Method and apparatus for registering cbct data and three-dimensional facial scan data
By manually selecting and automatically expanding feature points, combined with FPFH and iterative optimization methods, the rigid transformation and scale variation problems in the registration of CBCT data and 3D facial scan data were solved, achieving high-precision registration results and supporting preoperative communication and plan adjustment for oral implant restoration surgery.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHANGZHOU BOEN ZHONGDING MEDICAL TECH
- Filing Date
- 2022-11-25
- Publication Date
- 2026-06-16
AI Technical Summary
Existing registration methods for CBCT data and 3D facial scan data are difficult to achieve effective registration when dealing with 3D data of different resolutions and containing interference factors, especially in terms of rigid transformation and scale changes, and their reliance on the initial position makes iterative searching difficult.
By manually selecting suggested feature points and automatically expanding the selection of feature points, a coarse registration matrix is calculated by combining FPFH features and the k-nearest neighbor method. A scale variation parameter is introduced, and a fine registration matrix is obtained through iterative optimization. Data preprocessing and standardization are then performed to reduce complexity.
It improves the accuracy and efficiency of registration, provides an intuitive facial view, assists physicians in preoperative communication and surgical plan adjustments, reduces algorithm complexity, and minimizes human interference.
Smart Images
Figure CN115937277B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of medical data processing technology, specifically to a registration method and device for CBCT (Cone beam Computer Tomography) data and three-dimensional facial scan data. Background Technology
[0002] With the emergence and development of digital dentistry, dental implant restoration has become a common procedure. During implant surgery, dentists need to communicate effectively with patients to alleviate pre-operative anxieties and fears, thereby increasing patient trust. Providing a visually appealing view of the improved facial appearance after implant surgery can effectively assist dentists in pre-operative communication and surgical plan adjustments. The registration of CBCT data and 3D facial scan data, as one of the most crucial techniques in visualizing facial results, directly impacts the post-operative facial appearance after procedures such as tooth alignment, implantation, and orthodontics.
[0003] In related technologies, the registration of CBCT data and 3D facial scan data mainly relies on point cloud registration algorithms. However, registration based on CBCT data and 3D facial scan data involves two different 3D data models from different sources and with different resolutions, resulting in varying data volumes. Furthermore, CBCT data contains many external interference factors (such as soft tissue, bone, dental implants, etc.). Therefore, the aforementioned point cloud registration algorithms have the following shortcomings:
[0004] (1) Most algorithms only target rigid transformations and do not incorporate scale variation parameters, making them unsuitable for registering objects with different resolutions.
[0005] (2) It is difficult to select appropriate feature points for different source data containing a lot of interference;
[0006] (3) Relying on a certain initial position, when two 3D data are far apart, it is difficult to find an effective nearest point pair through iteration. Summary of the Invention
[0007] To address the aforementioned technical problems, this invention proposes a registration method for CBCT data and three-dimensional facial scan data.
[0008] The present invention also proposes a registration device for CBCT data and three-dimensional facial scan data.
[0009] The technical solution adopted in this invention is as follows:
[0010] A first aspect of this invention provides a method for registering CBCT data and three-dimensional facial scan data, comprising the following steps: acquiring CBCT data and three-dimensional facial scan data and preprocessing them; guiding the user to manually select suggested feature points based on the CBCT data and three-dimensional facial scan data, and automatically expanding the selection of feature points; performing coarse registration of the CBCT data and three-dimensional facial scan data based on the feature points and calculating the scale change value, and generating a coarse registration similarity transformation matrix based on the scale change value; performing fine registration based on the coarse registration result and obtaining a fine registration rigid transformation matrix; and calculating the registered three-dimensional facial scan data based on the coarse registration similarity transformation matrix and the fine registration rigid transformation matrix.
[0011] The registration method for CBCT data and three-dimensional facial scan data described above in this invention also has the following additional technical features:
[0012] According to one embodiment of the present invention, the system guides and prompts the user to manually select suggested feature points based on CBCT data and three-dimensional facial scan data, and automatically expands the feature point selection. Specifically, the system guides and prompts the user to manually select three suggested feature points in each of the CBCT data and three-dimensional facial scan data to form a suggested feature point pair; takes each manually selected suggested feature point as the center, obtains all feature points within a set radius; recalculates the distance from the feature points within the set radius to the corresponding center feature point, and filters out feature points within a specified range as expanded feature points.
[0013] According to one embodiment of the present invention, coarse registration of CBCT data and three-dimensional facial scan data is performed based on the feature points, and scale change values are calculated. A similarity transformation matrix for coarse registration is generated based on the scale change values. Specifically, the process includes: designating the feature point sets selected from the CBCT data and three-dimensional facial scan data as the target point cloud Q and the source point cloud P, respectively; calculating the surface normal vector and FPFH (Fast Point Feature Histograms) for each feature point using the k-nearest neighbor method; randomly selecting three sampling points from either the target point cloud Q or the source point cloud P based on the distance between the feature points; finding sampling points corresponding to the three sampling points from another feature point set based on the FPFH of the feature points to form a coarse registration point pair; calculating the rigid transformation matrix for coarse registration based on the coarse registration point pair; calculating the variance of the CBCT data and the three-dimensional facial scan data, respectively; calculating the scale change values based on the variances; and generating the similarity transformation matrix for coarse registration based on the rigid transformation matrix and the scale change values.
[0014] According to one embodiment of the present invention, fine registration is performed based on the coarse registration result, and a rigid transformation matrix of fine registration is obtained. Specifically, this includes: obtaining an initial point set for fine registration based on the coarse registration result; determining fine registration point pairs from the initial point set of fine registration based on the nearest distance principle; obtaining an objective function for the rigid transformation matrix of fine registration based on the fine registration point pairs; and iteratively solving the objective function to obtain the rigid transformation matrix of fine registration.
[0015] According to one embodiment of the present invention, acquiring CBCT data and three-dimensional facial scan data and performing preprocessing specifically includes: resampling, segmenting and cropping the CBCT data; resampling and cropping the three-dimensional facial scan data; and standardizing the x, y and z direction data of the CBCT data and the three-dimensional facial scan data.
[0016] A second aspect of the present invention provides a registration device for CBCT data and three-dimensional facial scan data, comprising: a preprocessing module for acquiring and preprocessing CBCT data and three-dimensional facial scan data; a selection module for guiding and prompting the user to manually select suggested feature points based on the CBCT data and three-dimensional facial scan data, and automatically expanding the selection of feature points; a coarse registration module for performing coarse registration of the CBCT data and three-dimensional facial scan data based on the feature points and calculating scale change values, and generating a coarse registration similarity transformation matrix based on the scale change values; a fine registration module for performing fine registration based on the coarse registration result and obtaining a fine registration rigid transformation matrix; and a calculation module for calculating the registered three-dimensional facial scan data based on the coarse registration similarity transformation matrix and the fine registration rigid transformation matrix.
[0017] The CBCT data and three-dimensional facial scan data registration device described above in this invention also has the following additional technical features:
[0018] According to an embodiment of the present invention, the selection module is specifically used to: guide and prompt the user to manually select 3 suggested feature points from CBCT data and 3D facial scan data in sequence to form a suggested feature point pair; obtain all feature points within a set radius with each manually selected suggested feature point as the center; recalculate the distance from the feature points within the set radius to the corresponding center feature point, and filter out feature points within a specified range as expanded feature points.
[0019] According to an embodiment of the present invention, the coarse registration module is specifically used for: denoting the feature point sets selected from the CBCT data and the three-dimensional facial scan data as the target point cloud Q and the source point cloud P, respectively; calculating the surface normal vector and FPFH of each feature point according to the k-nearest neighbor method; randomly selecting three sampling points from either the target point cloud Q or the source point cloud P based on the distance between the feature points; finding sampling points corresponding to the above three sampling points from another feature point set based on the FPFH of the feature points to form a coarse registration point pair; calculating the rigid transformation matrix of the coarse registration based on the coarse registration point pair; calculating the variance of the CBCT data and the three-dimensional facial scan data respectively; calculating the scale change value based on the variance; and generating the similarity transformation matrix of the coarse registration based on the rigid transformation matrix and the scale change value of the coarse registration.
[0020] According to one embodiment of the present invention, the fine registration module is specifically used for: obtaining an initial point set for fine registration based on the coarse registration result; determining fine registration point pairs from the initial point set for fine registration based on the nearest distance principle; obtaining an objective function for the rigid transformation matrix of fine registration based on the fine registration point pairs; and iteratively solving the objective function to obtain the rigid transformation matrix of fine registration.
[0021] According to one embodiment of the present invention, the preprocessing module is used to: resample, segment, and crop the CBCT data; resample and crop the three-dimensional facial scan data; and standardize the x, y, and z direction data of the CBCT data and the three-dimensional facial scan data.
[0022] The beneficial effects of this invention are:
[0023] This invention registers CBCT data and three-dimensional facial scan data to provide an intuitive facial view for oral implant restoration surgery, which can effectively assist physicians in preoperative doctor-patient communication and adjustment of surgical plans.
[0024] To address the non-rigid transformation problem between CBCT data and 3D facial scan data, a scale variation parameter is introduced to simplify the non-rigid transformation problem into a rigid transformation problem, effectively reducing the complexity of the registration algorithm.
[0025] For 3D data models from different sources, a manual-automatic feature point selection method is proposed. A small number of suggested feature points are manually selected, and feature points are automatically expanded according to the distance. This effectively reduces human interference and improves the time utilization of registration.
[0026] To address the issue that registration algorithms rely on a certain initial position, a data standardization method is used to establish data from two different sources in the same coordinate system, reducing the difficulty of algorithm iteration and optimization and improving the accuracy of model registration. Attached Figure Description
[0027] Figure 1 This is a flowchart of a method for registering CBCT data and three-dimensional facial scan data according to an embodiment of the present invention;
[0028] Figure 2 This is a flowchart of coarse registration according to an embodiment of the present invention;
[0029] Figure 3 This is a flowchart of fine registration according to an embodiment of the present invention;
[0030] Figure 4 This is a block diagram of a registration device for CBCT data and three-dimensional facial scan data according to an embodiment of the present invention. Detailed Implementation
[0031] This invention was made by the inventor based on research and understanding of the following issues:
[0032] The scalp model reconstructed from CBCT images contains rich 3D information, allowing for the free extraction of skin and bone information, and also recording the dental condition before and after dental implant restoration surgery. A 3D facial scan model is a 3D scan of the human face, primarily containing 3D geometric coordinates and color texture information, vividly displaying realistic facial features. It has wide applications in orthodontics, orthognathic surgery, and cosmetic restoration. Registration involves adjusting the coordinate systems of different sources of 3D data for the same object, ensuring that the positions of parts belonging to the same structure are consistent within the same coordinate system. Registration of CBCT data and 3D facial scan data is essential for demonstrating the improved facial appearance after dental implant restoration surgery.
[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0034] Figure 1 This is a flowchart of a registration method for CBCT data and three-dimensional facial scan data according to an embodiment of the present invention, as follows: Figure 1 As shown, the method includes the following steps:
[0035] S1: Acquire CBCT data and 3D facial scan data and perform preprocessing.
[0036] Furthermore, CBCT data and 3D facial scan data are acquired and preprocessed, specifically including: resampling, segmenting, and cropping the CBCT data; resampling and cropping the 3D facial scan data; and standardizing the x, y, and z direction data of the CBCT data and 3D facial scan data.
[0037] Specifically, after acquiring oral CBCT data, the spacing values of the x, y, and z coordinates in the three-dimensional data can be obtained, and the spacing value of each coordinate can be modified according to the data volume (here it is uniformly set to 0.5), thereby greatly reducing the number of feature points after the CBCT data is converted into point cloud and improving the smoothness of data viewing.
[0038] The resampled CBCT data contains information about skin, bones, teeth, etc. Therefore, an appropriate Hu value is selected (here set to -599), and the CBCT data is segmented by isosurface extraction to visualize skin-level information and reduce interference from other useless information.
[0039] The segmented data typically contains information such as soft tissue and chewing sticks. To reduce interference from this information, this embodiment first calculates the cross-section in the y-coordinate direction, with a normal vector of (0,-1,0). Then, using the index value of the data's y-coordinate dimension, the world origin coordinates, and the data spacing, the world coordinate value corresponding to that index is calculated. The calculation formula is as follows:
[0040] y world =y index *y spacing +y o ;
[0041] Among them, y index This is the index value of the data in the y-direction dimension, which is chosen to be at position 7 / 12 of the y-direction dimension. spacing It is the spacing value in the y-direction of the data, y o It is the y-coordinate value of the world origin coordinates, y world This is the world coordinate value corresponding to the index value. Therefore, the origin coordinates of the cross-section selected in the y-coordinate direction are (0, y). world (0). The cross-section in the y-coordinate direction can be obtained by combining the normal vector and the selected origin coordinates, thus cropping the CBCT data. Similarly, the cross-section in the z-coordinate direction can be calculated, where the index value is at 1 / 12 of the z-direction dimension of the data, further yielding the cropped CBCT data.
[0042] Therefore, CBCT data preprocessing involves resampling, segmenting, and cropping oral CBCT data, which can reduce external interference factors such as soft tissue, bones, dental sticks, and teeth in oral CBCT data, thereby improving the extraction rate of effective feature points in subsequent automatic feature point expansion processing.
[0043] After acquiring the 3D facial scan data, the data is resampled according to the x, y, and z-axis spacing values selected for CBCT data resampling, ensuring that the distribution of surface feature points in the two datasets is essentially consistent. Using the cross-sectional method obtained in CBCT data preprocessing, cross-sections in the y and z directions of the 3D facial scan data are calculated, with index values at 1 / 2 of the y-axis dimension and 1 / 12 of the z-axis dimension, respectively, further yielding the cropped 3D facial scan data. Thus, 3D facial scan data preprocessing involves resampling and cropping the 3D facial scan data to reduce information such as hair and neck in the 3D facial scan data, constructing data with a structure similar to that of oral CBCT data.
[0044] The standardization of the x, y, and z directions of CBCT data and 3D facial scan data involves standardizing the x, y, and z directions of both data separately to establish the same coordinate system. By proportionally reducing the distance between the two data sets, the iterative optimization difficulty of coarse registration is reduced.
[0045] More specifically, the means in the x, y, and z directions of the data can be calculated separately; further, the means in each coordinate direction of all feature points are subtracted from the coordinates of all feature points to decenter the data; further, the variance of all feature points can be calculated; further, the coordinates of all feature points are divided by the variance to perform scale normalization on the data, resulting in standardized data. This data standardization method effectively solves the problem of location differences between data from different sources, improving the accuracy of registration.
[0046] S2 guides users to manually select suggested feature points based on CBCT data and 3D facial scan data, and automatically expands the selection of feature points.
[0047] Furthermore, according to one embodiment of the present invention, guiding the user to manually select suggested feature points based on CBCT data and three-dimensional facial scan data, and automatically expanding the selection of feature points, specifically includes: guiding the user to manually select 3 suggested feature points in each of the CBCT data and three-dimensional facial scan data to form suggested feature point pairs; taking each manually selected suggested feature point as the center, obtaining all feature points within a set radius; recalculating the distance from the feature points within the set radius to the corresponding center feature point, and filtering out feature points within a specified range as expanded feature points.
[0048] Specifically, three suggested feature point pairs can be manually selected sequentially according to the guidance prompts, such as the center points of the left and right eyes, the center point of the lips, etc. This embodiment only requires the selection of three feature points, and the guidance prompts effectively reduce human error caused by manual feature point selection. Taking each manually selected feature point as the center, all feature points within a set radius (radius of 3cm) are obtained. The distance from the feature points within the radius to the center feature point is recalculated, and feature points within a specified range (between 2cm and 3cm) are selected as expansion feature points to improve the robustness of the registration algorithm.
[0049] S3: Perform coarse registration of CBCT data and 3D facial scan data based on feature points and calculate scale change values. Generate a coarse registration similarity transformation matrix based on the scale change values.
[0050] According to one embodiment of the present invention, such as Figure 2 As shown, coarse registration of CBCT data and 3D facial scan data is performed based on feature points, and scale change values are calculated. A coarse registration similarity transformation matrix is then generated based on these scale change values. Specifically, this includes:
[0051] S301, the feature point sets selected from CBCT data and 3D facial scan data are denoted as target point cloud Q and source point cloud P, respectively. The surface normal vector and FPFH of each feature point are calculated according to the k-nearest neighbor method.
[0052] Specifically, the feature point sets selected from CBCT data and 3D facial scan data can be denoted as target point cloud Q and source point cloud P, respectively. The surface normal vector of each feature point in each point cloud can be calculated according to the k-nearest neighbor method (k is 1 / 2 of the number of feature points). Then, the FPFH feature of each feature point can be extracted using the k-nearest neighbor method (K is generally 2 to 3 times the value of k).
[0053] S302, based on the distance between feature points, randomly select 3 sampling points from any feature point set in the target point cloud Q and the source point cloud P.
[0054] S303, based on the FPFH of the feature points, find the sampling points corresponding to the above 3 sampling points from another feature point set to form a coarse registration point pair.
[0055] Specifically, three sampling points can be randomly selected from the source point cloud P. To ensure that the sampled feature points have different FPFH features, the pairwise distance between the three randomly selected sampling points should, as far as possible, satisfy the set minimum distance threshold d. min (d min (It can be 2cm); then find sampling points in the target point cloud Q that have similar FPFH features to the sampling points in the source point cloud P and put them into a list, and randomly select a sampling point from these similar sampling points as the corresponding coarse registration point pair in the target point cloud Q and the source point cloud P.
[0056] S304, calculate the rigid transformation matrix of coarse registration based on the coarse registration point pairs.
[0057] Specifically, the rigid transformation matrix for registration can be calculated based on three randomly selected coarse registration points; it can also be calculated using the "distance error sum" function. To evaluate the performance of the current rigid transformation matrix, the "distance error sum" function is used. The Huber evaluation function is typically used, and the calculation formula is as follows:
[0058]
[0059] Among them, H(l) i ) represents the distance error, ml is the preset value, and l i Let be the distance difference between corresponding points in the i-th group after the rigid transformation. Further, the above operation can be repeated until a given number of iterations is satisfied. The goal of iterative optimization is to find an optimal set of transformations that minimizes the value of the "distance error sum" function, which is the final coarse registration rigid transformation matrix.
[0060] S305 calculates the variance of CBCT data and 3D facial scan data respectively, and calculates the scale change based on the variance.
[0061] S306, Generate the coarse registration similarity transformation matrix based on the rigid transformation matrix and scale change value of the coarse registration.
[0062] Specifically, the variances of the CBCT data and the 3D facial scan data can be calculated separately, and the square root of the division between the two is the scale change value of the two data. Then, the similarity transformation matrix of the two data can be regenerated by combining the rigid transformation matrix of coarse registration, which can be specifically expressed as:
[0063]
[0064] Where M represents the similarity transformation matrix for coarse registration, s represents the scale change value, R represents the rotation matrix, T represents the translation vector, and [R, T] constitutes the similarity transformation matrix.
[0065] S4. Calculate the registered 3D facial scan data based on the coarse registration similarity transformation matrix and the fine registration rigid transformation matrix.
[0066] Furthermore, according to one embodiment of the present invention, such as Figure 3 As shown, fine registration is performed based on the coarse registration result, and the rigid transformation matrix of the fine registration is obtained, specifically including:
[0067] S401, Obtain the initial point set for fine registration based on the coarse registration result.
[0068] S402, determine the fine registration point pairs from the initial point set of fine registration according to the principle of closest distance.
[0069] Specifically, the coarsely registered point cloud P′ (P being the source point cloud after coordinate transformation based on the similarity transformation matrix) and Q are used as the initial point set for fine registration; for each point P′ in P′... i Find the nearest corresponding point Q in the target point cloud Q. i This serves as the corresponding point in the target point cloud, forming the initial point set for fine registration. Furthermore, a direction vector thresholding method can be used to remove erroneous fine registration point pairs.
[0070] S403, the objective function for obtaining the rigid transformation matrix of the fine registration based on the fine registration point pairs.
[0071] S404, iteratively solve the objective function to obtain the rigid transformation matrix for precise registration.
[0072] Specifically, the objective function can be determined as the evaluation function of the current rigid transformation matrix, and the objective function is defined as follows:
[0073]
[0074] Where E(R,T) represents the objective function of (R,T), E represents the distance error sum, i represents the index of the sampling point, n represents the number of sampling points, R represents the rotation matrix, and T represents the translation vector.
[0075] The objective function can be solved iteratively to minimize the mean square error between the corresponding point sets. If the error of the iteration is less than a given threshold (the threshold is 1e-10) or the number of iterations reaches the maximum value (the maximum iteration value is 50), the iteration ends. Otherwise, the registered point set is updated, and the above steps are repeated until the convergence condition is met to obtain the rigid transformation matrix of fine registration.
[0076] S5. Calculate the registered 3D facial scan data based on the coarse registration similarity transformation matrix and the fine registration rigid transformation matrix.
[0077] Specifically, after obtaining the original 3D facial scan data, the data is standardized. Based on the coarse registration similarity transformation matrix and the fine registration rigid transformation matrix, the facial scan data is regenerated. The regenerated data is then inversely standardized and saved as an OBJ (3D model file format) file.
[0078] In summary, the registration method for CBCT data and 3D facial scan data according to the embodiments of this invention provides an intuitive facial view for oral implant restoration surgery, effectively assisting physicians in preoperative doctor-patient communication and adjustment of surgical plans. To address the non-rigid transformation problem between CBCT data and 3D facial scan data, a scale variation parameter is introduced, simplifying the non-rigid transformation problem into a rigid transformation problem, effectively reducing the complexity of the registration algorithm. For 3D data models from different sources, a manual-automatic feature point selection method is proposed, manually selecting a small number of suggested feature points and automatically expanding the selection based on distance, effectively reducing human interference and improving the time utilization rate of registration. To address the issue of the registration algorithm relying on a certain initial position, a data standardization method is used to establish data from two different sources in the same coordinate system, reducing the difficulty of algorithm iterative optimization and improving the accuracy of model registration.
[0079] Corresponding to the above-described registration method for CBCT data and three-dimensional facial scan data, this invention also proposes a registration device for CBCT data and three-dimensional facial scan data. Since the device embodiments of this invention correspond to the above-described method embodiments, details not disclosed in the device embodiments can be referred to in the above-described method embodiments, and will not be repeated here.
[0080] Figure 4 This is a block diagram of a registration device for CBCT data and three-dimensional facial scan data according to an embodiment of the present invention, as shown below. Figure 4 As shown, the device includes: a preprocessing module 1, a selection module 2, a coarse registration module 3, a fine registration module 4, and a calculation module 5.
[0081] The preprocessing module 1 is used to acquire and preprocess CBCT data and 3D facial scan data; the selection module 2 is used to guide and prompt the user to manually select suggested feature points based on CBCT data and 3D facial scan data, and automatically expand the selection of feature points; the coarse registration module 3 is used to perform coarse registration of CBCT data and 3D facial scan data based on feature points and calculate the scale change value, and generate the coarse registration similarity transformation matrix based on the scale change value; the fine registration module 4 is used to perform fine registration based on the coarse registration result and obtain the fine registration rigid transformation matrix; the calculation module 5 is used to calculate the registered 3D facial scan data based on the coarse registration similarity transformation matrix and the fine registration rigid transformation matrix.
[0082] According to an embodiment of the present invention, the selection module 2 is specifically used to: guide and prompt the user to manually select 3 suggested feature points from CBCT data and 3D facial scan data in sequence to form a suggested feature point pair; take each manually selected suggested feature point as the center to obtain all feature points within a set radius; recalculate the distance from the feature points within the set radius to the corresponding center feature point, and filter out the feature points within the specified range as expanded feature points.
[0083] According to one embodiment of the present invention, the coarse registration module 3 is specifically used for: denoting the feature point sets selected from the CBCT data and the three-dimensional facial scan data as the target point cloud Q and the source point cloud P, respectively; calculating the surface normal vector and FPFH of each feature point according to the k-nearest neighbor method; randomly selecting three sampling points from either the target point cloud Q or the source point cloud P based on the distance between the feature points; finding sampling points corresponding to the above three sampling points from another feature point set based on the FPFH of the feature points to form a coarse registration point pair; calculating the rigid transformation matrix of coarse registration based on the coarse registration point pair; calculating the variance of the CBCT data and the three-dimensional facial scan data respectively; calculating the scale change value based on the variance; and generating the similarity transformation matrix of coarse registration based on the rigid transformation matrix and the scale change value of coarse registration.
[0084] According to one embodiment of the present invention, the fine registration module 4 is specifically used for: obtaining an initial point set for fine registration based on the coarse registration result; determining fine registration point pairs from the initial point set for fine registration based on the nearest distance principle; obtaining the objective function of the rigid transformation matrix of fine registration based on the fine registration point pairs; and iteratively solving the objective function to obtain the rigid transformation matrix of fine registration.
[0085] According to one embodiment of the present invention, the preprocessing module 1 is specifically used for: resampling, segmenting and cropping CBCT data; resampling and cropping three-dimensional facial scan data; and standardizing the x, y and z direction data of CBCT data and three-dimensional facial scan data.
[0086] The registration device for CBCT data and 3D facial scan data according to embodiments of the present invention registers CBCT data and 3D facial scan data, providing an intuitive facial view for oral implant restoration surgery, which can effectively assist physicians in preoperative doctor-patient communication and adjustment of surgical plans. Addressing the non-rigid transformation problem of CBCT data and 3D facial scan data, a scale variation parameter is introduced to simplify the non-rigid transformation problem into a rigid transformation problem, effectively reducing the complexity of the registration algorithm. For 3D data models from different sources, a manual-automatic feature point selection method is proposed, manually selecting a small number of suggested feature points and automatically expanding the selection based on distance, effectively reducing human interference and improving the time utilization rate of registration. To address the problem that the registration algorithm relies on a certain initial position, a data standardization method is used to establish data from two different sources in the same coordinate system, reducing the difficulty of algorithm iterative optimization and improving the accuracy of model registration.
[0087] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.
[0088] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.
[0089] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of the invention pertain.
[0090] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.
[0091] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0092] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.
[0093] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.
[0094] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.
Claims
1. A method for registering CBCT data and three-dimensional facial scan data, characterized in that, Includes the following steps: Acquire CBCT data and 3D facial scan data and perform preprocessing; The system guides users to manually select suggested feature points based on CBCT data and 3D facial scan data, and automatically expands the selection of feature points. Based on the feature points, coarse registration of CBCT data and 3D facial scan data is performed and scale change values are calculated. Based on the scale change values, a coarse registration similarity transformation matrix is generated. Based on the coarse registration results, fine registration is performed, and the rigid transformation matrix of the fine registration is obtained; The registered 3D facial scan data are calculated based on the coarse registration similarity transformation matrix and the fine registration rigid transformation matrix. Based on the feature points, coarse registration is performed between CBCT data and 3D facial scan data, and scale change values are calculated. A coarse registration similarity transformation matrix is then generated based on these scale change values, specifically including: The feature point sets selected from CBCT data and 3D facial scan data are denoted as target point cloud Q and source point cloud P, respectively. The surface normal vector and FPFH of each feature point are calculated according to the k-nearest neighbor method. Based on the distance between feature points, three sampling points are randomly selected from any set of feature points in the target point cloud Q and the source point cloud P; Based on the FPFH of the feature points, find the sampling points corresponding to the above 3 sampling points from another feature point set to form a coarse registration point pair; Calculate the rigid transformation matrix of coarse registration based on the coarse registration point pairs; Calculate the variances of CBCT data and 3D facial scan data respectively, and calculate the scale variation value based on the variances; The coarse registration similarity transformation matrix is generated based on the rigid transformation matrix and scale change value of the coarse registration.
2. The registration method for CBCT data and three-dimensional facial scan data according to claim 1, characterized in that, The system guides users to manually select suggested feature points based on CBCT data and 3D facial scan data, and automatically expands the selection of feature points, specifically including: The system guides users to manually select three suggested feature points from both the CBCT data and the 3D facial scan data to form a suggested feature point pair. Using each manually selected suggested feature point as the center, obtain all feature points within a set radius; The distance from the feature point within the set radius to the corresponding center feature point is recalculated, and feature points within the specified range are selected as expanded feature points.
3. The registration method for CBCT data and three-dimensional facial scan data according to claim 1, characterized in that, Based on the coarse registration results, fine registration is performed, and the rigid transformation matrix of the fine registration is obtained, specifically including: Obtain the initial point set for fine registration based on the coarse registration results; Based on the principle of closest distance, fine registration point pairs are determined from the initial point set of the fine registration. The objective function for obtaining the rigid transformation matrix of the fine registration based on the fine registration point pair; The objective function is solved iteratively to obtain the rigid transformation matrix of the fine registration.
4. The registration method for CBCT data and three-dimensional facial scan data according to claim 1, characterized in that, Acquire CBCT data and 3D facial scan data and perform preprocessing, specifically including: The CBCT data is resampled, segmented, and cropped. The three-dimensional facial scan data is resampled and cropped; The x, y, and z direction data of the CBCT data and the three-dimensional facial scan data are standardized.
5. A registration device for CBCT data and three-dimensional facial scan data, characterized in that, include: The preprocessing module is used to acquire CBCT data and three-dimensional facial scan data and perform preprocessing. The selection module guides and prompts the user to manually select suggested feature points based on CBCT data and 3D facial scan data, and automatically expands the selection of feature points. A coarse registration module is used to perform coarse registration of CBCT data and three-dimensional facial scan data based on the feature points and calculate the scale change value, and generate a coarse registration similarity transformation matrix based on the scale change value. A fine registration module is used to perform fine registration based on the coarse registration result and obtain the rigid transformation matrix of the fine registration. The calculation module is used to calculate the registered three-dimensional facial scan data based on the similarity transformation matrix of the coarse registration and the rigid transformation matrix of the fine registration. The coarse registration module is specifically used for: The feature point sets selected from CBCT data and 3D facial scan data are denoted as target point cloud Q and source point cloud P, respectively. The surface normal vector and FPFH of each feature point are calculated according to the k-nearest neighbor method. Based on the distance between feature points, three sampling points are randomly selected from any set of feature points in the target point cloud Q and the source point cloud P; Based on the FPFH of the feature points, find the sampling points corresponding to the above 3 sampling points from another feature point set to form a coarse registration point pair; Calculate the rigid transformation matrix of coarse registration based on the coarse registration point pairs; Calculate the variances of CBCT data and 3D facial scan data respectively, and calculate the scale variation value based on the variances; The coarse registration similarity transformation matrix is generated based on the rigid transformation matrix and scale change value of the coarse registration.
6. The registration device for CBCT data and three-dimensional facial scan data according to claim 5, characterized in that, The selection module is specifically used for: The system guides users to manually select three suggested feature points from both the CBCT data and the 3D facial scan data to form a suggested feature point pair. Using each manually selected suggested feature point as the center, obtain all feature points within a set radius; The distance from the feature point within the set radius to the corresponding center feature point is recalculated, and feature points within the specified range are selected as expanded feature points.
7. The registration device for CBCT data and three-dimensional facial scan data according to claim 5, characterized in that, The precise registration module is specifically used for: Obtain the initial point set for fine registration based on the coarse registration results; Based on the principle of closest distance, fine registration point pairs are determined from the initial point set of the fine registration. The objective function for obtaining the rigid transformation matrix of the fine registration based on the fine registration point pair; The objective function is solved iteratively to obtain the rigid transformation matrix of the fine registration.
8. The registration device for CBCT data and three-dimensional facial scan data according to claim 5, characterized in that, The preprocessing module is specifically used for: The CBCT data is resampled, segmented, and cropped. The three-dimensional facial scan data is resampled and cropped; The x, y, and z direction data of the CBCT data and the three-dimensional facial scan data are standardized.