second display direction belongs; and using a second preset correspondence relationship to determine a second preset direction corresponding to the second category, wherein the second preset correspondence relationship is used to represent a correspondence between the category and the second preset direction.

[0014] Optionally, moving the tooth model to a preset position in three-dimensional space based on the point cloud data includes: determining the middle position of the point cloud data; determining the coordinate origin of the tooth model in three-dimensional space based on the middle position; and moving the tooth model to the preset position based on the coordinate origin.

[0015] According to one aspect of an embodiment of the present invention, a device for adjusting a tooth model is provided, comprising: an acquisition module for acquiring point cloud data of a tooth model in a three-dimensional space; a moving module for moving the tooth model to a preset position in the three-dimensional space based on the point cloud data; and an adjustment module for adjusting the tooth model at a preset position to a first preset orientation based on a neural network model.

[0016] According to one aspect of an embodiment of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein when the program is run, a processor of a device is controlled to execute the tooth model adjustment method of any one of the above embodiments.

[0017] According to one aspect of an embodiment of the present invention, a medical system is provided, comprising: one or more processors; a storage device for storing one or more programs; when the one or more programs are executed by one or more processors, the one or more processors execute the tooth model adjustment method of any one of the above-mentioned embodiments.

[0018] In an embodiment of the present invention, point cloud data of a tooth model in a three-dimensional space is obtained, the tooth model is moved to a preset position in the three-dimensional space based on the point cloud data, and the tooth model at the preset position is adjusted to a first preset orientation based on a neural network model, thereby improving the adjustment efficiency of the tooth model. It is easy to notice that in the process of adjusting the tooth model, the tooth model can be moved in combination with the point cloud data of the tooth model, which can reduce the manual adjustment process of the user, and the tooth model at the preset position can be adjusted to a first preset orientation based on the neural network model, which can further reduce the user's adjustment process, so as to improve the accuracy of the adjustment, and can improve the adjustment efficiency of the tooth model, thereby solving the technical problem of low adjustment efficiency of the tooth model in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0020] Figure 1 is a flow chart of a method for adjusting a tooth model according to an embodiment of the present invention;

[0021] Figure 2 is a schematic diagram of an upper jaw model in a tooth model according to an embodiment of the present application;

[0022] Figure 3a is a schematic diagram of a mandibular model in a tooth model according to an embodiment of the present application;

[0023] Figure 3b is a schematic diagram of a mandibular model in a tooth model according to an embodiment of the present application;

[0024] Figure 4 is a schematic diagram of a target tooth in a maxillary model according to an embodiment of the present application;

[0025] Figure 5 is a schematic diagram of a target tooth in a mandibular model according to an embodiment of the present application;

[0026] Figure 6 is a schematic diagram of a tooth occlusal surface in a maxillary model according to an embodiment of the present application;

[0027] Figure 7 is a schematic diagram of the wide surface of the maxillary model according to an embodiment of the present application;

[0028] Figure 8 is a rotated mandibular model according to an embodiment of the present application;

[0029] Fig. 9 A mandibular model in a first preset orientation according to an embodiment of the present application;

[0030] Fig.10 is a mandibular model in a second preset orientation according to an embodiment of the present application;

[0031] Fig.11 is a maxillary model in a second preset orientation according to an embodiment of the present application;

[0032] Fig.12 is a category schematic diagram according to an embodiment of the present application;

[0033] Fig.13 is another category schematic diagram according to an embodiment of the present application;

[0034] Fig.14 is a schematic diagram of an adjustment device for a tooth model according to an embodiment of the present invention. DETAILED DESCRIPTION

[0035] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0036] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0037] Example 1

[0038] According to an embodiment of the present invention, an embodiment of a method for adjusting a tooth model is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0039] Figure 1 is a flow chart of a method for adjusting a tooth model according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0040] Step S102, obtaining point cloud data of the tooth model in three-dimensional space.

[0041] The above-mentioned tooth model may be a tooth model pre-generated according to the patient's tooth information and tooth image, and the tooth model may be a tooth model that has not been processed.

[0042] The above-mentioned tooth model may be a three-dimensional model.

[0043] The above-mentioned point cloud data is used to represent the three-dimensional point cloud data of the tooth model in the three-dimensional space.

[0044] In an optional embodiment, the tooth model may be preliminarily moved by acquiring point cloud data of the tooth model in three-dimensional space, so that the tooth model may be moved to a suitable position.

[0045] Figure 2 is a schematic diagram of an upper jaw model in a tooth model according to an embodiment of the present application, such as Figure 2 The tooth model shown may be the maxillary model initially displayed in the three-dimensional space, and the coordinate axis on the left may be the three-dimensional coordinate axis in the three-dimensional space, wherein the three-dimensional coordinate axis includes an x-axis, a y-axis, and a z-axis. Figure 2The three-dimensional coordinate axes in can be world coordinate axes.

[0046] Figure 3a is a schematic diagram of a mandibular model in a tooth model according to an embodiment of the present application, such as Figure 3a The tooth model shown may be a mandibular model initially displayed in a three-dimensional space, and the coordinate axis on the left may be a three-dimensional coordinate axis in the three-dimensional space, wherein the three-dimensional coordinate axis includes an x-axis, a y-axis, and a z-axis. Figure 3a The three-dimensional coordinate axes in can be world coordinate axes.

[0047] Step S104: moving the tooth model to a preset position in the three-dimensional space based on the point cloud data.

[0048] The above-mentioned preset position can be the central position of the three-dimensional space where the tooth model is located, the preset position can also be the central position of the display interface, and the above-mentioned preset position can also be any pre-set position. This is only an example and is not limited to any example.

[0049] The above-mentioned preset position may also be the three-dimensional coordinate origin of the three-dimensional coordinate in the three-dimensional space.

[0050] In an optional embodiment, the preset position can be set according to actual needs so that the tooth model can be moved to a suitable position in the three-dimensional space.

[0051] In order to facilitate subsequent tooth detection and tooth segmentation on the tooth model, the tooth model needs to be moved to the central position of the three-dimensional space.

[0052] Figure 3b is a schematic diagram of a mandibular model in a tooth model according to an embodiment of the present application, such as Figure 3b The mandibular model shown may be a mandibular model obtained by adjusting the mandibular model initially displayed in the three-dimensional space to the central position of the three-dimensional space.

[0053] In an optional embodiment, in the process of actually applying the tooth model, in order to facilitate subsequent processing of the tooth model, it is necessary to move the tooth model to a preset position in the three-dimensional space when adjusting the axial direction of the tooth model.

[0054] Step S106: adjusting the tooth model at the preset position to a first preset orientation based on the neural network model.

[0055] The above-mentioned neural network model may be a convolutional neural network model, but is not limited thereto.

[0056] The number of the above-mentioned neural network models can be one or more, and the number of the neural network models is not limited here.

[0057] The above-mentioned first preset orientation can be the orientation of a pre-set tooth model. The first preset orientation here can be set according to the actual task scenario. For example, in subsequent processing, the widest surface of the jaw is often used as the display surface facing the user to facilitate subsequent tooth segmentation and tooth recognition.

[0058] In an optional embodiment, a neural network model can be used to adjust the tooth model in a preset position to a first preset orientation, thereby improving the adjustment efficiency and accuracy of the tooth model; a traditional algorithm can also be used to adjust the tooth model in a preset position to the first preset orientation.

[0059] Through the above steps, the point cloud data of the tooth model in the three-dimensional space is obtained, the tooth model is moved to a preset position in the three-dimensional space based on the point cloud data, and the tooth model at the preset position is adjusted to a first preset orientation based on the neural network model, thereby achieving the technical purpose of improving the adjustment efficiency of the tooth model. It is easy to notice that in the process of adjusting the tooth model, the tooth model can be moved in combination with the point cloud data of the tooth model, which can reduce the manual adjustment process of the user, and the tooth model at the preset position can be adjusted to a first preset orientation based on the neural network model, which can further reduce the user's adjustment process, so as to improve the accuracy of the adjustment, and can improve the adjustment efficiency of the tooth model, thereby solving the technical problem of low adjustment efficiency of the tooth model in the related art.

[0060] In the above embodiments of the present application, the first preset orientation is used to represent at least one of the following: the orientation of the target tooth in the dental model coincides with the orientation of the first axis in the three-dimensional space, the orientation of the occlusal surface of the tooth in the dental model coincides with the orientation of the second axis in the three-dimensional space, and the orientation of the wide surface of the jaw in the dental model coincides with the orientation of the third axis in the three-dimensional space.

[0061] The first axis, the second axis, and the third axis mentioned above may be different axes in the three-dimensional coordinate axes in the three-dimensional space.

[0062] The first axis mentioned above can be the z-axis of the three-dimensional coordinate axis in the three-dimensional space, the second axis mentioned above can be the y-axis of the three-dimensional coordinate axis, and the third axis mentioned above can be the x-axis of the three-dimensional coordinate axis in the three-dimensional space. This is only for example description without any limitation.

[0063] The target teeth mentioned above may be incisors in the tooth model. In the maxillary model of the tooth model, the target teeth may be upper incisors, and in the mandibular model of the tooth model, the target teeth may be lower incisors. The orientation of the target teeth may be determined by a first connecting line obtained by connecting the center point of the tooth model to a point between the two incisors. If the first connecting line is parallel to the first axis, it means that the orientation of the target teeth coincides with the orientation of the first axis in the three-dimensional space.

[0064] Figure 4 is a schematic diagram of a target tooth in a maxillary model according to an embodiment of the present application, such as Figure 4 As shown, in the maxillary model, the target tooth may be tooth No. 1, that is, the target tooth is the upper incisor in the maxillary model.

[0065] Figure 5 is a schematic diagram of a target tooth in a mandibular model according to an embodiment of the present application, such as Figure 5 As shown, in the mandibular model, the target tooth may be tooth No. 2, that is, the target tooth is the lower incisor in the mandibular model.

[0066] The above-mentioned tooth occlusal surface can be the surface where the tooth occlusal part is located in the tooth model. The direction of the tooth occlusal surface can be determined by the normal line of the tooth occlusal surface. If the normal line is parallel to the second axis, it means that the direction of the tooth occlusal surface coincides with the direction of the second axis in the three-dimensional space.

[0067] Figure 6 is a schematic diagram of the occlusal surface of teeth in a maxillary model according to an embodiment of the present application, such as Figure 6 As shown, in the maxillary model, the tooth occlusal surface can be Figure 6 The surface where the tooth occlusal part is located, that is, the surface composed of the x-axis and the y-axis is the tooth occlusal surface.

[0068] The above-mentioned maxillofacial wide surface may be the widest surface displayed in the tooth model. The maxillofacial wide surface may be represented by a connecting line between the teeth at both ends. If the connecting line is parallel to the third axis, it means that the direction of the maxillofacial wide surface coincides with the direction of the third axis in the three-dimensional space.

[0069] Figure 7 is a schematic diagram of the wide surface of the maxillary model according to an embodiment of the present application, such as Figure 7 As shown, in the maxillary model, the wide surface of the jaw can be based on Figure 7 The teeth at both ends are determined.

[0070] By adjusting the tooth model to the first preset orientation, the user can have a better experience when the tooth model faces the user, and subsequent operations on the tooth model are convenient for the user.

[0071] It should be noted that the first preset orientation may be changed according to user needs, and the first preset orientation may also be a pre-set orientation, which is not limited here and can be flexibly set according to actual needs.

[0072] In the above embodiment of the present application, the tooth model in a preset position is adjusted to a first preset orientation based on a neural network model, including: determining a feature vector of the point cloud data; rotating the tooth model based on the feature vector to obtain a rotated tooth model; and adjusting the rotated tooth model to the first preset orientation based on the neural network model.

[0073] In an optional embodiment, the spatial placement of the tooth model can be normalized by determining the eigenvector of the point cloud data, that is, the tooth model is rotated to obtain the rotated tooth model. The rotated tooth model can be processed by a neural network to obtain the rotation angle of the rotated tooth model. The tooth model can be adjusted according to the rotation angle so that the rotated tooth model is adjusted to a first preset orientation.

[0074] The feature vector of the point cloud data may be obtained by a principal component analysis (PCA) algorithm, wherein the principal component analysis direction may be a data dimensionality reduction algorithm. The feature vector may be a feature vector obtained by analyzing the point cloud data by PCA.

[0075] In an optional embodiment, PCA is used to analyze the point cloud data to obtain a feature vector, wherein the feature vector is used to represent the main direction of the point cloud data, and the feature vector is multiplied by the tooth model in the reverse direction so that the tooth model can be rotated to the main direction of the point cloud data, thereby obtaining a rotated tooth model. The reverse multiplication refers to rotating the tooth model in the reverse direction based on the feature vector of the tooth model in three-dimensional space.

[0076] When the tooth model is a maxillary model, the eigenvector of the point cloud data corresponding to the maxillary model can be determined, and the maxillary model can be rotated based on the eigenvector to obtain the rotated maxillary model. Optionally, PCA can be used to analyze the point cloud data to obtain the eigenvector of the maxillary model, and the eigenvector is multiplied in the opposite direction with the tooth model so that the maxillary model can be rotated to the main direction of the point cloud data, thereby obtaining the rotated maxillary model.

[0077] When the tooth model is a mandibular model, the characteristic vector of the point cloud data corresponding to the mandibular model can be determined, and the mandibular model can be rotated based on the characteristic vector to obtain a rotated mandibular model. Figure 8 is a rotated mandibular model according to an embodiment of the present application, such as Figure 8 Optionally, PCA can be used to analyze the point cloud data to obtain the eigenvector of the mandibular model, and the eigenvector is multiplied in the opposite direction with the tooth model so that the mandibular model can be rotated to the main direction of the point cloud data, thereby obtaining the rotated mandibular model.

[0078] In another optional embodiment, the direction of the rotated tooth model in three-dimensional space can be determined according to the neural network model, and the tooth model can be adjusted to a first preset orientation based on the direction.

[0079] like Figure 7 The maxillary model adjusted to the first preset orientation is shown, the orientation of the target tooth in the maxillary model coincides with the orientation of z in the three-dimensional space, the orientation of the tooth occlusal surface in the maxillary model coincides with the orientation of y in the three-dimensional space, and the orientation of the wide surface of the maxillary model coincides with the orientation of x in the three-dimensional space. Figure 7 Shown is the final presentation of the maxillary model.

[0080] Fig. 9 is a mandibular model in a first preset orientation according to an embodiment of the present application, such as Fig. 9 As shown in , the orientation of the target tooth in the mandibular model coincides with the orientation of z in three-dimensional space, the orientation of the tooth occlusal surface in the tooth model coincides with the orientation of y in three-dimensional space, and the orientation of the jaw width surface in the mandibular model coincides with the orientation of x in three-dimensional space. Fig. 9 Shown is the final presentation of the mandibular model.

[0081] In the above embodiment of the present application, the rotated tooth model is adjusted to a first preset orientation based on a neural network model, including: using a point cloud dynamic graph convolutional neural network model to adjust the rotated tooth model so that the rotated tooth model is adjusted to a second preset orientation; using a residual neural network model to adjust the second preset orientation so that the rotated tooth model is adjusted to the first preset orientation.

[0082] In an optional embodiment, since the randomness of adjusting the rotated tooth model in space using the point cloud dynamic graph convolutional neural network model is very large, it is necessary to add a residual neural network model to adjust the rotated tooth model. The second preset orientation is to place the tooth model on a suitable plane in space, and the first preset orientation is to place the tooth model in a position facing the display.

[0083] The above-mentioned point cloud dynamic graph convolutional neural network model (Dynamic Graph CNN for Learning on Point Clouds, referred to as DGCNN) is used for point cloud extraction and point cloud segmentation.

[0084] In an optional embodiment, the rotated tooth model can be extracted using a point cloud dynamic graph convolutional neural network to obtain point cloud data of the rotated tooth model. The point cloud data can be segmented to obtain the approximate direction of the rotated tooth model in three-dimensional space, and the rotated tooth model can be rotated to a second preset orientation so that the residual neural network model can be used for subsequent fine-tuning so that the rotated tooth model is adjusted to the first preset orientation.

[0085] The above residual neural network model is not limited. For example, it can be a semi-supervised semantic segmentation network (MinkResUnet) based on deep learning, or a convolutional neural network (MinkowskiNet) for point cloud segmentation and semantic segmentation.

[0086] The above-mentioned residual neural network model is used to finely adjust the first preset orientation and adjust the rotated tooth model to the second preset orientation. By adding the residual neural network model, the adjustment accuracy of the tooth model can be improved.

[0087] In an optional embodiment, the rotated tooth model can be roughly adjusted using a point cloud dynamic graph convolutional neural network, and then the model can be finely adjusted using a residual neural network model, so that the model can finally be accurately adjusted to the first preset orientation.

[0088] Fig.10 is a mandibular model in a second preset orientation according to an embodiment of the present application, such as Fig.10 The mandibular model adjusted to the second preset orientation is shown, wherein the orientation of the target tooth in the rotated tooth model coincides with the z orientation, and the tooth occlusal surface in the rotated mandibular model coincides with the target plane, and the target plane is constructed by the z-axis and the x-axis.

[0089] Fig.11 is a maxillary model in a second preset orientation according to an embodiment of the present application, such as Fig.11 The maxillary model adjusted to the second preset orientation is shown, wherein the orientation of the target tooth in the rotated tooth model coincides with the z orientation, and the occlusal surface of the tooth in the rotated maxillary model coincides with the target plane, and the target plane is constructed by the z-axis and the x-axis.

[0090] In the above embodiment of the present application, the rotated tooth model is adjusted using a point cloud dynamic graph convolutional neural network model so that the rotated tooth model is adjusted to a second preset orientation, including: using a point cloud dynamic convolutional neural network model to determine a first preset direction of the rotated tooth model in three-dimensional space; using the first preset direction to adjust the rotated tooth model so that the rotated tooth model is adjusted to a second preset orientation.

[0091] The first preset direction may be a preset direction. The first preset direction corresponding to the rotated tooth model may be obtained by identifying the point cloud data of the rotated tooth model.

[0092] The point cloud dynamic convolutional neural network model can be used to identify the rotated tooth model and determine the first preset direction of the rotated tooth model in the three-dimensional space, so as to adjust the rotated tooth model according to the first preset direction of the rotated tooth model in the three-dimensional space, so that the rotated tooth model is adjusted to the second preset orientation. Optionally, the rotation relationship between the preset direction and the preset orientation can be preset, and after the preset direction is determined, the preset direction can be rotated to the preset orientation according to the rotation relationship. For example, the first preset direction is 20°, and the second preset orientation is 40°. According to the preset rotation relationship, the first preset direction is rotated by 20°, and the rotated tooth model in the first preset direction can be adjusted to the second preset orientation.

[0093] In the above embodiment of the present application, a point cloud dynamic convolutional neural network model is used to determine a first preset direction of the rotated tooth model in three-dimensional space, including: using a point cloud dynamic convolutional neural network model to identify a first display direction of the rotated tooth model in three-dimensional space; classifying the first display direction to obtain a first category to which the first display direction belongs; and determining a first preset direction corresponding to the first category using a first preset correspondence relationship, wherein the first preset correspondence relationship is used to represent a correspondence between a category and a first preset direction.

[0094] The first display direction mentioned above may be the direction in which the rotated tooth model is actually displayed in the three-dimensional space.

[0095] The above-mentioned first preset correspondence can be a correspondence between a pre-designed category and a preset direction, wherein the category is used to represent the category of the model direction. The three-dimensional space can be divided into 24 categories, and the 24 categories are equally divided into rotation angles in the three-dimensional space. The 24 categories here are only for example. The three-dimensional space can also be divided into other numbers of categories, and the angles of the directions between different categories can be the same.

[0096] Fig.12 is a schematic diagram of a category according to an embodiment of the present application, such as Fig.12As shown, 24 categories can be described in three-dimensional space by displaying spheres, wherein different categories can be displayed by cones. It should be noted that the first category includes but is not limited to a first preset category and a second preset category. The first preset direction corresponding to the first preset category is the first direction, and the second preset direction corresponding to the second preset category is the second direction. The first display direction is the direction in which the rotated tooth model is actually displayed in the three-dimensional space; if the first display direction is displayed in the cone of the first preset category, it means that the first category to which the first display direction belongs is the first preset category, and the first preset correspondence relationship can be used to determine that the first preset direction corresponding to the first preset category is the first direction; if the first display direction is displayed in the cone of the second preset category, it means that the first category to which the first display direction belongs is the second preset category, and the first preset correspondence relationship can be used to determine that the second preset direction corresponding to the first preset category is the second direction.

[0097] In an optional embodiment, a point cloud dynamic convolutional neural network model can be used to identify the rotated tooth model to obtain a first display direction of the rotated tooth model in the three-dimensional space, that is, the actual display direction of the rotated tooth model in the three-dimensional space. The first display direction can be classified to obtain a first category. Optionally, a first category corresponding to the first display direction can be determined from multiple preset categories in the three-dimensional space, and the first preset direction referred to by the first category can be determined by calling the first preset correspondence.

[0098] In the above embodiments of the present application, the second preset orientation is used to represent at least one of the following: the orientation of the target tooth in the rotated tooth model coincides with the orientation of the first axis, the occlusal surface of the tooth in the rotated tooth model coincides with the target plane, and the target plane is constructed by the first axis and the third axis.

[0099] The first axis may be the z axis in the three-dimensional coordinate axis, the second axis may be the y axis in the three-dimensional coordinate axis, and the third axis may be the x axis in the three-dimensional coordinate axis. The target plane may be a plane composed of the x axis and the z axis, wherein the tooth occlusal surface may coincide with the target plane.

[0100] In the above embodiment of the present application, the second preset orientation is adjusted using the residual neural network model so that the rotated tooth model is adjusted to the first preset orientation, including: using the residual neural network model to determine the second preset direction of the rotated tooth model in three-dimensional space; using the second preset direction to adjust the second preset orientation so that the rotated tooth model is adjusted to the first preset orientation.

[0101] In an optional embodiment, after the tooth model is rotated to the second preset orientation, the orientation of the tooth model may be further finely adjusted so that the tooth model may eventually face the user.

[0102] like Figure 4 As shown, the orientation of the target tooth in the rotated tooth model coincides with the orientation of the z-axis, the orientation of the tooth occlusal surface in the rotated tooth model coincides with the orientation of the y-axis, and the orientation of the wide surface of the jaw in the rotated tooth model coincides with the x-axis in the three-dimensional space.

[0103] The second preset direction may be a pre-set direction. The second preset direction corresponding to the rotated tooth model may be obtained by identifying the point cloud data of the rotated tooth model.

[0104] The residual neural network model can be used to identify the rotated tooth model and determine the second preset direction of the rotated tooth model in the three-dimensional space. The rotated tooth model can be adjusted according to the rotational relationship between the preset second preset direction and the first preset orientation, so that the rotated tooth model can be adjusted according to the second preset direction of the rotated tooth model in the three-dimensional space so that the rotated tooth model is adjusted to the first preset orientation.

[0105] It should be noted that the convolutional neural network can be used to accurately adjust the general direction of the tooth model first, and then the residual neural network can be used to further fine-tune the tooth model to improve the adjustment accuracy of the tooth model, thereby obtaining a tooth model with better display effect.

[0106] In the above embodiment of the present application, a residual neural network model is used to determine the second preset direction of the rotated tooth model in the three-dimensional space, including: using the residual neural network model to identify the second display direction of the rotated tooth model in the three-dimensional space; classifying the second display direction to obtain the second category to which the second display direction belongs; and using a second preset correspondence relationship to determine the second preset direction corresponding to the second category, wherein the second preset correspondence relationship is used to represent the correspondence between the category and the second preset direction.

[0107] The second display direction mentioned above may be the direction in which the rotated tooth model is actually displayed in the three-dimensional space.

[0108] The number of categories in the second preset corresponding relationship may be greater than the number of categories in the first preset corresponding relationship.

[0109] The second preset corresponding relationship can be used for more refined category classification.

[0110] The above-mentioned second preset correspondence can be a correspondence between a pre-designed category and a preset direction, wherein the category is used to represent the category of the model direction. The three-dimensional space can be divided into 37 categories, and the 37 categories are equally divided into rotation angles in the three-dimensional space. The degrees of the preset directions corresponding to the classes in the 37 categories can differ by 5°. The 37 categories and 5° here are only for illustration. The three-dimensional space can also be divided into other numbers of classes, and the degrees of the directions between different classes can be the same.

[0111] Fig.13 is another type of schematic diagram according to an embodiment of the present application, such as Fig.13 As shown, the second display direction of the rotated tooth model in three-dimensional space can be abstractly described. It should be noted that the second category includes but is not limited to the third preset category and the fourth preset category. The second preset direction corresponding to the third preset category is the third direction, and the second preset direction corresponding to the fourth preset category is the fourth direction. The second display direction is the actual display direction of the rotated tooth model in three-dimensional space. The second display direction can be projected onto the plane constructed by the x-axis and the z-axis. If the projection direction of the second display direction is displayed in the third preset category, it means that the second category to which the second display direction belongs is the third preset category. The second preset correspondence relationship can be used to determine that the second preset direction corresponding to the third preset category is the third direction; if the second category to which the second display direction belongs is the fourth preset category, the second preset correspondence relationship can be used to determine that the second preset direction corresponding to the fourth preset category is the fourth direction.

[0112] In an optional embodiment, the rotated tooth model can be identified using a residual neural network model to obtain a second display direction of the rotated tooth model in the three-dimensional space, that is, the actual display direction of the rotated tooth model in the three-dimensional space. The second display direction can be classified to obtain a second category. Optionally, the second category corresponding to the second display direction can be determined from the categories pre-set in the three-dimensional space, and the second preset direction referred to by the second category can be determined by calling the second preset correspondence.

[0113] In the above embodiment of the present application, the tooth model is moved to a preset position in the three-dimensional space based on the point cloud data, including: determining the middle position of the point cloud data; determining the coordinate origin of the tooth model in the three-dimensional space based on the middle position; and moving the tooth model to the preset position based on the coordinate origin.

[0114] In an optional embodiment, the middle position of the point cloud data can be calculated by a relevant calculation method, and reverse movement can be performed based on the middle position to obtain the three-dimensional coordinate origin of the tooth model in the three-dimensional space. The tooth model can be moved to a preset position based on the three-dimensional coordinate origin. Optionally, the preset position can be the three-dimensional coordinate origin, and the tooth model can be moved to the three-dimensional coordinate origin.

[0115] Example 2

[0116] According to another aspect of an embodiment of the present invention, a schematic diagram of a tooth model adjustment device is also provided, and the device can execute the tooth model adjustment method in the above-mentioned embodiment 1. The specific implementation scheme and application scenario in this embodiment are the same as those in the above-mentioned embodiment 1, and will not be repeated here.

[0117] Fig.14 is a schematic diagram of an adjustment device for a tooth model according to an embodiment of the present invention. Fig.14 As shown, the device 1400 includes: an acquisition module 1402 , a movement module 1404 , and an adjustment module 1406 .

[0118] Among them, the acquisition module is used to obtain point cloud data of the tooth model in the three-dimensional space; the moving module is used to move the tooth model to a preset position in the three-dimensional space based on the point cloud data; the adjustment module is used to adjust the tooth model at the preset position to a first preset orientation based on the neural network model.

[0119] In the above embodiments of the present application, the first preset orientation is used to represent at least one of the following: the orientation of the target tooth in the dental model coincides with the orientation of the first axis in the three-dimensional space, the orientation of the occlusal surface of the tooth in the dental model coincides with the orientation of the second axis in the three-dimensional space, and the orientation of the wide surface of the jaw in the dental model coincides with the orientation of the third axis in the three-dimensional space.

[0120] In the above embodiment of the present application, the adjustment module is used to determine the feature vector of the point cloud data, rotate the tooth model based on the feature vector to obtain the rotated tooth model, and adjust the rotated tooth model to a first preset orientation based on the neural network model.

[0121] In the above embodiment of the present application, the adjustment module is used to adjust the rotated tooth model using the point cloud dynamic graph convolutional neural network model, so that the rotated tooth model is adjusted to the second preset orientation, and the second preset orientation is adjusted using the residual neural network model, so that the rotated tooth model is adjusted to the first preset orientation.

[0122] In the above embodiment of the present application, the adjustment module is used to use the point cloud dynamic convolutional neural network model to determine the first preset direction of the rotated tooth model in the three-dimensional space, and use the first preset direction to adjust the rotated tooth model so that the rotated tooth model is adjusted to the second preset orientation.

[0123] In the above embodiment of the present application, the adjustment module is used to use the point cloud dynamic convolutional neural network model to identify the first display direction of the rotated tooth model in the three-dimensional space, classify the first display direction, obtain the first category to which the first display direction belongs, and use the first preset correspondence relationship to determine the first preset direction corresponding to the first category, wherein the first preset correspondence relationship is used to represent the correspondence between the category and the first preset direction.

[0124] In the above embodiments of the present application, the second preset orientation is used to represent at least one of the following: the orientation of the target tooth in the rotated tooth model coincides with the orientation of the first axis, the occlusal surface of the tooth in the rotated tooth model coincides with the target plane, and the target plane is constructed by the second axis and the third axis.

[0125] In the above embodiment of the present application, the adjustment module is used to use the residual neural network model to determine the second preset direction of the rotated tooth model in the three-dimensional space, and use the second preset direction to adjust the second preset orientation so that the rotated tooth model is adjusted to the first preset orientation.

[0126] In the above embodiment of the present application, the adjustment module is used to use the residual neural network model to identify the second display direction of the rotated tooth model in the three-dimensional space, classify the second display direction, obtain the second category to which the second display direction belongs, and use the second preset correspondence relationship to determine the second preset direction corresponding to the second category, wherein the second preset correspondence relationship is used to represent the correspondence between the category and the second preset direction.

[0127] In the above embodiments of the present application, the moving module is also used to determine the middle position of the point cloud data, determine the coordinate origin of the tooth model in the three-dimensional space based on the middle position, and move the tooth model to a preset position based on the coordinate origin.

[0128] Example 3

[0129] According to one aspect of an embodiment of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein when the program is run, a processor of a device is controlled to execute the tooth model adjustment method of any one of the above embodiments.

[0130] Example 4

[0131] According to one aspect of an embodiment of the present invention, a medical system is provided, comprising: one or more processors; a storage device for storing one or more programs; when the one or more programs are executed by one or more processors, the one or more processors execute the tooth model adjustment method of any one of the above-mentioned embodiments.

[0132] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0133] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0134] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, 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 interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0135] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0136] 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. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0137] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0138] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for adjusting a tooth model, characterized in that: include: Obtaining point cloud data of the tooth model in three-dimensional space; moving the tooth model to a preset position in the three-dimensional space based on the point cloud data; The tooth model at the preset position is adjusted to a first preset orientation based on the neural network model.

2. The method for adjusting a tooth model according to claim 1, characterized in that: The first preset orientation is used to represent at least one of the following: the orientation of the target tooth in the dental model coincides with the orientation of the first axis in the three-dimensional space, the orientation of the occlusal surface of the tooth in the dental model coincides with the orientation of the second axis in the three-dimensional space, and the orientation of the wide surface of the jaw in the dental model coincides with the orientation of the third axis in the three-dimensional space.

3. The method for adjusting a tooth model according to claim 2, characterized in that: Adjusting the tooth model at the preset position to a first preset orientation based on the neural network model includes: Determining a feature vector of the point cloud data; Rotating the tooth model based on the feature vector to obtain a rotated tooth model; The rotated tooth model is adjusted to the first preset orientation based on the neural network model.

4. The method for adjusting a tooth model according to claim 3, characterized in that: Adjusting the rotated tooth model to the first preset orientation based on the neural network model includes: Using a point cloud dynamic graph convolutional neural network model to adjust the rotated tooth model so that the rotated tooth model is adjusted to a second preset orientation; The second preset orientation is adjusted using a residual neural network model so that the rotated tooth model is adjusted to the first preset orientation.

5. The method for adjusting a tooth model according to claim 4, characterized in that: The method of adjusting the rotated tooth model by using a point cloud dynamic graph convolutional neural network model so that the rotated tooth model is adjusted to the second preset orientation includes: Determining a first preset direction of the rotated tooth model in the three-dimensional space using the point cloud dynamic convolutional neural network model; The rotated tooth model is adjusted using the first preset direction so that the rotated tooth model is adjusted to the second preset orientation.

6. The method for adjusting a tooth model according to claim 5, characterized in that: Determining a first preset direction of the rotated tooth model in the three-dimensional space using the point cloud dynamic convolutional neural network model includes: Using the point cloud dynamic convolutional neural network model to identify a first display direction of the rotated tooth model in the three-dimensional space; Classifying the first display direction to obtain a first category to which the first display direction belongs; The first preset direction corresponding to the first category is determined by using a first preset correspondence relationship, wherein the first preset correspondence relationship is used to represent a correspondence relationship between a category and the first preset direction.

7. The method for adjusting a tooth model according to claim 4, characterized in that: The second preset orientation is used to represent at least one of the following: the orientation of the target tooth in the rotated tooth model coincides with the orientation of the first axis, the tooth occlusal surface in the rotated tooth model coincides with the target plane, and the target plane is constructed by the first axis and the third axis.

8. The method for adjusting a tooth model according to claim 4, characterized in that: The method of adjusting the second preset orientation by using the residual neural network model so that the rotated tooth model is adjusted to the first preset orientation includes: Determining a second preset direction of the rotated tooth model in the three-dimensional space using the residual neural network model; The second preset orientation is adjusted using the second preset direction, so that the rotated tooth model is adjusted to the first preset orientation.

9. The method for adjusting a tooth model according to claim 8, characterized in that: Determining a second preset direction of the rotated tooth model in the three-dimensional space by using the residual neural network model includes: Using the residual neural network model to identify a second display direction of the rotated tooth model in the three-dimensional space; classifying the second display direction to obtain a second category to which the second display direction belongs; The second preset direction corresponding to the second category is determined by using a second preset correspondence relationship, wherein the second preset correspondence relationship is used to represent a correspondence relationship between a category and the second preset direction.

10. The method for adjusting a tooth model according to claim 1, characterized in that: Moving the tooth model to a preset position in the three-dimensional space based on the point cloud data includes: Determining a middle position of the point cloud data; Determine the coordinate origin of the tooth model in the three-dimensional space based on the intermediate position; The tooth model is moved to the preset position based on the coordinate origin.

11. A tooth model adjustment device, characterized in that: include: An acquisition module, used for acquiring point cloud data of a tooth model in three-dimensional space; A moving module, used for moving the tooth model to a preset position in the three-dimensional space based on the point cloud data; An adjustment module is used to adjust the tooth model at the preset position to a first preset orientation based on a neural network model.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the method for adjusting the tooth model according to any one of claims 1 to 10 is executed in a processor of a device where the program is controlled.

13. A medical system, characterized in that: include: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors execute the tooth model adjustment method according to any one of claims 1 to 10.

Images