Three-dimensional image-guided correction planning method, system, device and medium

Abstract

The present invention relates to a three-dimensional image-guided correction planning method, system, device and medium. The method includes: obtaining a three-dimensional CT image and a two-dimensional X-ray image of a patient's oral cavity, integrating and processing the three-dimensional CT image and the two-dimensional X-ray image to obtain a corresponding comprehensive image set; extracting features from the comprehensive image set to obtain a tooth image and a jaw image, performing three-dimensional construction based on the tooth image and the jaw image to obtain an initial three-dimensional model; aligning and fusing the tooth image and the jaw image to obtain a corresponding oral fusion image; registering the oral fusion image with the initial three-dimensional model based on a preset registration method to obtain a corresponding oral model; performing trajectory analysis on the patient's oral cavity to obtain an initial trajectory path, iteratively optimizing the initial trajectory path to obtain an optimized trajectory scheme; simulating the trajectory of the oral model based on the optimized trajectory scheme to obtain a corresponding simulated trajectory result. The present invention can achieve refined management of oral details.

Description

Technical Field

[0001] The present invention relates to the technical field of digital health, and in particular to a three-dimensional image-guided correction planning method, system, device and medium. Background Art

[0002] As an innovative oral correction technology, the three-dimensional image-guided correction planning method has been increasingly widely used in modern oral medicine. With the continuous improvement of patients' demands for oral health and aesthetics, how to accurately plan and perform oral correction has become one of the research focuses. Existing correction planning methods usually rely on a single type of image data. The limitations of this single image data may lead to inaccurate oral trajectory planning, thereby affecting the final trajectory effect and patient satisfaction. In addition, due to the incompleteness of a single image data, the initial trajectory path may have a large deviation, requiring multiple manual adjustments and optimizations. Summary of the invention

[0003] The main purpose of the present invention is to provide an air source heat pump aggregation prediction and peak-shaving task instruction coordinated control method and system, which can realize the refined management of oral details, and formulate a more accurate correction trajectory path to avoid unnecessary adjustments.

[0004] To achieve the above object, the present invention provides a three-dimensional image-guided correction planning method, which is characterized by comprising:

[0005] Acquire a three-dimensional CT image and a two-dimensional X-ray image of the patient's oral cavity, and integrate the three-dimensional CT image and the two-dimensional X-ray image to obtain a corresponding comprehensive image set;

[0006] Extracting features from the comprehensive image set to obtain tooth images and jaw images, and performing three-dimensional construction based on the tooth images and the jaw images to obtain an initial three-dimensional model;

[0007] Aligning and fusing the tooth image and the jaw image to obtain a corresponding oral fusion image;

[0008] Based on a preset registration method, the oral fusion image is registered with the initial three-dimensional model to obtain a corresponding oral model;

[0009] Performing trajectory analysis on the patient's oral cavity according to the oral cavity model to obtain an initial trajectory path, and iteratively optimizing the initial trajectory path to obtain an optimized trajectory solution;

[0010] The trajectory of the oral model is simulated based on the optimized trajectory solution to obtain a corresponding simulation trajectory result.

[0011] Furthermore, the three-dimensional CT image and the two-dimensional X-ray image are integrated to obtain a corresponding comprehensive image set, including:

[0012] Performing distortion correction processing on the three-dimensional CT image to obtain a corrected three-dimensional CT image, and performing distortion correction on the two-dimensional X-ray image to obtain a corrected two-dimensional X-ray image;

[0013] Align and match the corrected three-dimensional CT image and the corrected two-dimensional X-ray image to obtain a preliminary aligned image;

[0014] Perform multi-resolution pyramid decomposition on the preliminary registered images to obtain image levels of different scales, and integrate all image levels to obtain a multi-scale image set;

[0015] Perform wavelet transform on the multi-scale image set and extract the features of different frequency components to obtain the wavelet coefficient matrix, perform feature fusion on the wavelet coefficient matrix to obtain the fused feature representation;

[0016] The fused feature representation is inversely transformed and reconstructed to obtain the comprehensive image set.

[0017] Furthermore, the feature extraction of the comprehensive image set to obtain a tooth image and a jaw image, and the three-dimensional construction based on the tooth image and the jaw image to obtain an initial three-dimensional model include:

[0018] Performing image segmentation processing on the comprehensive image set to obtain the segmented tooth image and the jaw image;

[0019] Extracting tooth contour features from the tooth image according to a preset edge detection algorithm to obtain tooth contour information;

[0020] Extracting the features of the jaw bone line from the jaw image according to a preset bone line extraction algorithm to obtain jaw bone line information;

[0021] Perform three-dimensional reconstruction based on the tooth contour information and jaw bone line information to obtain a preliminary reconstruction model;

[0022] The model surface of the preliminary reconstructed model is smoothed to obtain the initial three-dimensional model.

[0023] Furthermore, aligning and fusing the tooth image and the jaw image to obtain a corresponding oral fusion image includes:

[0024] Performing single segmentation processing on the tooth image according to the tooth contour information to obtain corresponding single tooth contour information;

[0025] Performing edge detection on the jaw image according to the jaw skeleton line information to obtain jaw boundary information;

[0026] The relationship between the single tooth contour information and the jaw boundary information is constructed to obtain the corresponding tooth-jaw correspondence matrix;

[0027] Performing rigid registration on the tooth image and the jaw image according to a tooth-jaw correspondence matrix to obtain a preliminary alignment result;

[0028] Perform non-rigid deformation processing on the preliminary alignment result to obtain a refined alignment result;

[0029] The tooth image and the jaw image are interpolated and fused according to the fine alignment result to obtain the oral fusion image.

[0030] Furthermore, the oral fusion image is registered with the initial three-dimensional model based on a preset registration method to obtain a corresponding oral model, including:

[0031] Extracting feature points from the oral fusion image to obtain a first feature point set;

[0032] Extracting feature points from the initial three-dimensional model to obtain a second feature point set;

[0033] Performing a rough registration on the oral fusion image and the initial three-dimensional model according to the first feature point set and the second feature point set to obtain a rough registration result;

[0034] Based on the coarse registration result, fine registration is performed on the oral fusion image and the initial three-dimensional model to obtain a fine registration result;

[0035] According to the fine registration result, the oral fusion image and the initial three-dimensional model are fused to obtain a fusion model;

[0036] Perform bounding box analysis on the fusion model to obtain a model bounding box, and construct a three-dimensional regular grid based on the model bounding box to obtain a model voxel space;

[0037] Densely sample the surface of the fusion model to obtain surface sampling points;

[0038] Detect whether each voxel in the model voxel space contains a surface sampling point, and mark the voxel containing the surface sampling point as occupied;

[0039] Fill the model voxel space from the boundary to the inside according to a preset filling rule, and mark the internal voxels of the model voxel space as occupied;

[0040] Encode all occupied voxels according to the model voxel space to obtain a voxelized model;

[0041] The voxelized model is subjected to surface reconstruction to obtain the oral cavity model.

[0042] Furthermore, performing trajectory analysis on the patient's oral cavity according to the oral cavity model to obtain an initial trajectory path, and iteratively optimizing the initial trajectory path to obtain an optimized trajectory solution includes:

[0043] Segmenting the oral cavity model to obtain tooth units and jaw units, and performing position analysis on the tooth units and jaw units to obtain a relative position relationship between the tooth units and jaw units;

[0044] Classify and label the tooth units according to their relative positional relationships to obtain tooth type information;

[0045] Analyze and process tooth type information based on preset tooth arrangement rules to obtain the ideal arrangement position;

[0046] Calculating the displacement vector of each tooth unit according to the ideal arrangement position and the current position of the tooth unit, and comprehensively processing the displacement vectors of all tooth units to obtain the initial trajectory path;

[0047] Performing three-dimensional decomposition on each displacement vector in the initial trajectory path to obtain a corresponding three-dimensional component vector, and performing force analysis on the three-dimensional component vector to obtain corresponding force data;

[0048] Perform vector adjustment on the three-dimensional component vector according to the force data to obtain the optimized component vector;

[0049] All optimized component vectors are synthesized to obtain an optimized displacement vector;

[0050] The initial trajectory path is iteratively optimized based on the optimized displacement vector to obtain the optimized trajectory solution.

[0051] Furthermore, simulating the trajectory of the oral model based on the optimized trajectory scheme to obtain a corresponding simulation trajectory result includes:

[0052] Segmenting the oral cavity model according to the optimization trajectory scheme to obtain a plurality of oral cavity sub-models;

[0053] Perform force analysis on each oral cavity model to obtain the corresponding force distribution diagram;

[0054] Based on the force distribution diagram, each oral cavity sub-model is deformed by simulation to obtain a deformed sub-model, and all deformed sub-models are combined to obtain a trajectory oral cavity overall model after the simulation trajectory;

[0055] Compare and analyze the trajectory oral cavity overall model with the preset standard oral cavity model to obtain deviation data, and fine-tune the trajectory oral cavity overall model according to the deviation data to obtain a fine-tuned oral cavity overall model;

[0056] The fine-tuned oral cavity overall model is subjected to three-dimensional rendering processing to obtain the simulation trajectory result.

[0057] The present invention further provides a three-dimensional image-guided correction planning system, which is applied to any one of the three-dimensional image-guided correction planning methods mentioned above, comprising:

[0058] An acquisition module, the acquisition module is used to obtain a three-dimensional CT image and a two-dimensional X-ray image of the patient's oral cavity, and integrate the three-dimensional CT image and the two-dimensional X-ray image to obtain a corresponding comprehensive image set;

[0059] An analysis module, the analysis module is used to extract features from the comprehensive image set to obtain a tooth image and a jaw image, and to perform a three-dimensional construction based on the tooth image and the jaw image to obtain an initial three-dimensional model;

[0060] An association module, the association module is used to align and fuse the tooth image and the jaw image to obtain a corresponding oral fusion image;

[0061] A processing module, the processing module is used to register the oral fusion image with the initial three-dimensional model based on a preset registration method to obtain a corresponding oral model;

[0062] A control module is used to perform trajectory analysis on the patient's oral cavity according to the oral cavity model to obtain an initial trajectory path, and iteratively optimize the initial trajectory path to obtain an optimized trajectory solution.

[0063] An execution module is used to simulate the trajectory of the oral model based on the optimized trajectory scheme to obtain a corresponding simulation trajectory result.

[0064] The present invention also provides a three-dimensional image-guided correction planning device, comprising:

[0065] Memory, used to store programs;

[0066] The processor is used to execute the program to implement each step of any one of the above-mentioned three-dimensional image-guided correction planning methods.

[0067] The present invention provides a three-dimensional image-guided correction planning method, system, device and medium, which have the following beneficial effects:

[0068] By integrating 3D CT images and 2D X-ray image data, the features of teeth and jaws can be extracted more accurately, thereby improving the accuracy of oral model construction and providing a more reliable basis for trajectory planning. By aligning and fusing image data and performing registration analysis based on fused images, refined management of oral details can be achieved, which helps to develop more accurate trajectory paths and avoid unnecessary adjustments. The initial path is iteratively optimized based on the optimized trajectory plan to ensure that the trajectory process can be carried out efficiently at different stages, reducing unnecessary errors and improving the efficiency of the entire trajectory process. By comprehensively analyzing the oral model and trajectory path data, a more reasonable trajectory strategy is formulated, and the trajectory plan is optimized through comprehensive adjustments, thereby effectively improving the trajectory effect and reducing treatment time and patient discomfort. And by considering the impact of individual differences on the trajectory plan, the trajectory strategy can be flexibly adjusted according to the oral characteristics and demand changes of different patients, making the trajectory process more adaptable to diverse demand scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 It is a flow chart of a three-dimensional image-guided correction planning method provided by the present invention;

[0070] Figure 2 It is a structural diagram of a three-dimensional image-guided correction planning system provided by the present invention;

[0071] Figure 3 This is a structural diagram of a three-dimensional image-guided correction planning device provided by the present invention.

[0072] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0073] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0074] The present invention is further described below in conjunction with the accompanying drawings and specific implementation methods.

[0075] Reference Figure 1 As shown, the present invention provides a three-dimensional image-guided correction planning method, comprising:

[0076] Step S1: obtaining a three-dimensional CT image and a two-dimensional X-ray image of the patient's oral cavity, integrating the three-dimensional CT image and the two-dimensional X-ray image to obtain a corresponding comprehensive image set;

[0077] Step S2: extracting features from the comprehensive image set to obtain tooth images and jaw images, and performing three-dimensional construction based on the tooth images and jaw images to obtain an initial three-dimensional model;

[0078] Step S3: aligning and fusing the tooth image and the jaw image to obtain a corresponding oral fusion image;

[0079] Step S4: registering the oral fusion image with the initial three-dimensional model based on a preset registration method to obtain a corresponding oral model;

[0080] Step S5: performing trajectory analysis on the patient's oral cavity according to the oral cavity model to obtain an initial trajectory path, and iteratively optimizing the initial trajectory path to obtain an optimized trajectory solution;

[0081] Step S6: Simulate the trajectory of the oral model based on the optimized trajectory solution to obtain the corresponding simulation trajectory result.

[0082] Based on the above steps, the detailed process is as follows:

[0083] Step S1: Obtain detailed three-dimensional CT image data of the patient's oral cavity by reading the CT scanning program. The data includes the full view of the oral cavity structure such as teeth, gums and jaws.

[0084] By reading the X-ray machine, two-dimensional X-ray images of the oral cavity are obtained, especially the side and front images, ensuring that the contrast and clarity of the X-ray images are sufficient to show the details of the teeth and jaws.

[0085] De-noising the acquired 3D CT images, using non-local mean filtering or other image denoising algorithms to reduce the noise in the images. Contrast enhancement of 2D X-ray images, using histogram equalization or adaptive contrast enhancement technology to improve the image detail display.

[0086] Distortion correction is performed on CT images and X-ray images respectively, and the distortion caused by the movement of the imaging device or the patient is corrected using the geometric transformation method.

[0087] The preprocessed three-dimensional CT images and two-dimensional X-ray images are registered, and the two modal images are aligned using a registration algorithm based on feature point matching (such as SIFT, SURF, etc.).

[0088] Use multimodal image fusion technology (such as mutual information-based fusion algorithm) to integrate the registered images to obtain a comprehensive image set.

[0089] Step S2: Apply an image segmentation algorithm (such as an algorithm based on region growing, threshold segmentation or active contour model) to the comprehensive image set to extract images of the tooth region and the jaw region respectively.

[0090] Use edge detection algorithms (such as Canny edge detection) to process tooth images and extract tooth contour features. Use skeleton line extraction algorithms (such as morphological skeleton extraction) to process jaw images and extract jaw skeleton line features.

[0091] Based on the extracted tooth and jaw images, a 3D reconstruction algorithm (such as the Marching Cubes algorithm based on volume rendering) is used to construct a preliminary 3D model of the teeth and jaw. The preliminary 3D model is meshed and optimized, and the model surface is optimized using mesh simplification and mesh smoothing algorithms (such as Laplacian smoothing) to reduce the number of polygons and improve surface smoothness.

[0092] According to the optimized three-dimensional mesh model, the texture mapping technology is used to perform texture mapping on the model surface to obtain an initial three-dimensional model with real surface features.

[0093] Step S3: Apply feature point extraction algorithm to the tooth image and jaw image in the preliminary 3D model to extract key feature point sets of the tooth and jaw, respectively. Use the iterative closest point (ICP) algorithm to perform rigid registration of the tooth image and jaw image based on the extracted feature point set to obtain a preliminary alignment result.

[0094] Based on the preliminary alignment results, a non-rigid registration algorithm (such as a thin plate spline deformation algorithm) is used to finely align the tooth image and the jaw image to obtain a fine alignment result.

[0095] According to the fine alignment results, the tooth image and the jaw image are interpolated and fused, and linear interpolation or more advanced interpolation methods are used to obtain the final oral fusion image.

[0096] Step S4: Preprocess the oral fusion image and the initial 3D model, such as denoising, normalization and coordinate system one, so that the registration algorithm can work effectively. Extract key feature points from the oral fusion image and the initial 3D model, and use feature point matching algorithms (such as SIFT, SURF, etc.) for preliminary matching.

[0097] Initialize the registration parameters, perform rigid registration based on the preliminary matched feature points, and adjust the translation and rotation parameters so that the oral fusion image is roughly aligned with the initial 3D model.

[0098] A non-rigid registration method is used for fine registration, and a combination of global deformation and local deformation is used to further adjust the image and model to make them highly consistent in spatial position and morphology.

[0099] The registered oral model is checked to ensure the spatial consistency of all key structures (such as teeth and jaws). If registration errors are found, local optimization methods are used to correct them to obtain an oral model with high accuracy and consistency.

[0100] Step S5: Perform trajectory analysis based on the teeth and jaw data in the oral model to identify key problems in the patient's oral cavity, such as tooth misalignment, excessive or insufficient tooth gaps, and jaw asymmetry.

[0101] Based on the trajectory goals (such as aligning teeth, adjusting occlusion, etc.), a preliminary trajectory path is generated.

[0102] The preliminary trajectory path is optimized multiple times iteratively, with each optimization making adjustments and improvements based on the results of the previous step.

[0103] The simulation technology is used to simulate the trajectory effect of each iteration, predict the oral model after the trajectory, and evaluate its gap with the target state.

[0104] According to the simulation results, the parameters and strategies of the optimization algorithm are adjusted to gradually approach the optimal trajectory solution.

[0105] After multiple iterations of optimization, a trajectory plan with the best overall effect is determined.

[0106] The final trajectory plan is compared with the oral model to verify that it meets the expected requirements in all aspects (such as function, appearance, stability, etc.), and finally a detailed optimized trajectory plan is generated.

[0107] Step S6: simulating the trajectory of the oral model based on the optimized trajectory scheme, where the simulation trajectory includes modules such as mechanical simulation and tissue response simulation.

[0108] According to the steps and plans of the optimization trajectory scheme, the oral model is gradually adjusted. Each step of adjustment requires precise displacement and rotation of the teeth and jaws. After each adjustment step, the model data of the current state is recorded and the gap with the target state is evaluated.

[0109] During the simulation trajectory, mechanical analysis was performed to evaluate the stress and strain of the teeth and jaws under the action of trajectory forces.

[0110] After completing all trajectory steps, the final simulation trajectory results are evaluated to check whether the oral model after the trajectory meets the expected goals, including tooth arrangement, occlusal relationship, jaw symmetry, etc.

[0111] Generate a comparative analysis report before and after the trajectory, showing each key step and the final effect of the trajectory process. Conduct feedback analysis based on the simulation trajectory results. If it is found that the simulation results are significantly different from expectations, readjust the optimized trajectory plan and simulate again to obtain the final simulation trajectory results.

[0112] The present invention provides a three-dimensional image-guided correction planning method, which can more accurately extract the features of teeth and jaws by integrating three-dimensional CT images and two-dimensional X-ray image data, thereby improving the construction accuracy of the oral model and providing a more reliable basis for trajectory planning. By aligning and fusing the image data, and performing registration analysis based on the fused image, the refined management of oral details is achieved, which helps to formulate a more accurate trajectory path and avoid unnecessary adjustments. The initial path is iteratively optimized based on the optimized trajectory scheme to ensure that the trajectory process can be carried out efficiently at different stages, reduce unnecessary errors, and improve the efficiency of the entire trajectory process. By comprehensively analyzing the oral model and trajectory path data, a more reasonable trajectory strategy is formulated, and the trajectory scheme is optimized through comprehensive adjustment, thereby effectively improving the trajectory effect and reducing treatment time and patient discomfort. And by considering the impact of individual differences on the trajectory scheme, the trajectory strategy can be flexibly adjusted according to the oral characteristics and demand changes of different patients, so that the trajectory process is more adaptable to diverse demand scenarios.

[0113] In one embodiment, the three-dimensional CT image and the two-dimensional X-ray image are integrated and processed to obtain a corresponding comprehensive image set, including:

[0114] The 3D CT image generated by the CT scanner scanning the target area is obtained. The geometric deformation and intensity unevenness generated during the shooting process are eliminated through the geometric correction algorithm and the intensity balance algorithm to obtain the corrected 3D CT image.

[0115] A two-dimensional X-ray image is obtained by scanning the same target area with an X-ray machine. The distortion of the two-dimensional X-ray image is corrected by applying image correction techniques, such as nonlinear transformation and histogram equalization, to obtain a corrected two-dimensional X-ray image.

[0116] Align and match the corrected 3D CT image and the corrected 2D X-ray image. Detect and match the key feature points of the two images through feature point extraction and matching algorithms, such as SIFT (Scale Invariant Feature Transform) or SURF (Speeded Up Robust Features). Use a rigid or non-rigid registration algorithm to spatially align the 3D CT image and the 2D X-ray image based on the detected key feature points to obtain a preliminary aligned image.

[0117] The preliminary aligned images are decomposed into multi-resolution pyramids, which includes downsampling the preliminary aligned images layer by layer to generate a series of image levels with different resolutions. Each level represents the details and structural information of the image at different scales.

[0118] All image levels are integrated to form a multi-scale image set, which contains image information of the target area at different resolutions.

[0119] By performing wavelet transform on the multi-scale image set and extracting the wavelet coefficient matrix, the characteristics of the image at different frequency components are obtained.

[0120] The wavelet coefficient matrix is ​​subjected to feature fusion, and the features of different frequency components are processed comprehensively to obtain a fused feature representation.

[0121] The fused feature representation is inversely transformed to restore the fused features to image form. Through inverse wavelet transform, a comprehensive image set containing all feature information is reconstructed. In the reconstruction process, interpolation algorithms and resampling techniques are used to ensure the spatial resolution and quality of the image.

[0122] This embodiment improves the accuracy of the image and reduces the error caused by distortion during the acquisition process by correcting the distortion of the three-dimensional CT image and the two-dimensional X-ray image. The corrected images are aligned and matched to ensure the spatial consistency of the multi-source image data and improve the basic accuracy of image fusion. Through multi-resolution pyramid decomposition and multi-scale image set integration, this method can capture image details and global structures at different scales, enhancing the comprehensiveness and meticulousness of image analysis. Wavelet transform and feature extraction further extract the multi-frequency components of the image to ensure the comprehensiveness and richness of the image features. Through feature fusion and inverse wavelet transform, the generated comprehensive image set not only retains the key features of the original image, but also eliminates redundant information, thereby improving the quality and practicality of the image.

[0123] In one embodiment, feature extraction is performed on the comprehensive image set to obtain tooth images and jaw images, and three-dimensional construction is performed based on the tooth images and jaw images to obtain an initial three-dimensional model, including:

[0124] The comprehensive image set is segmented to separate the tooth image and the jaw image. The image segmentation techniques include threshold segmentation, region growing, graph cut algorithm, etc.

[0125] Based on the obtained tooth image, the preset edge detection algorithm is used to extract the tooth contour features. The edge detection algorithm includes Canny edge detection, Sobel operator, Laplacian operator, etc.

[0126] The jaw bone image obtained by segmentation is subjected to jaw bone line feature extraction using a preset skeleton line extraction algorithm, and the skeleton line extraction algorithm uses a gradient-based skeleton extraction method or a morphology-based skeleton extraction method.

[0127] Based on the extracted tooth contour information and jaw bone line information, three-dimensional reconstruction is performed to obtain a preliminary reconstructed model. The three-dimensional reconstruction uses a surface reconstruction algorithm, such as the Marching Cubes algorithm, Poisson surface reconstruction, etc.

[0128] The surface of the preliminary reconstructed model is smoothed by a preset smoothing algorithm (Laplacian smoothing) to obtain an initial three-dimensional model.

[0129] This embodiment can obtain tooth images and jaw images respectively through image segmentation processing and feature extraction, thereby achieving more precise separation and identification. This not only improves the accuracy of data processing, but also provides a reliable basis for subsequent three-dimensional reconstruction. Using the preset edge detection algorithm and skeleton line extraction algorithm, the tooth contour and jaw skeleton line information can be efficiently obtained to ensure the accuracy of three-dimensional reconstruction. By performing surface smoothing on the preliminary reconstructed model, the aesthetics and realism of the model are further improved, and the errors in the subsequent tracker design are reduced. The accuracy of the three-dimensional model provides a scientific basis for the formulation of the trajectory plan and reduces the risk and uncertainty in the trajectory process.

[0130] In one embodiment, the tooth image and the jaw image are aligned and fused to obtain a corresponding oral fusion image, including:

[0131] The tooth image is analyzed according to the tooth contour information, each tooth is segmented from the overall tooth image, and the contour information of each single tooth is obtained.

[0132] The edge detection of the jaw image is performed based on the jaw bone line information, and the boundary information of the jaw is extracted by using a gradient-based edge detection method.

[0133] Based on the obtained single tooth contour information and jaw boundary information, a tooth-jaw correspondence matrix is ​​constructed, and the jaw area corresponding to each tooth is calculated. The tooth-jaw correspondence matrix reflects the spatial relationship between teeth and jaws.

[0134] The tooth image and the jaw image are rigidly registered according to the tooth-jaw correspondence matrix to obtain a preliminary alignment result. The purpose of rigid registration is to align the tooth image and the jaw image in space preliminarily through rotation and translation operations.

[0135] Based on the preliminary alignment results, the images are subjected to non-rigid deformation processing to obtain fine alignment results. The purpose of non-rigid deformation processing is to further adjust the tooth images and jaw images to make them more consistent in local details.

[0136] Based on the fine alignment results, the tooth image and the jaw image are interpolated and fused by trilinear interpolation to obtain an oral fusion image. The purpose of interpolation fusion is to fuse the tooth image and the jaw image based on the fine alignment results to generate a unified oral image. Interpolation fusion can also use nearest neighbor interpolation or image fusion algorithm-based methods.

[0137] This embodiment ensures that the spatial relationship between teeth and jaws is accurately expressed through multi-step fine operations such as segmentation, edge detection, rigid registration, and non-rigid deformation processing, thereby generating high-quality oral fusion images. Compared with the traditional single image processing method, this method can more comprehensively display the internal structure of the oral cavity, help doctors formulate more accurate trajectory plans, and improve the treatment effect and patient comfort. In addition, through interpolation fusion technology, the visual effect of the image is further improved, allowing doctors to observe and judge more intuitively when planning and implementing trajectory treatments, thereby improving the convenience and efficiency of clinical operations. In short, this method not only improves the accuracy and quality of image processing, but also brings significant clinical application value to the field of oral trajectory.

[0138] In one embodiment, the oral fusion image is registered with the initial three-dimensional model based on a preset registration method to obtain a corresponding oral model, including:

[0139] Feature points are extracted from the oral fusion image to obtain a first feature point set. The feature points of the first feature point set are key points in the image, such as the edges of teeth or other anatomical feature points. Feature points are extracted from the initial three-dimensional model to obtain a second feature point set, and the feature points of the second feature point set correspond to the feature points of the first feature point set.

[0140] Based on the first feature point set and the second feature point set, the oral fusion image and the initial three-dimensional model are roughly registered. The purpose of the rough registration is to roughly align the two data sets to provide a good initial position. In the rough registration process, the iterative closest point algorithm is used for analysis to obtain the rough registration result.

[0141] According to the coarse registration result, the oral fusion image and the initial 3D model are finely registered. Fine registration is based on an optimization algorithm, such as the least square method or other optimization methods. Through fine registration, a more accurate fine registration result is obtained.

[0142] According to the result of fine registration and the weighted average method, the oral fusion image and the initial three-dimensional model are fused to obtain a fusion model. The fusion process includes combining the two data sets together while retaining the advantages of each.

[0143] Perform bounding box analysis on the fusion model to determine the external boundary of the fusion model and obtain the model bounding box.

[0144] A three-dimensional regular grid is constructed based on the model bounding box to obtain a model voxel space. The three-dimensional regular grid divides the model into multiple small voxels.

[0145] The surface of the fused model is densely sampled to obtain surface sampling points, which are key points on the model surface. Each voxel in the model voxel space is checked to see if it contains a surface sampling point, and the voxel containing the surface sampling point is marked as occupied, so that it can also be determined which voxels are part of the model.

[0146] According to the preset filling rules, the model voxel space is filled with voxels from the boundary to the inside, and the internal voxels of the model voxel space are marked as occupied. The filling rules include geometric shape-based filling rules, voxel exclusion rules, voxel density rules and voxel morphology rules.

[0147] Among them, voxel exclusion rules: set voxels that are not filled under certain conditions. For example, avoid filling voxels outside the model or marked as invalid areas.

[0148] Voxel density rule: Control the density of filled voxels according to the preset density requirements. Ensure that the filled model has a uniform density distribution and avoid over-filling or under-filling.

[0149] Voxel shape rules: Follow the natural shape of the model and avoid generating voxels that do not conform to the actual shape during the filling process. For example, avoid generating protruding voxels on a smooth surface.

[0150] According to the model voxel space, all occupied voxels are encoded to obtain a voxelized model, which represents the three-dimensional model as a combination of multiple small voxels. The surface of the voxelized model is reconstructed to obtain the final oral model.

[0151] This embodiment provides a good initial alignment position through feature point extraction and rough registration steps, greatly reducing the computational complexity and time of subsequent fine registration. Fine registration is based on an optimization algorithm, which enables the two data sets to be aligned with high precision, improving the accuracy of registration. The fusion process retains the advantages of each data set through techniques such as weighted averaging, ensuring the accuracy and integrity of the fusion model. Bounding box analysis and three-dimensional regular grid construction make the division of the model voxel space more standardized and easy to handle. Dense sampling and voxel labeling steps not only ensure the integrity of the model, but also improve the reliability of the data. The final voxelized model and surface reconstruction steps provide a high-precision oral model through precise encoding and reconstruction algorithms. This embodiment provides reliable three-dimensional data support for oral trajectories, helps to improve trajectory effects and patient satisfaction, and significantly improves the convenience and efficiency of clinical applications.

[0152] In one embodiment, trajectory analysis is performed on the patient's oral cavity according to the oral cavity model to obtain an initial trajectory path, and the initial trajectory path is iteratively optimized to obtain an optimized trajectory solution, including:

[0153] The acquired oral model is segmented into tooth units and jaw units. The segmentation process can use image segmentation algorithms such as region growing method and level set method. After the segmentation is completed, the tooth unit and jaw unit are positionally analyzed to determine the relative position relationship between them. The position analysis includes calculating the geometric features such as the center point coordinates and direction vectors of each unit, and establishing the spatial relationship between them.

[0154] According to the relative position relationship between the tooth unit and the jaw unit, the tooth unit is classified and labeled to identify the type of each tooth, such as incisor, canine, premolar and molar, etc. The classification process adopts a rule-based method.

[0155] After obtaining the tooth type information, the tooth type information is analyzed and processed based on the preset tooth arrangement rules. The tooth arrangement rules are set based on the professional knowledge of stomatology, including ideal tooth spacing, occlusal relationship, etc. The tooth arrangement rules are used to determine an ideal arrangement position for each tooth.

[0156] The displacement vector of each tooth unit from the current position to the ideal arrangement position is calculated. This process involves coordinate transformation and vector calculation in three-dimensional space. The displacement vectors of all tooth units are processed comprehensively to obtain the initial trajectory path. The initial trajectory path represents the initial movement plan of the tooth from the current state to the ideal state.

[0157] Each displacement vector is decomposed in three dimensions to obtain its component vectors in the x, y, and z directions, and the force analysis is performed on these component vectors. This step also includes various forces that may be encountered during tooth movement, such as the resistance of the periodontal membrane and the influence of adjacent teeth. By establishing a mechanical relationship for calculation, the force data corresponding to each component vector is obtained.

[0158] The three-dimensional component vector is adjusted based on the results of the force analysis. This adjustment process includes operations such as scaling, rotating or redirecting the vector. The purpose of the adjustment is to make the trajectory process smoother and more efficient while reducing unnecessary force effects. The adjusted vector is called the optimized component vector.

[0159] All optimized component vectors are synthesized to obtain new optimized displacement vectors, which constitute the optimized trajectory path.

[0160] This optimization process is repeated and iterated until the preset termination conditions are met, such as the displacement change is less than a preset threshold, and the number of iterations reaches the upper limit, etc. Finally, the patient's personalized tooth trajectory plan is obtained, which is the optimized trajectory plan.

[0161] This embodiment can accurately obtain the relative position relationship between the tooth unit and the jaw unit by performing precise three-dimensional segmentation and position analysis on the oral model, thereby providing a high-precision data basis for subsequent trajectory planning. Secondly, based on the preset tooth arrangement rules and classification labels, the system can automatically generate the ideal tooth arrangement position, reduce the error of human operation, and improve the scientificity and rationality of the trajectory plan. Furthermore, through the three-dimensional decomposition and force analysis of the initial trajectory path, various mechanical problems that may be encountered during tooth movement can be fully considered, and the trajectory path can be optimized to make the trajectory process smoother and more efficient. Finally, the iterative optimization process ensures the continuous improvement of each trajectory plan, and the final optimized trajectory plan is more personalized and accurate.

[0162] In one embodiment, the trajectory of the oral model is simulated based on the optimized trajectory solution to obtain the corresponding simulation trajectory result, including:

[0163] The oral model is segmented according to the optimized trajectory scheme. Assuming that the oral model includes the tooth arrangement of the maxilla and mandible, the entire oral model is divided into several sub-models according to the optimized trajectory scheme, and each sub-model corresponds to a part of the teeth. Specifically, the maxilla can be divided into the left anterior tooth area, the right anterior tooth area, the left posterior tooth area, and the right posterior tooth area; the mandible can also be divided into the left anterior tooth area, the right anterior tooth area, the left posterior tooth area, and the right posterior tooth area. In this way, the entire oral model is decomposed into eight sub-models.

[0164] Perform force analysis on each oral sub-model to obtain the corresponding force distribution diagram. Perform force analysis on each sub-model to simulate the distribution of trajectory force in teeth and periodontal tissues. For example, by analyzing the force conditions in the left anterior teeth area, the force magnitude and distribution of each tooth in the area in different directions are obtained. Similarly, perform the same force analysis on the other seven sub-models to obtain their respective force distribution diagrams.

[0165] Based on the force distribution diagram, each oral sub-model is deformed and simulated to obtain a deformed sub-model. In this process, the deformation of the teeth under the action of the trajectory force is simulated according to the force distribution diagram. For example, the teeth in the left anterior area may be displaced and rotated under the action of force. After deformation simulation, the deformed sub-model of this area is obtained. The other seven sub-models are also subjected to similar deformation simulations to obtain their own deformed sub-models.

[0166] All the deformed sub-models are combined to obtain the trajectory oral cavity overall model after the simulated trajectory. Specifically, the eight deformed sub-models are combined according to the actual structure of the oral cavity model to form a complete trajectory oral cavity model, which shows the overall changes of the oral cavity during the trajectory process under the optimized trajectory scheme.

[0167] The trajectory oral model is compared with the preset standard oral model to obtain deviation data. Through comparative analysis, the differences between the trajectory oral model and the standard oral model are identified, including deviation data in tooth position, angle, and arrangement. Based on these deviation data, the trajectory oral model is fine-tuned. For example, if the position of a tooth is found to deviate from the standard model, the position of the tooth is adjusted to the standard position through fine-tuning. After multiple fine-tuning, the fine-tuned oral model is finally obtained.

[0168] The fine-tuned oral model is rendered in three dimensions to obtain the simulation trajectory results. In this process, the fine-tuned oral model is rendered using three-dimensional rendering software to generate a high-precision three-dimensional image. The image can intuitively display the oral model after the trajectory, including information such as the arrangement and morphology of the teeth.

[0169] This embodiment performs segmented processing and force analysis on the oral model based on the optimized trajectory scheme, which can more accurately simulate the distribution of trajectory force in the oral cavity, thereby improving the accuracy of the trajectory scheme. Through deformation simulation and combined processing, the overall changes of the oral cavity during the trajectory process can be fully displayed, making the trajectory results more visual and intuitive. Comparing and analyzing the trajectory oral cavity overall model with the standard model, and fine-tuning the processing according to the deviation data, it can effectively reduce the possible errors in the trajectory process and improve the accuracy of the trajectory effect. Finally, the high-precision three-dimensional image obtained by three-dimensional rendering processing provides doctors with an intuitive reference, making it more convenient and efficient to evaluate and optimize the trajectory scheme. On the whole, the method of this embodiment not only improves the accuracy and visibility of the trajectory scheme, but also significantly improves the trajectory effect, providing doctors with a reliable basis for decision-making.

[0170] Reference Figure 2 As shown, the present invention also provides a three-dimensional image-guided correction planning system, which is applied to any one of the three-dimensional image-guided correction planning methods mentioned above, comprising:

[0171] An acquisition module, which is used to obtain a three-dimensional CT image and a two-dimensional X-ray image of the patient's oral cavity, and integrate and process the three-dimensional CT image and the two-dimensional X-ray image to obtain a corresponding comprehensive image set;

[0172] An analysis module is used to extract features from the comprehensive image set to obtain tooth images and jaw images, and to perform three-dimensional construction based on the tooth images and jaw images to obtain an initial three-dimensional model;

[0173] An association module is used to align and fuse the tooth image and the jaw image to obtain a corresponding oral fusion image;

[0174] A processing module, which is used to register the oral fusion image with the initial three-dimensional model based on a preset registration method to obtain a corresponding oral model;

[0175] The control module is used to perform trajectory analysis on the patient's oral cavity based on the oral cavity model, obtain an initial trajectory path, iteratively optimize the initial trajectory path, and obtain an optimized trajectory solution.

[0176] The execution module is used to simulate the trajectory of the oral model based on the optimized trajectory scheme to obtain the corresponding simulation trajectory results.

[0177] The three-dimensional image-guided correction planning system provided by the present invention can more accurately extract the features of teeth and jaws by integrating three-dimensional CT images and two-dimensional X-ray image data, thereby improving the construction accuracy of the oral model and providing a more reliable basis for trajectory planning. By aligning and fusing the image data and performing registration analysis based on the fused image, the refined management of oral details can be achieved, which helps to formulate a more accurate trajectory path and avoid unnecessary adjustments. The initial path is iteratively optimized based on the optimized trajectory scheme to ensure that the trajectory process can be carried out efficiently at different stages, reduce unnecessary errors, and improve the efficiency of the entire trajectory process. By comprehensively analyzing the oral model and trajectory path data, a more reasonable trajectory strategy is formulated, and the trajectory scheme is optimized through comprehensive adjustment, thereby effectively improving the trajectory effect and reducing treatment time and patient discomfort. And by considering the impact of individual differences on the trajectory scheme, the trajectory strategy can be flexibly adjusted according to the oral characteristics and demand changes of different patients, so that the trajectory process is more adaptable to diverse demand scenarios.

[0178] Reference Figure 3 As shown, the present invention also provides a three-dimensional image-guided correction planning device, comprising:

[0179] Memory, used to store programs;

[0180] The processor is used to execute the program to implement each step of any one of the above-mentioned three-dimensional image-guided correction planning methods.

[0181] In this embodiment, the processor and the memory may be connected via a bus or other means. The memory may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a read-only memory, a flash memory, a hard disk, or a solid-state drive. The processor may be a general-purpose processor, such as a central processing unit, a digital signal processor, an application-specific integrated circuit, or one or more integrated circuits configured to implement the embodiments of the present invention.

[0182] The present invention also provides a storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute any of the above methods.

[0183] It should be noted that technicians in the relevant technical field can clearly understand that for the convenience and conciseness of description, the specific working process of the system and each module described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0184] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A three-dimensional image-guided correction planning method, characterized in that: include: Acquire a three-dimensional CT image and a two-dimensional X-ray image of the patient's oral cavity, and integrate the three-dimensional CT image and the two-dimensional X-ray image to obtain a corresponding comprehensive image set; Extracting features from the comprehensive image set to obtain tooth images and jaw images, and performing three-dimensional construction based on the tooth images and the jaw images to obtain an initial three-dimensional model; Aligning and fusing the tooth image and the jaw image to obtain a corresponding oral fusion image; Based on a preset registration method, the oral fusion image is registered with the initial three-dimensional model to obtain a corresponding oral model; Performing trajectory analysis on the patient's oral cavity according to the oral cavity model to obtain an initial trajectory path, and iteratively optimizing the initial trajectory path to obtain an optimized trajectory solution; Based on the optimized trajectory scheme, the oral model is simulated to obtain a corresponding simulation trajectory result; The step of performing trajectory analysis on the patient's oral cavity according to the oral cavity model to obtain an initial trajectory path, and iteratively optimizing the initial trajectory path to obtain an optimized trajectory solution includes: Segmenting the oral cavity model to obtain tooth units and jaw units, and performing position analysis on the tooth units and jaw units to obtain a relative position relationship between the tooth units and jaw units; Classify and label the tooth units according to their relative positional relationships to obtain tooth type information; Analyze and process tooth type information based on preset tooth arrangement rules to obtain the ideal arrangement position; Calculating the displacement vector of each tooth unit according to the ideal arrangement position and the current position of the tooth unit, and comprehensively processing the displacement vectors of all tooth units to obtain the initial trajectory path; Performing three-dimensional decomposition on each displacement vector in the initial trajectory path to obtain a corresponding three-dimensional component vector, and performing force analysis on the three-dimensional component vector to obtain corresponding force data; Perform vector adjustment on the three-dimensional component vector according to the force data to obtain the optimized component vector; All optimized component vectors are synthesized to obtain an optimized displacement vector; The initial trajectory path is iteratively optimized based on the optimized displacement vector to obtain the optimized trajectory solution.

2. The three-dimensional image-guided correction planning method according to claim 1, characterized in that: The integrating and processing the three-dimensional CT image and the two-dimensional X-ray image to obtain a corresponding comprehensive image set includes: Performing distortion correction processing on the three-dimensional CT image to obtain a corrected three-dimensional CT image, and performing distortion correction on the two-dimensional X-ray image to obtain a corrected two-dimensional X-ray image; Align and match the corrected three-dimensional CT image and the corrected two-dimensional X-ray image to obtain a preliminary aligned image; Perform multi-resolution pyramid decomposition on the preliminary registered images to obtain image levels of different scales, and integrate all image levels to obtain a multi-scale image set; Perform wavelet transform on the multi-scale image set and extract the features of different frequency components to obtain the wavelet coefficient matrix, perform feature fusion on the wavelet coefficient matrix to obtain the fused feature representation; The fused feature representation is inversely transformed and reconstructed to obtain the comprehensive image set.

3. The three-dimensional image-guided correction planning method according to claim 1, characterized in that: The step of extracting features from the comprehensive image set to obtain a tooth image and a jaw image, and performing three-dimensional construction based on the tooth image and the jaw image to obtain an initial three-dimensional model includes: Performing image segmentation processing on the comprehensive image set to obtain the segmented tooth image and the jaw image; Extracting tooth contour features from the tooth image according to a preset edge detection algorithm to obtain tooth contour information; Extracting the features of the jaw bone line from the jaw image according to a preset bone line extraction algorithm to obtain jaw bone line information; Perform three-dimensional reconstruction based on the tooth contour information and jaw bone line information to obtain a preliminary reconstruction model; The model surface of the preliminary reconstructed model is smoothed to obtain the initial three-dimensional model.

4. The three-dimensional image-guided correction planning method according to claim 3, characterized in that: The step of aligning and fusing the tooth image and the jaw image to obtain a corresponding oral fusion image includes: Performing single segmentation processing on the tooth image according to the tooth contour information to obtain corresponding single tooth contour information; Performing edge detection on the jaw image according to the jaw skeleton line information to obtain jaw boundary information; The relationship between the single tooth contour information and the jaw boundary information is constructed to obtain the corresponding tooth-jaw correspondence matrix; Performing rigid registration on the tooth image and the jaw image according to a tooth-jaw correspondence matrix to obtain a preliminary alignment result; Perform non-rigid deformation processing on the preliminary alignment result to obtain a refined alignment result; The tooth image and the jaw image are interpolated and fused according to the fine alignment result to obtain the oral fusion image.

5. The three-dimensional image-guided correction planning method according to claim 1, characterized in that: The method of registering the oral fusion image with the initial three-dimensional model based on a preset registration method to obtain a corresponding oral model includes: Extracting feature points from the oral fusion image to obtain a first feature point set; Extracting feature points from the initial three-dimensional model to obtain a second feature point set; Performing a rough registration on the oral fusion image and the initial three-dimensional model according to the first feature point set and the second feature point set to obtain a rough registration result; Based on the coarse registration result, fine registration is performed on the oral fusion image and the initial three-dimensional model to obtain a fine registration result; According to the fine registration result, the oral fusion image and the initial three-dimensional model are fused to obtain a fusion model; Perform bounding box analysis on the fusion model to obtain a model bounding box, and construct a three-dimensional regular grid based on the model bounding box to obtain a model voxel space; Densely sample the surface of the fusion model to obtain surface sampling points; Detect whether each voxel in the model voxel space contains a surface sampling point, and mark the voxel containing the surface sampling point as occupied; Fill the model voxel space from the boundary to the inside according to a preset filling rule, and mark the internal voxels of the model voxel space as occupied; Encode all occupied voxels according to the model voxel space to obtain a voxelized model; The voxelized model is subjected to surface reconstruction to obtain the oral cavity model.

6. The three-dimensional image-guided correction planning method according to claim 1, characterized in that: The simulating trajectory of the oral model based on the optimized trajectory scheme to obtain a corresponding simulation trajectory result includes: Segmenting the oral cavity model according to the optimization trajectory scheme to obtain a plurality of oral cavity sub-models; Perform force analysis on each oral cavity model to obtain the corresponding force distribution diagram; Based on the force distribution diagram, each oral cavity sub-model is deformed by simulation to obtain a deformed sub-model, and all deformed sub-models are combined to obtain a trajectory oral cavity overall model after the simulation trajectory; Compare and analyze the trajectory oral cavity overall model with the preset standard oral cavity model to obtain deviation data, and fine-tune the trajectory oral cavity overall model according to the deviation data to obtain a fine-tuned oral cavity overall model; The fine-tuned oral cavity overall model is subjected to three-dimensional rendering processing to obtain the simulation trajectory result.

7. A three-dimensional image-guided correction planning system, characterized in that: The three-dimensional image-guided correction planning method applied to any one of claims 1 to 6 above comprises: An acquisition module, the acquisition module is used to obtain a three-dimensional CT image and a two-dimensional X-ray image of the patient's oral cavity, and integrate the three-dimensional CT image and the two-dimensional X-ray image to obtain a corresponding comprehensive image set; An analysis module, the analysis module is used to extract features from the comprehensive image set to obtain a tooth image and a jaw image, and to perform a three-dimensional construction based on the tooth image and the jaw image to obtain an initial three-dimensional model; An association module, the association module is used to align and fuse the tooth image and the jaw image to obtain a corresponding oral fusion image; A processing module, the processing module is used to register the oral fusion image with the initial three-dimensional model based on a preset registration method to obtain a corresponding oral model; A control module, the control module is used to perform trajectory analysis on the patient's oral cavity according to the oral cavity model to obtain an initial trajectory path, and iteratively optimize the initial trajectory path to obtain an optimized trajectory solution; An execution module, the execution module is used to simulate the trajectory of the oral model based on the optimized trajectory scheme to obtain a corresponding simulation trajectory result; The step of performing trajectory analysis on the patient's oral cavity according to the oral cavity model to obtain an initial trajectory path, and iteratively optimizing the initial trajectory path to obtain an optimized trajectory solution includes: Segmenting the oral cavity model to obtain tooth units and jaw units, and performing position analysis on the tooth units and jaw units to obtain a relative position relationship between the tooth units and jaw units; Classify and label the tooth units according to their relative positional relationships to obtain tooth type information; Analyze and process tooth type information based on preset tooth arrangement rules to obtain the ideal arrangement position; Calculating the displacement vector of each tooth unit according to the ideal arrangement position and the current position of the tooth unit, and comprehensively processing the displacement vectors of all tooth units to obtain the initial trajectory path; Performing three-dimensional decomposition on each displacement vector in the initial trajectory path to obtain a corresponding three-dimensional component vector, and performing force analysis on the three-dimensional component vector to obtain corresponding force data; Perform vector adjustment on the three-dimensional component vector according to the force data to obtain the optimized component vector; All optimized component vectors are synthesized to obtain an optimized displacement vector; The initial trajectory path is iteratively optimized based on the optimized displacement vector to obtain the optimized trajectory solution.

8. A three-dimensional image-guided correction planning device, characterized in that: include: Memory, used to store programs; A processor is used to execute the program to implement each step of a three-dimensional image-guided correction planning method as described in any one of claims 1-6.

9. A storage medium, characterized in that: Computer instructions are stored, and the computer instructions are used to make a computer execute the method according to any one of claims 1 to 6.

Images