CN118021474B - Dental implant model forming method based on image processing - Google Patents

Abstract

The invention relates to the field of image processing, in particular to a dental implant model forming method based on image processing. The method comprises the following steps: acquiring an oral CT image of a patient; carrying out multi-channel convolution treatment on the CT image of the oral cavity of the patient to construct an oral cavity convolution characteristic diagram; performing pixel-level semantic segmentation on the oral convolution feature map to segment the map in an oral area; generating a tooth missing region map based on the oral region segmentation map; carrying out morphological structure analysis on the tooth missing region graph to obtain tooth groove shape structure data; carrying out three-dimensional construction on the tooth socket shape structure data to generate a three-dimensional tooth socket bone model; performing topological form fitting based on the three-dimensional alveolar bone model to construct a first dental implant model; performing pressure distribution calculation around the dental implant on the first dental implant model to generate dental implant structure pressure data; the invention realizes the efficient molding of the dental implant model.

Description

Dental implant model forming method based on image processing

Technical Field

The invention relates to the field of image processing, and in particular to a dental implant model forming method based on image processing.

Background technique

Dental implant modeling is a common dental procedure used to restore missing teeth. Traditional dental implant modeling methods usually rely on photography, impressions, and manual production. These methods have some limitations, such as low efficiency, long time consumption, and introduction of human errors. In order to meet the needs of modern dental diagnosis and treatment, an intelligent, fast, and efficient dental implant modeling method is needed. A dental implant modeling method based on image processing came into being.

Summary of the invention

The present invention provides a dental implant model forming method based on image processing to solve at least one of the above technical problems.

To achieve the above object, the present invention provides a dental implant model forming method based on image processing, comprising the following steps:

Step S1: obtaining an oral CT image of a patient; performing multi-channel convolution processing on the oral CT image of the patient to construct an oral convolution feature map; performing pixel-level semantic segmentation on the oral convolution feature map to obtain an oral region segmentation map;

Step S2: generating a tooth missing region map based on the oral region segmentation map; performing morphological structure analysis on the tooth missing region map to obtain alveolar morphological structure data; performing three-dimensional construction on the alveolar morphological structure data to generate a three-dimensional alveolar bone model;

Step S3: performing topological morphology fitting based on the three-dimensional alveolar bone model to construct a first dental implant model; calculating the pressure distribution around the dental implant on the first dental implant model to generate dental implant structure pressure data;

Step S4: performing oral multi-time series node prediction based on the dental implant structure pressure data to obtain oral multi-time point prediction results; performing tissue morphological structure evolution analysis on the oral multi-time point prediction results to obtain oral morphological evolution prediction data;

Step S5: performing time series trend curve fitting on the oral morphology evolution prediction data to obtain the morphology evolution time series trend curve; performing boundary buffer area analysis on the first dental implant model based on the morphology evolution time series trend curve to obtain a deformation intelligent buffer zone;

Step S6: Simulate implantation of the first dental implant model based on the deformation intelligent buffer zone to obtain a holographic alveolus-dental implant model; optimize the stress area design based on the holographic alveolus-dental implant model to construct an optimized dental implant model.

The present invention processes and segments the patient's oral CT image to obtain an accurate segmentation map of the oral region, extracts key features in the oral image through multi-channel convolution processing, and helps subsequent analysis and processing, while pixel-level semantic segmentation further classifies each pixel in the oral convolution feature map so that each pixel can be accurately attributed to the oral region or other regions, thereby obtaining an oral region segmentation map, and further analyzes and processes the oral region segmentation map to obtain a tooth loss region map, and performs morphological structure analysis on the tooth loss region map to obtain morphological structure data of the alveolar, including shape, size, position and other information, and then, through three-dimensional construction technology, the obtained alveolar morphological structure data is converted into a three-dimensional alveolar bone model for subsequent processing and analysis, and the three-dimensional alveolar bone model is used for topological morphological fitting, and a first dental implant model is constructed according to the morphological structure data of the alveolar, and the pressure distribution around the dental implant is calculated for the first dental implant model to obtain dental implant structure pressure data, which is used to evaluate the stability and adaptability of the dental implant in the oral environment, and provide for subsequent optimization design. Based on the data of dental implant structure pressure, oral multi-time node prediction is performed, that is, the changes of the oral cavity at different time points are predicted to obtain the state of the oral cavity in the future. The tissue morphological structure evolution analysis is performed on the oral multi-time point prediction results, the changing trend of oral morphology is observed, and the oral morphology evolution prediction data is obtained to provide guidance for subsequent design and optimization. According to the trend of morphological evolution, the boundary of the dental implant model is adjusted to create an intelligent buffer zone, taking into account the changes in oral morphology and providing a certain fault tolerance space to adapt to future changes and improve the stability and adaptability of dental implants. The first dental implant model is simulated and implanted into the oral model using the deformation intelligent buffer zone to obtain a complete holographic alveolar-dental implant model, taking into account the trend of oral morphological evolution and providing a buffer zone to adapt to future changes. Based on the holographic alveolar-dental implant model, stress area optimization design is performed. By analyzing the stress distribution in the oral cavity, the stress area in the dental implant model is determined, and the optimization design is performed to improve the stability and adaptability of the dental implant and ensure its long-term success in the oral environment.

Preferably, step S1 comprises the following steps:

Step S11: Acquire a CT image of the patient's oral cavity;

Step S12: performing multi-channel convolution processing on the patient's oral CT image to extract cross-band feature data;

Step S13: constructing an oral cavity convolution feature map based on the cross-band feature data;

Step S14: performing branch sampling processing on the oral cavity convolution feature map to obtain a spatial detail convolution map;

Step S15: Perform pixel-level semantic segmentation on the spatial detail convolution map to obtain an oral region segmentation map.

The present invention obtains a CT image of the patient's oral cavity, which provides detailed information on the oral structure, including the morphology and position of teeth, alveolar bones and surrounding tissues. By obtaining the oral CT image, basic data is provided for subsequent processing and analysis. Multi-channel convolution processing technology is used to extract cross-band feature data from the oral CT image. The multi-channel convolution processing captures the correlation and feature representation between different bands, which helps to extract richer and more accurate feature information of the oral image. The oral convolution feature map is obtained by performing a convolution operation on the cross-band feature data, which better represents the feature information such as structure, texture and morphology in the oral image. The branch sampling processing further highlights the spatial detail information in the oral image, such as the edge, morphology and microstructure of the teeth, which helps the subsequent semantic segmentation and morphological analysis. The pixel-level semantic segmentation classifies each pixel in the oral convolution feature map and accurately attributes it to the oral region or other regions, thereby obtaining an oral region segmentation map. The oral region segmentation map provides accurate boundaries and segmentation results of various structures in the oral cavity, providing a basis for subsequent dental implant modeling.

Preferably, the specific steps of step S14 are:

Step S141: performing multi-layer downsampling on the oral cavity convolution feature map to obtain global large-granularity feature data;

Step S142: performing multi-scale upsampling on the oral cavity convolution feature map to obtain local detail feature data;

Step S143: performing residual connection processing on the global large-grained feature data and the local detail feature data to generate missing detail feature data;

Step S144: Obtain a spatial detail convolution map based on the missing detail feature data.

The present invention performs dimensionality reduction processing on the oral convolution feature map through multi-layer downsampling operations, thereby obtaining global large-grained feature data. Multi-layer downsampling gradually reduces the spatial size of the feature map and increases the number of channels at the same time, so that the model can better capture the feature information of the overall oral structure and reduce the computational complexity. The oral convolution feature map is enlarged through multi-scale upsampling operations to obtain local detail feature data. Multi-scale upsampling restores the spatial detail information of the feature map, so that the model can better capture the local fine structure and texture features in the oral image, which helps to improve the precision and accuracy of the model. The global large-grained feature data is processed through residual connection processing. The missing detail feature data is fused with the local detail feature data to generate the missing detail feature data. The residual connection effectively transmits and fuses feature information of different scales, thereby filling the information missing between the global and local levels and improving the model's perception of detail information. The generated missing detail feature data is used to obtain the final spatial detail convolution map. By integrating the global large-grained feature data, the local detail feature data and the missing detail feature data, a spatial detail convolution map with more comprehensive information is obtained, which contains rich detail feature information such as oral structure, texture and morphology. Such a spatial detail convolution map provides a more accurate and comprehensive input for the subsequent molding of the dental implant model.

Preferably, the specific steps of step S2 are:

Step S21: performing fine-grained structure recognition on the oral region segmentation map to obtain a tooth missing region map;

Step S22: extracting the alveolar contour from the tooth missing area map to obtain alveolar area contour data;

Step S23: performing morphological structure analysis on the alveolar area contour data to obtain alveolar morphological structure data;

Step S24: performing tooth topological gap calculation on the tooth missing area map to obtain peripheral tooth gap range data;

Step S25: constructing the alveolar morphological structure data in three dimensions based on the peripheral tooth space range data to generate a three-dimensional alveolar bone model.

The present invention performs fine-grained structural recognition on an oral region segmentation map to accurately locate and identify tooth-missing regions, analyzes and identifies the oral region segmentation map to determine which regions correspond to tooth-missing locations, thereby generating a tooth-missing region map, processes the tooth-missing region map to extract contour information of the alveolar, performs edge detection and contour extraction on the tooth-missing region map to obtain shape and boundary information of the alveolar, thereby obtaining contour data of the alveolar region, performs morphological analysis on the alveolar region contour data to obtain morphological structural information of the alveolar, performs morphological transformation on the alveolar region contour by applying morphological operations, such as expansion, corrosion, opening or closing operations, and the like, thereby analyzing the shape, The size and structural characteristics of the alveolar shape and structure data are obtained, the tooth missing area map is processed, and the topological gap of the tooth, that is, the range of the tooth gap around the missing tooth, is calculated. By analyzing the position of the teeth in the tooth missing area map and the distribution of the surrounding teeth, the range of the tooth gap around the missing tooth is determined, providing a basis for the positioning and design of the implant model. Based on the surrounding tooth gap range data, the alveolar morphological structure data is used for three-dimensional construction to generate a three-dimensional alveolar bone model. By combining the alveolar morphological structure data with the surrounding tooth gap range data, the key parameters such as the position, angle and size of the implant in the alveolar bone are determined, and then an accurate three-dimensional alveolar bone model is generated, providing a basis for the subsequent molding of the dental implant model.

Preferably, the specific steps of step S3 are:

Step S31: performing topological morphology fitting based on the three-dimensional alveolar bone model to construct a first dental implant model;

Step S32: performing elastic deformation dynamic constraint analysis on the first dental implant model to obtain elastic deformation dynamic constraint data;

Step S33: performing chewing wear simulation on the elastic deformation dynamics constraint data to obtain wear simulation data;

Step S34: Calculate the pressure distribution around the dental implant based on the wear simulation data to generate dental implant structure pressure data.

The present invention generates a first dental implant model by matching and fitting parameters such as the shape, size and position of the implant with the alveolar bone model to ensure good fit and stability between the implant and the alveolar bone, performs elastic deformation dynamic constraint analysis on the first dental implant model to obtain relevant constraint data, determines the force constraint of the implant including parameters such as bearing capacity and stress distribution by analyzing the elastic deformation and mechanical behavior of the implant during chewing and biting, thereby obtaining elastic deformation dynamic constraint data, evaluates the contact and wear between the implant and surrounding tissues by simulating chewing motion and mechanical action, obtains wear simulation data, and helps to understand the durability and service life of the implant, calculates the pressure distribution around the implant including parameters such as contact force and stress distribution by analyzing the mechanical action and contact during chewing, provides dental implant structure pressure data, and helps to evaluate the mechanical interaction and stability between the implant and surrounding tissues.

Preferably, the specific steps of step S4 are:

Step S41: performing oral multi-time series node prediction based on the dental implant structure pressure data to obtain oral multi-time point prediction results;

Step S42: performing bone tissue morphology change analysis on the oral multi-time point prediction results to obtain bone tissue morphology change data;

Step S43: performing soft tissue change analysis on the oral multi-time point prediction results to obtain oral soft tissue change data;

Step S44: performing structural feature point change identification based on the bone tissue morphology change data and the oral soft tissue change data to obtain changed structural feature points;

Step S45: performing position evolution rate analysis on the changed structural feature points to obtain the structural evolution law;

Step S46: Performing tissue morphological structure evolution analysis on the oral multi-time point prediction results according to the structure evolution law to obtain oral morphological evolution prediction data.

The present invention predicts the oral state of the oral cavity at different time points, including the stability of the implant, the changes in the surrounding bone tissue, etc., by analyzing the mechanical characteristics and changes in the pressure data of the dental implant structure, so as to obtain the oral multi-time point prediction results. By comparing and analyzing the oral multi-time point prediction results, the morphological changes of the oral bone tissue at different time points are determined, including processes such as bone absorption and bone formation, so as to provide the oral bone tissue morphological change data, which is helpful to understand the interaction and changes between the implant and the surrounding bone tissue. By analyzing the position and morphological changes of the soft tissue in the oral multi-time point prediction results, the changes of the oral soft tissue at different time points are determined, including the morphology of the gums, the thickness of the soft tissue, etc., so as to provide the oral soft tissue change data, which is helpful to understand the interaction and changes between the implant and the surrounding soft tissue. By analyzing the changes in bone tissue and soft tissue, the position and morphological changes of key structural feature points in the oral cavity are determined, such as changes in bone tissue around implants, changes in the position of the gum line, etc., so as to obtain the changed structural feature points and provide quantitative data on the evolution of oral structure. By analyzing the position and change amplitude of the changed structural feature points, the rate and trend of oral structure evolution are calculated, and the evolution of oral structure at different time points is understood, including the absorption of bone tissue, changes in soft tissue, etc., so as to obtain the evolution law of oral structure. By combining the evolution law of oral structure with the results of oral multi-time point prediction, the morphological changes of the oral cavity at future time points are inferred, including the stability of implants, changes in surrounding bone tissue and soft tissue, etc., so as to provide predictive data on oral morphological evolution, which is helpful for clinical decision-making and treatment planning.

Preferably, the specific steps of step S5 are:

Step S51: calculating the maximum deformation area of the oral morphology evolution prediction data to obtain the maximum deformation area of the oral morphology;

Step S52: fitting a time series trend curve for the maximum deformation area of the oral morphology to obtain a morphology evolution time series trend curve;

Step S53: performing boundary buffer area analysis on the first dental implant model based on the morphological evolution time series trend curve to generate an implant deformation buffer zone;

Step S54: performing stress topology optimization on the implant deformation buffer zone to obtain a deformation intelligent buffer zone.

The present invention determines the area with the most dramatic morphological changes in the oral cavity, that is, the maximum deformation area, by comparing and analyzing the oral morphological evolution prediction data. This area involves bone tissue, soft tissue, etc. around the implant. By calculating the maximum deformation area, the part of the oral cavity that needs to be focused on and treated is located. By analyzing the changes in the maximum deformation area of the oral morphology, a temporal trend curve of morphological evolution is fitted to reflect the trend of oral morphology changes over time, so as to better understand the evolution speed and trend of oral morphology and provide a basis for further analysis and decision-making. By analyzing the relationship between the temporal trend of oral morphological evolution and the first dental implant model, the boundary buffer area around the implant, that is, the area where the implant is affected by deformation, is determined to improve the stability and success rate of the implant. By analyzing and optimizing the implant deformation buffer area, the optimal deformation buffer area is determined to provide better support and protection for the implant. Stress topology optimization helps determine the material distribution and structural design in the deformation buffer area to minimize the adverse effects of stress concentration and deformation on the implant. Such a deformation intelligent buffer area improves the stability of the implant, reduces adverse stress and damage, and promotes the molding of the dental implant model.

Preferably, the specific steps of step S53 are:

Step S531: performing deformation rate matching analysis on the first dental implant model based on the morphological evolution time series trend curve to generate deformation rate data of the dental implant model;

Step S532: analyzing the regional force variation law of the dental implant model deformation rate data to obtain the regional force variation law;

Step S533: performing boundary buffer region analysis based on the regional force variation law to generate an implant deformation buffer region.

The present invention determines the deformation degree of the implant model at different time points through matching analysis between the time trend curve and the dental implant model, understands the morphological changes of the implant at different time points, and provides a basis for subsequent steps. By analyzing the deformation rate data of different areas of the dental implant model, the change law of the force exerted on different areas during the deformation process is determined, the deformation of each area in the implant model is understood, and the areas with large or uneven forces are identified to determine the areas that need to be analyzed for boundary buffer areas. By analyzing the change law of regional force, the areas that need special attention and treatment in the implant model, that is, the areas that are subjected to large or uneven forces, are determined. These areas are considered to be potential risk areas for deformation. By performing boundary buffer area analysis, the boundary areas in the implant model that need to be set with buffers are determined to provide better support and protection and reduce the adverse effects of deformation on the implant. The generated deformation buffer zone will contribute to the molding of the dental implant model and the success of clinical treatment.

Preferably, the specific steps of step S6 are:

Step S61: reconstructing the topological shape of the first dental implant model based on the deformation intelligent buffer zone to construct a second dental implant model;

Step S62: performing simulated implantation on the three-dimensional alveolar bone model through the second dental implant model to obtain a holographic alveolar-dental implant model;

Step S63: performing dental implant position recognition on the holographic alveolus-dental implant model to obtain embedding position data;

Step S64: quantifying the distribution of tissue pressure at the edge of the holographic alveolus-dental implant model based on the embedding position data to obtain quantitative tissue pressure data;

Step S65: performing stress area optimization design on the second dental implant model based on the tissue pressure quantification data to construct an optimized dental implant model.

The present invention reconstructs the shape of the first dental implant model by utilizing the data of the deformation intelligent buffer zone, so that its shape is more adapted to the changes in the oral shape, so as to obtain a second dental implant model, whose shape is more in line with the actual situation, and improves the adaptability and stability of the model in the oral cavity. By combining the second dental implant model with the three-dimensional alveolar bone model, the position and interaction of the dental implant in the alveolar bone are simulated, and the obtained holographic alveolar-dental implant model more accurately reflects the embedding situation of the dental implant, and provides a more realistic oral structure model. By analyzing the holographic alveolar-dental implant model, the specific position of the dental implant in the oral cavity is determined, and the embedding position data obtained in this way are provided to clinicians for reference, helping them to accurately locate the dental implant in actual operation and ensure the precision and accuracy of the implantation. By analyzing the holographic alveolar-dental implant model, the pressure distribution of the adjacent tissues around the dental implant is determined. The obtained tissue pressure quantitative data is used to evaluate the stress of the tissue around the implant and is provided to clinicians for reference to determine the stability of the implant and the degree of impact on the surrounding tissues. By analyzing the tissue pressure quantitative data, the areas with high or uneven stress in the dental implant model are determined. Based on these data, the second dental implant model is optimized and designed, and its morphology and structure are adjusted to reduce the stress concentration areas, improve the stability of the dental implant and the adaptability of the alveolar bone, and the optimized dental implant model better shares the chewing force, reduces the adverse effects on the surrounding tissues, improves the long-term success rate of the implant, and contributes to the molding of the dental implant model and the success of clinical treatment.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of the steps of a dental implant model forming method based on image processing according to the present invention;

FIG2 is a schematic diagram of a detailed implementation process of step S1;

FIG3 is a schematic diagram of a detailed implementation process of step S2;

FIG. 4 is a schematic flow chart of the detailed implementation steps of step S3.

Detailed ways

It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

The present application example provides a dental implant model forming method based on image processing. The execution subject of the dental implant model forming method based on image processing includes but is not limited to the following: mechanical equipment, data processing platform, cloud server node, network upload device, etc. equipped with the system can be regarded as the general computing node of the present application, and the data processing platform includes but is not limited to: at least one of an audio image management system, an information management system, and a cloud data management system.

Referring to FIGS. 1 to 4 , the present invention provides a dental implant model forming method based on image processing, comprising the following steps:

Step S1: obtaining an oral CT image of a patient; performing multi-channel convolution processing on the oral CT image of the patient to construct an oral convolution feature map; performing pixel-level semantic segmentation on the oral convolution feature map to obtain an oral region segmentation map;

Step S2: generating a tooth missing region map based on the oral region segmentation map; performing morphological structure analysis on the tooth missing region map to obtain alveolar morphological structure data; performing three-dimensional construction on the alveolar morphological structure data to generate a three-dimensional alveolar bone model;

Step S3: performing topological morphology fitting based on the three-dimensional alveolar bone model to construct a first dental implant model; calculating the pressure distribution around the dental implant on the first dental implant model to generate dental implant structure pressure data;

Step S4: performing oral multi-time series node prediction based on the dental implant structure pressure data to obtain oral multi-time point prediction results; performing tissue morphological structure evolution analysis on the oral multi-time point prediction results to obtain oral morphological evolution prediction data;

Step S5: performing time series trend curve fitting on the oral morphology evolution prediction data to obtain the morphology evolution time series trend curve; performing boundary buffer area analysis on the first dental implant model based on the morphology evolution time series trend curve to obtain a deformation intelligent buffer zone;

Step S6: Simulate implantation of the first dental implant model based on the deformation intelligent buffer zone to obtain a holographic alveolus-dental implant model; optimize the stress area design based on the holographic alveolus-dental implant model to construct an optimized dental implant model.

The present invention processes and segments the patient's oral CT image to obtain an accurate segmentation map of the oral region, extracts key features in the oral image through multi-channel convolution processing, and helps subsequent analysis and processing, while pixel-level semantic segmentation further classifies each pixel in the oral convolution feature map so that each pixel can be accurately attributed to the oral region or other regions, thereby obtaining an oral region segmentation map, and further analyzes and processes the oral region segmentation map to obtain a tooth loss region map, and performs morphological structure analysis on the tooth loss region map to obtain morphological structure data of the alveolar, including shape, size, position and other information, and then, through three-dimensional construction technology, the obtained alveolar morphological structure data is converted into a three-dimensional alveolar bone model for subsequent processing and analysis, and the three-dimensional alveolar bone model is used for topological morphological fitting, and a first dental implant model is constructed according to the morphological structure data of the alveolar, and the pressure distribution around the dental implant is calculated for the first dental implant model to obtain dental implant structure pressure data, which is used to evaluate the stability and adaptability of the dental implant in the oral environment, and provide for subsequent optimization design. Based on the data of dental implant structure pressure, oral multi-time node prediction is performed, that is, the changes of the oral cavity at different time points are predicted to obtain the state of the oral cavity in the future. The tissue morphological structure evolution analysis is performed on the oral multi-time point prediction results, the changing trend of oral morphology is observed, and the oral morphology evolution prediction data is obtained to provide guidance for subsequent design and optimization. According to the trend of morphological evolution, the boundary of the dental implant model is adjusted to create an intelligent buffer zone, taking into account the changes in oral morphology and providing a certain fault tolerance space to adapt to future changes and improve the stability and adaptability of dental implants. The first dental implant model is simulated and implanted into the oral model using the deformation intelligent buffer zone to obtain a complete holographic alveolar-dental implant model, taking into account the trend of oral morphological evolution and providing a buffer zone to adapt to future changes. Based on the holographic alveolar-dental implant model, stress area optimization design is performed. By analyzing the stress distribution in the oral cavity, the stress area in the dental implant model is determined, and the optimization design is performed to improve the stability and adaptability of the dental implant and ensure its long-term success in the oral environment.

In an embodiment of the present invention, referring to FIG. 1 , which is a schematic flow chart of a dental implant model forming method based on image processing of the present invention, in this example, the steps of the dental implant model forming method based on image processing include:

Step S1: obtaining an oral CT image of a patient; performing multi-channel convolution processing on the oral CT image of the patient to construct an oral convolution feature map; performing pixel-level semantic segmentation on the oral convolution feature map to obtain an oral region segmentation map;

In this embodiment, an appropriate medical imaging device, such as a CT scanner, is used to scan the patient's oral cavity to obtain three-dimensional CT image data of the oral cavity. The oral CT image is used as input and processed by a multi-channel convolutional neural network (CNN). The image is preprocessed, such as resizing, cropping, or padding, to meet the input requirements of the network. The preprocessed image is input into the convolutional network. The convolutional network extracts feature information of the image through a series of convolutional layers, activation functions, and pooling layers. Oral convolution feature maps are obtained from the last layer or middle layer of the convolutional network. These feature maps are high-dimensional data representations, in which each channel corresponds to different features, and a label is predicted for each pixel. The oral region is segmented from other regions to generate an oral region segmentation map.

Step S2: generating a tooth missing region map based on the oral region segmentation map; performing morphological structure analysis on the tooth missing region map to obtain alveolar morphological structure data; performing three-dimensional construction on the alveolar morphological structure data to generate a three-dimensional alveolar bone model;

In this embodiment, a tooth missing area map is generated through pixel-level operations. In the oral region segmentation map, the tooth missing area is marked as a positive sample, and other areas are marked as negative samples. The tooth missing area map is processed using morphological operations such as corrosion, dilation, opening or closing operations to remove noise, fill cavities, smooth boundaries, etc. The morphological structure analysis further optimizes the shape and edges of the tooth missing area. This is achieved by converting two-dimensional morphological structure data into three-dimensional coordinate points or voxel representations. Commonly used methods include surface reconstruction, volume rendering or voxelization, etc. Through three-dimensional construction technology, an alveolar bone model with a geometric shape and topological structure is generated.

Step S3: performing topological morphology fitting based on the three-dimensional alveolar bone model to construct a first dental implant model; calculating the pressure distribution around the dental implant on the first dental implant model to generate dental implant structure pressure data;

In this embodiment, the design parameters of the first dental implant (such as diameter, length, angle, etc.) are fitted with the three-dimensional alveolar bone model. The goal of topological fitting is to find a suitable position and direction for the implant in the alveolar bone model to achieve stable implantation of the dental implant. According to the result of topological fitting, a three-dimensional model of the first dental implant is generated, which is achieved by placing a model of appropriate shape and size in the alveolar bone model to represent the position and geometric shape of the first dental implant. A pressure analysis is performed on the first dental implant model to simulate the stress conditions of the tissue around the implant during chewing or biting. According to the simulation results, the pressure data of the dental implant structure is calculated and generated.

Step S4: performing oral multi-time series node prediction based on the dental implant structure pressure data to obtain oral multi-time point prediction results; performing tissue morphological structure evolution analysis on the oral multi-time point prediction results to obtain oral morphological evolution prediction data;

In this embodiment, oral multi-time series dental implant structure pressure data is used as input, and a time series prediction method is applied to perform node prediction at multiple time points in the oral cavity, which is achieved through statistical models, machine learning, or deep learning. The prediction result is the state or characteristics of each node (such as teeth, alveolar bones, etc.) in the oral cavity at multiple time points in the future. According to the prediction result, a prediction model or morphological feature of the oral cavity at different time points is generated, including information such as tooth position, alveolar bone density, and tissue pressure distribution. For each time point, a corresponding oral model or morphological parameter is generated. By comparing the prediction models or morphological parameters of the oral cavity at different time points, a morphological structure evolution analysis is performed, including calculating indicators such as the morphological change amount and the morphological change rate, so as to understand the dynamic changes in the oral morphology.

Step S5: fitting the oral morphology evolution prediction data with a time series trend curve to obtain a morphology evolution time series trend curve; performing boundary buffer area analysis on the first dental implant model based on the morphology evolution time series trend curve to obtain a deformation intelligent buffer zone;

In this embodiment, the oral morphology evolution prediction data is fitted into a time-series trend curve by fitting the prediction data at known time points. The evolution trend of the oral morphology is analyzed based on the fitted time-series trend curve, including calculating indicators such as the slope and change rate of the curve to understand the changes in the oral morphology. Based on the morphology evolution time-series trend curve, a boundary buffer area of the first dental implant model is defined, and the buffer area adjusts its shape and size according to changes in the time-series trend curve. A dynamic boundary buffer area is defined, which is dynamically adjusted according to the slope or change rate of the time-series trend curve.

Step S6: Simulate implantation of the first dental implant model based on the deformation intelligent buffer zone to obtain a holographic alveolus-dental implant model; optimize the stress area design based on the holographic alveolus-dental implant model to construct an optimized dental implant model.

In this embodiment, the first dental implant model is aligned with the oral model, and the dental implant model is placed at the corresponding position in the oral model for numerical simulation. The material properties, geometric shapes and boundary conditions of the teeth, alveolar bones and surrounding tissues are considered. During the simulation, the boundary information of the deformation intelligent buffer zone will be used to limit the deformation range of the dental implant. The stress distribution in the holographic alveolar-dental implant model is calculated through the simulation results. This is achieved through finite element analysis and other methods. The stress analysis reveals the stress concentration areas and stress conditions of the dental implant and surrounding tissues. According to the stress analysis results, the stress area is optimized. By adjusting the shape, structure or material distribution of the dental implant, the stress concentration is reduced and the stability of the dental implant is improved.

In this embodiment, referring to FIG. 2 , it is a schematic flow chart of detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include:

Step S11: Acquire a CT image of the patient's oral cavity;

Step S12: performing multi-channel convolution processing on the patient's oral CT image to extract cross-band feature data;

Step S13: constructing an oral cavity convolution feature map based on the cross-band feature data;

Step S14: performing branch sampling processing on the oral cavity convolution feature map to obtain a spatial detail convolution map;

Step S15: Perform pixel-level semantic segmentation on the spatial detail convolution map to obtain an oral region segmentation map.

In this embodiment, CT scanning is used to obtain three-dimensional image data of the patient's oral cavity. A convolutional neural network (CNN) is used to apply multiple convolution kernels to the oral CT image for convolution operations. Each convolution kernel extracts specific cross-band features. Multiple convolution operations are performed to obtain feature maps of multiple channels. The feature data obtained through multi-channel convolution processing are integrated to construct an oral convolution feature map. This is achieved by combining, superimposing or other operations on the feature maps of each channel. A branch sampling operation is performed on the oral convolution feature map to extract spatial detail information. This is achieved by using filters, pooling operations or other sampling techniques. The sampling operation helps focus on local areas of the oral image and extract more detailed information. A spatial detail convolution map is obtained through branch sampling processing, which contains local detail information of the oral image. A pixel-level semantic segmentation algorithm is used to perform pixel-level classification on the spatial detail convolution map, and each pixel of the oral image is assigned to a specific category, such as teeth, gums, bones, etc. Through pixel-level semantic segmentation, an oral region segmentation map is obtained, in which each pixel is labeled with a corresponding category.

In this embodiment, the specific steps of step S14 are:

Step S141: performing multi-layer downsampling on the oral cavity convolution feature map to obtain global large-granularity feature data;

Step S142: performing multi-scale upsampling on the oral cavity convolution feature map to obtain local detail feature data;

Step S143: performing residual connection processing on the global large-grained feature data and the local detail feature data to generate missing detail feature data;

Step S144: Obtain a spatial detail convolution map based on the missing detail feature data.

In this embodiment, when downsampling is performed, the feature map is divided into non-overlapping areas, and the pixel values in each area are pooled to obtain a single value, thereby reducing the size of the feature map and extracting global large-grained feature data. When upsampling is performed, each pixel value in the feature map is copied to the target position according to the size that needs to be enlarged, and other pixel values are filled by interpolation or convolution, etc., to increase the size of the feature map, and extract local detail feature data. The global large-grained feature data and the local detail feature data are residually connected by adding the corresponding pixel values of the two feature maps. The residual connection helps to transmit the missing detail information and improve the expressiveness of the features. The generated missing detail feature data is further used for convolution operations, filtering operations, etc., and a spatial detail convolution map is obtained through processing based on the missing detail feature data, which contains richer detail information of the oral image.

In this embodiment, referring to FIG. 3 , which is a schematic flow chart of detailed implementation steps of step S2, in this embodiment, the detailed implementation steps of step S2 include:

Step S21: performing fine-grained structure recognition on the oral region segmentation map to obtain a tooth missing region map;

Step S22: extracting the alveolar contour from the tooth missing area map to obtain alveolar area contour data;

Step S23: performing morphological structure analysis on the alveolar area contour data to obtain alveolar morphological structure data;

Step S24: performing tooth topological gap calculation on the tooth missing area map to obtain peripheral tooth gap range data;

Step S25: constructing the alveolar morphological structure data in three dimensions based on the peripheral tooth space range data to generate a three-dimensional alveolar bone model.

In this embodiment, fine-grained structure recognition is performed on the basis of the oral region segmentation map to identify the tooth missing region, and a deep learning method, such as a convolutional neural network (CNN), is used to classify or pixel-level predict the segmentation map to mark the missing tooth region, and an edge detection or contour extraction algorithm is performed on the tooth missing region map to extract the contour of the alveolar region. A gradient-based algorithm (such as a Sobel operator or a Canny edge detection) or an algorithm based on a region boundary (such as a boundary tracking algorithm) is used to perform a morphological analysis on the alveolar region contour data, including calculating the morphological features of the contour, such as the length, area, and perimeter, and applying morphological operations (such as corrosion, expansion, opening, and closing operations). ) to further improve the contour shape, remove noise or fill holes, analyze the tooth missing area map, calculate the topological gaps between teeth, and achieve this by measuring the distance between the missing area and the surrounding teeth or the distance between the boundaries. Use distance transformation, boundary detection and other technologies to calculate the range and distribution of the gaps. Use the surrounding tooth gap range data and alveolar morphological structure data to perform a three-dimensional reconstruction algorithm to generate a three-dimensional alveolar bone model. Use computer graphics and computer-aided design/manufacturing (CAD/CAM) technology to convert the alveolar morphological structure data into three-dimensional coordinate points or grid models, and interpolate or correct them according to the surrounding tooth gap range data to generate an accurate three-dimensional alveolar bone model.

In this embodiment, referring to FIG. 4 , which is a flowchart of detailed implementation steps of step S3, in this embodiment, the detailed implementation steps of step S3 include:

Step S31: performing topological morphology fitting based on the three-dimensional alveolar bone model to construct a first dental implant model;

Step S32: performing elastic deformation dynamic constraint analysis on the first dental implant model to obtain elastic deformation dynamic constraint data;

Step S33: performing chewing wear simulation on the elastic deformation dynamics constraint data to obtain wear simulation data;

Step S34: Calculate the pressure distribution around the dental implant based on the wear simulation data to generate dental implant structure pressure data.

In this embodiment, the shape of the first dental implant is fitted to the three-dimensional alveolar bone model to obtain the position and posture of the first dental implant, and an elastic deformation dynamic constraint analysis is performed on the first dental implant model to simulate the force and pressure applied to the implant during chewing. This is achieved using finite element analysis (FEA) or other simulation methods. The implant model is combined with material properties, and the elastic behavior of the tissue and the effect of chewing force are taken into account. Based on the elastic deformation dynamic constraint data, the contact and wear between the teeth during chewing are simulated. Simulation methods such as contact force analysis and friction models are used to combine chewing motion data and tooth surface characteristics to simulate the contact, sliding and wear processes between the teeth. The pressure distribution around the dental implant is calculated based on the wear simulation data. The pressure distribution around the dental implant is calculated by combining the wear simulation data with the pressure transfer model and taking into account the conduction, distribution and influence of the chewing force.

In this embodiment, step S4 includes the following steps:

Step S41: performing oral multi-time series node prediction based on the dental implant structure pressure data to obtain oral multi-time point prediction results;

Step S42: performing bone tissue morphology change analysis on the oral multi-time point prediction results to obtain bone tissue morphology change data;

Step S43: performing soft tissue change analysis on the oral multi-time point prediction results to obtain oral soft tissue change data;

Step S44: performing structural feature point change identification based on the bone tissue morphology change data and the oral soft tissue change data to obtain changed structural feature points;

Step S45: performing position evolution rate analysis on the changed structural feature points to obtain the structural evolution law;

Step S46: Performing tissue morphological structure evolution analysis on the oral multi-time point prediction results according to the structure evolution law to obtain oral morphological evolution prediction data.

In this embodiment, oral models and dental implant structure pressure data are used to perform oral multi-time series node prediction, and time series analysis methods, such as regression analysis, machine learning or deep learning models, are used to establish the correlation between the dental implant structure pressure data and the time nodes, so as to predict the state and changes of the oral cavity at different time points. The bone tissue morphological change analysis is performed on the oral multi-time point prediction results to identify and quantify the morphological changes of the bone tissue. Image processing and analysis methods, such as image registration, segmentation and morphological analysis, are used to compare the oral models at different time points to detect and measure the morphological changes of the bone tissue. The soft tissue change analysis is performed on the oral multi-time point prediction results to identify and quantify the changes of the oral soft tissue. Image processing and analysis methods, such as image registration, segmentation and morphological analysis, are used to compare the oral models at different time points. Oral model, detect and measure the morphological changes of soft tissues, combine bone tissue morphological change data and oral soft tissue change data to identify the changes of structural feature points in the oral model, detect and quantify the changes in the position, shape or other attributes of the structural feature points by comparing the oral models at different time points, and perform position evolution rate analysis on the changed structural feature points to determine the evolution law of the oral structure. Use statistical analysis and pattern recognition methods to compare the changes in the position and attributes of the feature points at different time points, and calculate the rates and trends between them to reveal the law of oral structure evolution. Based on the law of structural evolution, perform tissue morphological structure evolution analysis on the oral multi-time point prediction results, and use interpolation, deformation models or other methods to predict the morphological evolution of the oral cavity at future time points based on the initial state of the oral cavity and the law of structural evolution.

In this embodiment, the specific steps of step S5 are:

Step S51: calculating the maximum deformation area of the oral morphology evolution prediction data to obtain the maximum deformation area of the oral morphology;

Step S52: fitting a time series trend curve for the maximum deformation area of the oral morphology to obtain a morphology evolution time series trend curve;

Step S53: performing boundary buffer area analysis on the first dental implant model based on the morphological evolution time series trend curve to generate an implant deformation buffer zone;

Step S54: performing stress topology optimization on the implant deformation buffer zone to obtain a deformation intelligent buffer zone.

In this embodiment, the oral morphological evolution prediction data is analyzed to find the area where the greatest change occurs during the morphological evolution process. Morphological analysis, point cloud comparison or other image processing methods are used to compare oral models at different time points, calculate the degree of morphological change, and determine the maximum deformation area. A time series analysis is performed on the morphological evolution of the maximum deformation area of the oral morphology, and a time series trend curve is fitted. Regression analysis, curve fitting or other time series analysis methods are used to model the morphological changes of the maximum deformation area at different time points, and a time series trend curve of morphological evolution is generated. The morphological evolution time series trend curve is used to perform a boundary buffer area analysis on the first dental implant model. A buffer area is defined and the boundary position of the implant model is determined according to the time series trend curve of morphological evolution to generate an implant deformation buffer zone. The implant deformation buffer zone is stress topologically optimized. Finite element analysis or other structural optimization methods are used to optimize the structure within the implant deformation buffer zone to improve its stress resistance and stability, thereby obtaining a deformation intelligent buffer zone.

In this embodiment, the specific steps of step S53 are:

Step S531: performing deformation rate matching analysis on the first dental implant model based on the morphological evolution time series trend curve to generate deformation rate data of the dental implant model;

Step S532: analyzing the regional force variation law of the dental implant model deformation rate data to obtain the regional force variation law;

Step S533: performing boundary buffer area analysis based on the regional force variation law to generate an implant deformation buffer area;

Step S54: performing stress topology optimization on the implant deformation buffer zone to obtain a deformation intelligent buffer zone.

In this embodiment, a deformation rate matching analysis is performed on the first dental implant model using a morphological evolution time series trend curve. By comparing the morphological differences of the models at different time points, the deformation rate of each point (i.e., the degree of morphological change) is calculated to generate deformation rate data of the dental implant model. The deformation rate data of the dental implant model is analyzed to find the law of force changes in different regions. By comparing the deformation rate data at different time points, the regions where the force changes significantly are detected and their change laws are determined. Using the regional force change law, a boundary buffer area analysis is performed on the first dental implant model. According to the law of force change, the boundary position of the implant model is determined to generate an implant deformation buffer zone. The implant deformation buffer zone is stress topologically optimized. Finite element analysis or other structural optimization methods are used to optimize the structure in the buffer zone to improve its stress resistance and stability, thereby obtaining a deformation intelligent buffer zone.

In this embodiment, the specific steps of step S6 are:

Step S61: reconstructing the topological shape of the first dental implant model based on the deformation intelligent buffer zone to construct a second dental implant model;

Step S62: performing simulated implantation on the three-dimensional alveolar bone model through the second dental implant model to obtain a holographic alveolar-dental implant model;

Step S63: performing dental implant position recognition on the holographic alveolus-dental implant model to obtain embedding position data;

Step S64: quantifying the distribution of tissue pressure at the edge of the holographic alveolus-dental implant model based on the embedding position data to obtain quantitative tissue pressure data;

Step S65: performing stress area optimization design on the second dental implant model based on the tissue pressure quantification data to construct an optimized dental implant model.

In this embodiment, the information of the deformation intelligent buffer zone is used to perform topological reconstruction on the first dental implant model by changing the shape, structure or topological connection of the model. The shape and structural characteristics of the deformation intelligent buffer zone are introduced into the first dental implant model to generate a second dental implant model with better stress resistance and stability. The second dental implant model is simulated and implanted with the three-dimensional alveolar bone model. The second dental implant model is placed at an appropriate position in the alveolar bone model to generate a holographic alveolar-dental implant model, which represents the relationship between the dental implant and the alveolar bone and includes embedded position information. The position of the dental implant is identified by analyzing and processing the holographic alveolar-dental implant model. This is achieved by calculating the relative position and geometric features of the dental implant and the alveolar bone, by identifying the embedding position, obtaining data on the exact position of the dental implant in the oral cavity, using the embedding position data to quantify the pressure distribution of the adjacent tissues of the holographic alveolar-dental implant model, and by calculating the pressure distribution of the tissues around the dental implant, by quantifying the tissue pressure, obtaining detailed data on the stress conditions of the tissues around the dental implant, and using the tissue pressure quantification data to optimize the design of the stress area of the second dental implant model, and by adjusting the shape, size or material of the dental implant, this is achieved, and through optimized design, the stress distribution of the dental implant in the oral cavity is made more uniform, reducing potential stress concentration areas, and improving its stability and long-term success rate.

Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims (1)

1. The method for forming the dental implant model based on image processing is characterized by comprising the following steps of:

Step S1: acquiring an oral CT image of a patient; carrying out multi-channel convolution treatment on the CT image of the oral cavity of the patient to construct an oral cavity convolution characteristic diagram; performing pixel-level semantic segmentation on the oral convolution feature map to obtain an oral region segmentation map; the specific steps of the step S1 are as follows:

step S11: acquiring an oral CT image of a patient;

step S12: carrying out multi-channel convolution processing on the CT image of the oral cavity of the patient so as to extract cross-band characteristic data;

Step S13: constructing an oral convolution feature map based on the cross-band feature data;

Step S14: carrying out branch sampling treatment on the oral cavity convolution characteristic map to obtain a space detail convolution map; the specific steps of step S14 are:

step S141: performing multi-layer downsampling on the oral cavity convolution feature map to obtain global large-granularity feature data;

Step S142: performing multi-scale up-sampling on the oral cavity convolution feature map so as to obtain local detail feature data;

Step S143: residual connection processing is carried out on the global large-granularity characteristic data and the local detail characteristic data so as to generate missing detail characteristic data;

step S144: obtaining a space detail convolution graph based on the missing detail feature data;

step S15: performing pixel-level semantic segmentation on the space detail convolution map to obtain an oral area segmentation map;

Step S2: generating a tooth missing region map based on the oral region segmentation map; carrying out morphological structure analysis on the tooth missing region graph to obtain tooth groove shape structure data; carrying out three-dimensional construction on the tooth socket shape structure data to generate a three-dimensional tooth socket bone model; the specific steps of the step S2 are as follows:

Step S21: carrying out fine grain structure identification on the oral area segmentation map to obtain a tooth missing area map;

step S22: extracting the tooth socket profile of the tooth missing region graph to obtain tooth socket region profile data;

step S23: carrying out morphological structure analysis on the outline data of the tooth socket area to obtain tooth socket shape structure data;

step S24: performing tooth topology gap calculation on the tooth missing region map to obtain peripheral tooth gap range data;

Step S25: three-dimensional construction is carried out on the tooth socket structural data based on the peripheral tooth gap range data so as to generate a three-dimensional tooth socket bone model;

step S3: performing topological form fitting based on the three-dimensional alveolar bone model to construct a first dental implant model; performing pressure distribution calculation around the dental implant on the first dental implant model to generate dental implant structure pressure data; the specific steps of the step S3 are as follows:

Step S31: performing topological form fitting based on the three-dimensional alveolar bone model to construct a first dental implant model;

step S32: performing elastic deformation dynamics constraint analysis on the first dental implant model to obtain elastic deformation dynamics constraint data;

step S33: performing chewing wear simulation on the elastic deformation dynamics constraint data to obtain wear simulation data;

Step S34: performing pressure distribution calculation around the dental implant based on the wear simulation data to generate dental implant structure pressure data;

step S4: carrying out oral cavity multi-time sequence node prediction based on dental implant structure pressure data to obtain an oral cavity multi-time point prediction result; carrying out tissue morphological structure evolution analysis on the oral cavity multi-time-point prediction result to obtain oral cavity morphological evolution prediction data; the specific steps of the step S4 are as follows:

Step S41: carrying out oral cavity multi-time sequence node prediction based on dental implant structure pressure data to obtain an oral cavity multi-time point prediction result;

Step S42: performing bone tissue morphological change analysis on the oral multi-time-point prediction result to obtain bone tissue morphological change data;

step S43: soft tissue change analysis is carried out on the oral cavity multi-time-point prediction result so as to obtain oral cavity soft tissue change data;

Step S44: performing structural feature point change identification based on the bone tissue morphology change data and the oral cavity soft tissue change data to obtain a change structural feature point;

step S45: analyzing the evolution rate of the part of the feature points of the change structure to obtain a structure evolution rule;

Step S46: carrying out tissue morphological structure evolution analysis on the oral cavity multi-time-point prediction result according to a structure evolution rule so as to obtain oral cavity morphological evolution prediction data;

Step S5: performing time sequence trend curve fitting on the oral cavity form evolution prediction data to obtain a form evolution time sequence trend curve; performing boundary buffer area analysis on the first dental implant model based on the morphological evolution time sequence trend curve to obtain a deformation intelligent buffer area; the specific steps of the step S5 are as follows:

Step S51: calculating the maximum deformation area of the oral morphology evolution prediction data to obtain the maximum deformation area of the oral morphology;

Step S52: performing time sequence trend curve fitting on the maximum deformation area of the oral cavity morphology to obtain a morphology evolution time sequence trend curve;

step S53: performing boundary buffer area analysis on the first dental implant model based on the morphological evolution time sequence trend curve to generate an implant deformation buffer area; the specific steps of step S53 are:

step S531: performing deformation rate matching analysis on the first dental implant model based on the morphological evolution time sequence trend curve to generate dental implant model deformation rate data;

step S532: analyzing the regional stress change rule of the deformation rate data of the dental implant model to obtain the regional stress change rule;

Step S533: performing boundary buffer area analysis based on the area stress change rule to generate an implant deformation buffer area;

step S54: performing stress topology optimization on the implant shape buffer area to obtain a deformation intelligent buffer area;

Step S6: performing simulated implantation on the first dental implant model based on the deformation intelligent buffer zone to obtain a holographic tooth socket-dental implant model; carrying out stress area optimization design based on the holographic tooth socket-dental implant model to construct an optimized dental implant model; the specific steps of the step S6 are as follows:

Step S61: performing topological reconstruction on the first dental implant model based on the deformation intelligent buffer zone to construct a second dental implant model;

Step S62: performing simulated implantation on the three-dimensional alveolar bone model through the second dental implant model to obtain a holographic alveolar-dental implant model;

Step S63: carrying out dental implant position identification on the holographic tooth socket-dental implant model to obtain embedded position data;

step S64: carrying out critical tissue pressure distribution quantification on the holographic tooth socket-dental implant model based on the embedded position data to obtain tissue pressure quantification data;

step S65: and carrying out stress area optimization design on the second dental implant model based on the tissue pressure quantification data so as to construct an optimized dental implant model.