A method for personalized design of dentures

Through the combination of automatic learning based on oral three-dimensional scanning and the U-Net network and gradient elevator, personalized design of dentures is realized, solving the problem of insufficient personalization in traditional design, improving the comfort and aesthetics of dentures, and satisfying the Customized needs of patients.

CN118615037BActive Publication Date: 2024-11-15NANTONG YADI DENTURES CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410897726.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-05
Publication Date
2024-11-15
Estimated Expiration
2044-07-05

AI Technical Summary

Technical Problem

The lack of personalization of traditional denture designs leads to insufficient dimensional matching and natural appearance, the design process is time-consuming and inefficient, making it difficult to fully consider patient needs, affecting comfort and functionality.

Method used

Automatic learning and feature extraction based on oral three-dimensional scanned images are adopted, combined with U-Net network and gradient hoist, geometric parameter measurement and customized design are carried out, and the optimal denture design scheme is generated through iterative optimization and patient feedback adjustment.

Benefits of technology

It improves the comfort, chewing efficiency and aesthetics of dentures, meets patients' needs for customized and functional aesthetics, enhances the automation and intelligence of the design process, and ensures that the design plan meets the actual needs of patients.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118615037B_ABST
    Figure CN118615037B_ABST
Patent Text Reader

Abstract

The present invention relates to the technical field of denture design, specifically a method for personalized denture design, comprising the following steps: S1: based on a three-dimensional scanning image of the oral cavity, detail capture and analysis are performed, oral structure features are automatically learned and extracted, including key structures of teeth and gum lines, and structural identifiers are determined to generate oral structure feature data; S2: based on the oral structure feature data, differential structural components in the oral cavity are identified and segmented. In the present invention, by using a U-Net network and a gradient boosting machine to measure and analyze geometric parameters, the automation and intelligence of the design process are enhanced, the accuracy and efficiency of the design are optimized, the influence of design parameters is analyzed in combination with historical design data and patient feedback, and the design parameters are adjusted accordingly to ensure that the design scheme fits the actual needs of the patient, effectively improving the comfort, chewing efficiency and aesthetics of the denture, and meeting the patient's needs for both customization and functional beauty.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of denture design, and in particular to a method for personalized denture design. Background Art

[0002] The field of denture design technology focuses on developing and improving the design and manufacturing technology of artificial teeth to meet individual differences and specific needs. This field combines multiple disciplines such as biomedical engineering, material science, computer-aided design (CAD) and computer-aided manufacturing (CAM) to create dentures that meet both functional and aesthetic needs. With the development of technology, denture design not only focuses on traditional mechanical properties and durability, but also pays more and more attention to customized design to better meet the oral structure and aesthetic preferences of different patients.

[0003] Among them, the core purpose of the personalized denture design method is to accurately capture the differences in individual oral structure, design and manufacture dentures that perfectly fit the patient's mouth, thereby improving comfort, chewing efficiency and aesthetics. It aims to overcome the limitations of traditional denture design, such as size mismatch or unnatural appearance. The effect achieved is to provide patients with a more customized, functional and aesthetic denture solution.

[0004] Traditional methods lack accuracy in dealing with individual differences, which can easily lead to insufficient size matching and natural appearance. The lack of intelligent design optimization makes the design process time-consuming and inefficient, making it difficult to fully consider patient needs and accurately quantify the geometric parameters of the oral structure, resulting in poor fit between the denture and the oral cavity, affecting comfort and functionality. The lack of effective feedback and data-supported design adjustment process makes it difficult for the design solution to achieve optimal customization and satisfaction. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings in the prior art and to propose a method for personalized design of dentures.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a method for personalized design of dentures, comprising the following steps:

[0007] S1: Based on the oral 3D scan image, it captures and analyzes details, automatically learns and extracts oral structure features, including the key structures of teeth and gum lines, determines structure identifiers, and generates oral structure feature data;

[0008] S2: Based on the oral structure feature data, identify and segment the difference structure components in the oral cavity, determine the boundary between teeth and gums through edge detection and area division, and draw the component outline to obtain structure component recognition data;

[0009] S3: Based on the structural component identification data, a U-Net network is used to measure and analyze geometric parameters, and each component is quantitatively evaluated through size measurement and shape analysis to obtain geometric parameter data;

[0010] S4: Based on the geometric parameter data, a gradient lifting machine is used to construct a customized denture design, and the shape and size of the denture are adjusted and optimized through the design software to match the patient's oral structure, and a preliminary customized design plan is established;

[0011] S5: Based on the preliminary customized design plan, combined with historical design data and patient feedback, perform influence analysis on design parameters, evaluate the contribution of multiple parameters to design quality, and obtain parameter influence analysis records;

[0012] S6: Adjust and optimize the design parameters according to the parameter impact analysis record, refine the design plan through iterative testing and patient feedback information, and obtain the optimal customized design plan.

[0013] The present invention has improvements in that the oral structure feature data includes the number of teeth, a description of the three-dimensional shape of the teeth, the outline of the gum line, and the spatial position relationship between the teeth; the structural component identification data includes the segmented tooth edge image, the gum boundary line image, and the annotation information of the teeth and gums; the geometric parameter data includes the identified tooth length, width, curvature measurement values ​​and shape parameters of the gum line; the preliminary customized design scheme includes the designed basic shape model, adjusted size parameters, and optimized shape characteristics; the parameter impact analysis record includes the parameter ranking affecting the design quality, key optimization parameters, and patient preference adjustment proposals; the optimal customized design scheme includes the optimal denture model, size specifications, and customized shape adjustment details.

[0014] The present invention is improved in that, based on the oral structure feature data, the difference structure components in the oral cavity are identified and segmented, the boundary between the teeth and the gums is determined through edge detection and area division, and the component outline is drawn. The specific steps of obtaining the structure component identification data are:

[0015] S201: Based on the oral structure feature data, image resolution is adjusted to optimize visibility of teeth and gum margins, and boundaries between differentiated oral structures are highlighted to obtain an image set with clear boundaries;

[0016] S202: Based on the boundary-clear image set, a structure segmentation operation is performed to divide the contours of teeth and gums, optimize the visual separation between structures, and generate contour refinement data;

[0017] S203: Based on the contour refinement data, the identified tooth and gum contours are converted into an operable data format, and the teeth and gums are classified and encoded to obtain structural component recognition data.

[0018] The present invention is improved in that, based on the structural component identification data, a U-Net network is used to measure and analyze geometric parameters, and each component is quantitatively evaluated through size measurement and shape analysis. The steps of obtaining geometric parameter data are specifically as follows:

[0019] S301: Based on the structural component identification data, use straight line and curve measurement tools to measure the size of each tooth and gum component, record the length and width, and generate component size data;

[0020] S302: Based on the component size data, using a U-Net network, performing shape analysis on the components, evaluating the curvature and morphological features of each component, and using a shape descriptor to quantify the shape details of the component to obtain shape analysis data;

[0021] S303: Based on the shape analysis data and component size data, the length, width and curvature of the component are uniformly encoded and converted into a standardized expression to obtain geometric parameter data.

[0022] The present invention has the following improvements: the U-Net network adopts the Hu invariant moment formula:

[0023]

[0024] Obtain shape analysis data;

[0025] Among them, H m′ is the invariant moment value of Hu, x is the horizontal coordinate of the pixel on the image, y is the vertical coordinate of the pixel on the image, is the horizontal coordinate of the image centroid, is the ordinate of the image centroid, p and q are the orders of the moments, δ x and δ y is the fine-tuning parameter for the centroid coordinates, α and β are the adjustment coefficients added to the original order, and w(x, y) is the weight coefficient that changes according to the pixel position.

[0026] The present invention is improved in that, according to the geometric parameter data, a gradient lifting machine is used to construct a customized denture design, and the shape and size of the denture are adjusted and optimized by design software to match the patient's oral structure. The steps of establishing a preliminary customized design scheme are specifically as follows:

[0027] S401: Based on the geometric parameter data, the oral structure model is imported using the design software, the model size and shape in the software are adjusted to match the patient's oral structure, and a preliminarily adjusted design model is generated;

[0028] S402: Based on the preliminarily adjusted design model, a gradient boosting machine is used to optimize the model design details, the shape and size of the denture are adjusted according to the geometric parameters, and the matching of the design is iteratively optimized to obtain an optimized design model;

[0029] S403: Based on the optimized design model, a genetic algorithm is used to compare and analyze the patient's oral structure to verify the feasibility of the design solution, and the design solution is adjusted to establish a preliminary customized design solution.

[0030] The present invention is improved in that the gradient boosting machine is according to the formula:

[0031] L(y,f(x,α,β,γ,δ))=(yf(x,α,β,γ,δ)) 2 +λ(α 2 +β 2 +γ 2 +δ 2 )

[0032] Denture design model for computational geometry parameter optimization;

[0033] Among them, L represents the loss function, y is the geometric parameter value of the denture design, f(x, α, β, γ, δ) is the geometric parameter value predicted by the model, α is the geometric shape adjustment coefficient, β is the size adjustment coefficient, γ is the adaptability adjustment coefficient, δ is the material elastic coefficient, and λ is the regularization coefficient.

[0034] The present invention is improved in that, based on the preliminary customized design scheme, combined with historical design data and patient feedback, the influence analysis of design parameters is performed to evaluate the contribution of multiple parameters to the design quality, and the steps of obtaining parameter influence analysis records are specifically as follows:

[0035] S501: Based on the preliminary customized design plan, collect and organize relevant historical design data and current patient feedback information, mark frequently mentioned improvement points in the design, and generate a summary record of improvement points;

[0036] S502: using the summary record of the improvement points, assigning weights to each improvement point, evaluating the potential impact on the design quality, determining the priority of the optimization strategy through comparative analysis, and obtaining an optimization strategy priority list;

[0037] S503: According to the optimization strategy priority list, optimization measures and adjustment plans are formulated, and the influence of design parameters is analyzed to obtain parameter influence analysis records.

[0038] Compared with the prior art, the advantages and positive effects of the present invention are:

[0039] In this invention, U-Net network and gradient boosting machine are used to measure and analyze geometric parameters, which enhances the automation and intelligence of the design process and optimizes the accuracy and efficiency of the design. The influence of design parameters is analyzed by combining historical design data and patient feedback, and the design parameters are adjusted accordingly to ensure that the design plan meets the actual needs of patients, effectively improve the comfort, chewing efficiency and aesthetics of dentures, and meet the needs of patients for both customization and functional beauty. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A flowchart of a method for personalized design of dentures is proposed for the present invention;

[0041] Figure 2 A schematic diagram of the detailed flow of step S1 in a method for personalized denture design is proposed in the present invention;

[0042] Figure 3 A schematic diagram of the detailed process of step S2 in a method for personalized denture design is proposed in the present invention;

[0043] Figure 4 A schematic diagram of a detailed flow chart of step S3 in a method for personalized denture design is provided for the present invention;

[0044] Figure 5 A detailed flow chart of step S4 in a method for personalized denture design is provided for the present invention;

[0045] Figure 6 A detailed flow chart of step S5 in a method for personalized denture design is provided in the present invention;

[0046] Figure 7 The present invention proposes a detailed flow chart of step S6 in a method for personalized denture design. DETAILED DESCRIPTION

[0047] 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.

[0048] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0049] Example

[0050] See also Figure 1 The present invention provides a technical solution: a method for personalized design of dentures, comprising the following steps:

[0051] S1: Based on the oral 3D scan image, it captures and analyzes details, automatically learns and extracts oral structure features, including the key structures of teeth and gum lines, determines structure identifiers, and generates oral structure feature data;

[0052] S2: Based on the oral structure feature data, the differential structural components in the oral cavity are identified and segmented. Through edge detection and area division, the boundary between teeth and gums is determined, and the component outline is drawn to obtain the structural component identification data;

[0053] S3: Based on the structural component identification data, the U-Net network is used to measure and analyze the geometric parameters. Through size measurement and shape analysis, each component is quantitatively evaluated, including length, width and curvature measurement, to obtain geometric parameter data;

[0054] S4: Based on the geometric parameter data, a gradient lifting machine is used to construct a customized denture design. The shape and size of the denture are adjusted and optimized through the design software to match the patient's oral structure and establish a preliminary customized design plan;

[0055] S5: Based on the preliminary customized design plan, combined with historical design data and patient feedback, the influence analysis of design parameters is carried out to evaluate the contribution of multiple parameters to design quality, and an optimization strategy is formulated to obtain parameter impact analysis records;

[0056] S6: Adjust and optimize the design parameters based on the parameter impact analysis records, refine the design plan through iterative testing and patient feedback information, and obtain the optimal customized design plan.

[0057] Oral structure feature data include the number of teeth, three-dimensional shape description of teeth, outline of the gum line, and spatial position relationship between teeth. Structural component identification data include segmented tooth edge images, gum boundary line images, and annotation information of teeth and gums. Geometric parameter data include the identified tooth length, width, curvature measurement values ​​and shape parameters of the gum line. The preliminary customized design plan includes the basic shape model of the design, adjusted size parameters, and optimized shape features. The parameter impact analysis record includes the ranking of parameters affecting design quality, key optimization parameters, and patient preference adjustment proposals. The optimal customized design plan includes the optimal denture model, size specifications, and customized shape adjustment details.

[0058] See also Figure 2 Based on the oral 3D scan image, the details are captured and analyzed, and the oral structure features are automatically learned and extracted, including the key structures of teeth and gum lines, and the structure identifiers are determined. The specific steps for generating oral structure feature data are as follows:

[0059] S101: Based on the oral 3D scan image, the image acquisition process is optimized, and the surface details of the teeth are covered by changing the scanning angle and lighting conditions to obtain the best image quality and generate an oral image dataset;

[0060] S102: performing image processing operations based on the oral image data set, including color correction and contrast adjustment of the image, optimizing the visual contrast of teeth and gum lines, and obtaining an optimized oral image set;

[0061] S103: Based on the optimized oral image set, perform structure marking, use image editing tools to manually outline the teeth and gum margins, record the shape, size and relative position, and generate oral structure feature data.

[0062] S101: Based on the oral 3D scanning images, the dynamic lighting adjustment method is used to control the lighting intensity and angle through the program to cover the tooth surface details, obtain the optimal image quality, and generate an oral image dataset. The OpenCV library is used to apply adaptive threshold processing to the image dataset, adjust the image lighting uniformity, and generate an optimized oral image dataset.

[0063] S102: An image enhancement algorithm was used to perform color correction and contrast adjustment on the oral image dataset using the Python image processing library PIL. Histogram equalization technology was applied to optimize the visual contrast of teeth and gum lines, and the image was sharpened to improve detail clarity, and an optimized oral image set was generated.

[0064] S103: A semi-automatic image annotation tool is used to perform structural labeling on the optimized oral image set using image processing technology. Through image segmentation algorithms, specifically segmentation methods based on region growing, the teeth and gum margins are automatically identified, allowing the user to fine-tune the edge position, record the shape, size and relative position, and generate oral structure feature data.

[0065] See also Figure 3 Based on the oral structure feature data, the difference structure components in the oral cavity are identified and segmented. Through edge detection and area division, the boundary between teeth and gums is determined, and the component outline is drawn. The specific steps to obtain the structure component identification data are as follows:

[0066] S201: Based on the oral structure feature data, image resolution is adjusted to optimize the clarity of teeth and gum margins, highlight the boundary between differentiated oral structures, and obtain an image set with clear boundaries;

[0067] S202: Based on the image set with clear boundaries, a structure segmentation operation is performed to divide the contours of teeth and gums, optimize the visual separation between structures, and generate contour refinement data;

[0068] S203: Based on the contour refinement data, the identified tooth and gum contours are converted into an operable data format, and the teeth and gums are classified and encoded to obtain structural component recognition data.

[0069] S201: Based on the oral structure feature data, the image resolution adjustment technology is used, and the bilinear interpolation method is applied through the image processing library OpenCV to perform image scaling, optimize the clarity of the teeth and gum margins, highlight the dividing lines between differentiated oral structures, and obtain a set of images with clear boundaries. The Gaussian fuzzy algorithm is used to reduce image noise and smooth image edges.

[0070] S202: Based on a set of images with clear boundaries, the Sobel edge detection algorithm is used, and the 0penCV library is used to perform edge detection. By setting a threshold to filter strong edges, a structure segmentation operation is performed to divide the contours of teeth and gums, optimize the visual separation between structures, generate contour refinement data, and apply corrosion and dilation operations to optimize edge clarity.

[0071] S203: Based on the contour refinement data, the identified tooth and gum contours are converted into SVG format using vectorization technology, and the image editing software Inkscape is used to operate, the teeth and gums are classified and encoded, and XML tags are used for structural description to obtain structural component recognition data.

[0072] See also Figure 4Based on the structural component identification data, the U-Net network is used to measure and analyze the geometric parameters. Through size measurement and shape analysis, each component is quantitatively evaluated, including length, width and curvature measurement. The specific steps for obtaining geometric parameter data are as follows:

[0073] S301: Based on the structural component identification data, use straight line and curve measurement tools to measure the size of each tooth and gum component, record the length and width, and generate component size data;

[0074] S302: Based on the component size data, a U-Net network is used to perform shape analysis on the components, the curvature and morphological characteristics of each component are evaluated, and shape descriptors are used to quantify the shape details of the components to obtain shape analysis data;

[0075] S303: Based on the shape analysis data and component size data, the length, width and curvature of the component are uniformly encoded and converted into a standardized expression to obtain geometric parameter data.

[0076] S301: Based on the structural component identification data, use straight line and curve measurement tools, operate through the size tool of Photoshop software, measure the size of each tooth and gum component, record the length and width, generate component size data, use the ruler tool to measure the straight line distance, and the arc tool to calculate the curve length.

[0077] S302: Based on the component size data, the U-Net network is used to build a deep learning model using the Python programming environment and the TensorFlow framework to perform shape analysis on the components, evaluate the curvature and morphological characteristics of each component, train the network through the training set, set the loss function to the cross entropy loss function, select the Adam optimizer as the optimizer, use the shape descriptor, specifically the Hu invariant moment, to quantify the shape details of the component and obtain the shape analysis data.

[0078] U-Net network uses Hu invariant moment formula:

[0079]

[0080] Obtain shape analysis data;

[0081] Among them, H m′ is the invariant moment value of Hu, x is the horizontal coordinate of the pixel on the image, y is the vertical coordinate of the pixel on the image, is the horizontal coordinate of the image centroid, is the ordinate of the image centroid, p and q are the orders of the moments, δ x and δ yis the fine-tuning parameter for the centroid coordinates, α and β are the adjustment coefficients added to the original order, and w(x, y) is the weight coefficient that changes according to the pixel position.

[0082] Execution process:

[0083] First, calculate the horizontal and vertical coordinates of the image centroid and This is achieved by weighted averaging the original pixel coordinates. Then, a weight coefficient w(x, y) is introduced for each pixel position. This coefficient varies according to the distance of the pixel position relative to the image centroid to enhance the influence of the center or edge area of ​​the image, plus a fine-tuning parameter δ for the centroid coordinates x and δ y , the position of the image centroid is slightly adjusted to adapt to the analysis requirements of specific shape features. At the same time, the flexibility and adaptability are further increased by adding adjustment coefficients α and β to the orders p and q of the original moments respectively. The exact value of the weight coefficient w(x, y) can be determined by preprocessing analysis of the image. For example, the weight can be adjusted according to the contrast of the image or the frequency of occurrence of a specific shape to ensure that the weight distribution matches the shape characteristics of the component to be analyzed.

[0084] S303: Based on the shape analysis data and component size data, unified coding technology is adopted, and Python script is used to encode the length, width and curvature, which are converted into JSON format for standardized expression, and the geometric parameter data is obtained. The data is organized in key-value pairs to ensure that the data of each component is easy to access and parse.

[0085] See also Figure 5 According to the geometric parameter data, the gradient lifting machine is used to construct the customized denture design. The shape and size of the denture are adjusted and optimized through the design software to match the patient's oral structure. The specific steps for establishing a preliminary customized design plan are as follows:

[0086] S401: Based on the geometric parameter data, the oral structure model is imported using the design software, the size and shape of the model in the software are adjusted to match the patient's oral structure, and a preliminarily adjusted design model is generated;

[0087] S402: Based on the initially adjusted design model, a gradient boosting machine is used to optimize the model design details, the shape and size of the denture are adjusted according to the geometric parameters, and the design adaptability is iteratively optimized to obtain the optimized design model;

[0088] S403: Based on the optimized design model, a genetic algorithm is used to compare and analyze the patient's oral structure to verify the practicality of the design plan, and the design plan is adjusted to establish a preliminary customized design plan.

[0089] S401: Based on the geometric parameter data, the oral structure model is imported using the design software SolidWorks. The model size and shape in the software are adjusted to match the patient's oral structure. Through the parametric design function, the length and width values ​​are input to adjust the model size and generate a preliminary adjusted design model. The geometric shape of the teeth and gums is refined using 3D modeling tools.

[0090] S402: Based on the preliminary adjusted design model, the gradient boosting machine is used. The gradient boosting machine algorithm is implemented using the Python programming environment and the Scikit-learn library to optimize the model design details. The shape and size of the denture are adjusted according to the geometric parameters. The loss function is set to the least squares loss, the optimizer is set to the stochastic gradient descent, and the design adaptability is iteratively optimized. The model is trained with the geometric parameter data to obtain the optimized design model, and the denture shape is adjusted to adapt to the specific geometric characteristics of the oral structure.

[0091] Gradient boosting machine, according to the formula:

[0092] L(y,f(x,α,β,γ,δ))=(yf(x,α,β,γ,δ)) 2 +λ(α 2 +β 2 +γ 2 +δ 2 )

[0093] Denture design model for computational geometry parameter optimization;

[0094] Among them, L represents the loss function, y is the geometric parameter value of the denture design, f (x, α, β, γ, δ) is the geometric parameter value predicted by the model, α is the geometric shape adjustment coefficient, β is the size adjustment coefficient, γ is the adaptability adjustment coefficient, δ is the material elastic coefficient, and λ is the regularization coefficient, which controls the balance of the influence of factors on the model.

[0095] Execution process:

[0096] Firstly, the factors including geometry adjustment coefficient, size adjustment coefficient, adaptability adjustment coefficient and material elasticity coefficient are defined. The factors are added to the model prediction function through α, β, γ and δ parameters. Considering the complexity and overfitting of the model, λ is introduced as the regularization coefficient. Its specific value is determined through cross-validation to achieve a balance between the generalization ability of the model and the prediction accuracy. The stochastic gradient descent optimizer adjusts the model parameters according to the formula. Taking into account the influence of the parameters, the geometric parameter optimization of the denture design is realized.

[0097] S403: Based on the optimized design model, genetic algorithm is applied, using MATLAB environment and genetic algorithm toolbox, and comparative analysis is performed with the patient's oral structure to verify the practicality of the design solution. The population size, crossover rate and mutation rate are set to configure the algorithm. The matching degree of the oral structure is used as the fitness function to iteratively search for the optimal solution, adjust the design solution, establish a preliminary customized design solution, and use genetic algorithm to optimize parameters to ensure that the design solution is most suitable for the patient's oral cavity.

[0098] See also Figure 6 Based on the preliminary customized design plan, combined with historical design data and patient feedback, the influence analysis of design parameters is carried out to evaluate the contribution of multiple parameters to design quality and formulate optimization strategies. The specific steps to obtain parameter impact analysis records are as follows:

[0099] S501: Based on the preliminary customized design plan, collect and organize relevant historical design data and current patient feedback information, mark the improvement points frequently mentioned in the design, and generate a summary record of the improvement points;

[0100] S502: using the summary record of the improvement points, assigning weights to each improvement point, evaluating the potential impact on the design quality, determining the priority of the optimization strategy through comparative analysis, and obtaining a priority list of the optimization strategies;

[0101] S503: According to the optimization strategy priority list, formulate optimization measures and adjustment plans, verify the pertinence of each strategy on the design quality, analyze the influence of design parameters, and obtain parameter influence analysis records.

[0102] S501: Based on the preliminary customized design plan, data collection and organization tools are used, and the Python programming environment and Pandas library are used to process data. The relevant historical design data and current patient feedback information are collected and organized, and the frequently mentioned improvement points in the design are marked. A summary record of the improvement points is generated, and the improvement points with higher frequency are identified through text analysis methods.

[0103] S502: Utilize the summary records of improvement points, adopt the weight allocation method, use Excel to perform data analysis, assign weights to each improvement point, evaluate the potential impact on design quality, determine the priority of the optimization strategy through comparative analysis, obtain the optimization strategy priority list, apply the VL00KUP function to match the improvement points with the corresponding weights, and use the sorting function to determine the priority of the optimization strategy.

[0104] S503: According to the optimization strategy priority list, formulate optimization measures and adjustment plans, use decision tree analysis method, implement decision tree algorithm through Python programming environment and Scikit-learn library, verify the pertinence of each strategy on design quality, and analyze the influence of design parameters to obtain parameter impact analysis records, set the tree depth and branch criterion as information gain to evaluate the influence of parameters on design quality.

[0105] See also Figure 7 , according to the parameter impact analysis records, adjust and optimize the design parameters, refine the design plan through iterative testing and patient feedback information, and obtain the optimal customized design plan. The specific steps are as follows:

[0106] S601: Based on the parameter impact analysis record, identify the design parameters that need to be adjusted, use the design software to manually adjust the parameters, ensure that each adjustment is based on patient feedback and test results, and generate a draft of the adjusted design plan;

[0107] S602: Iteratively test the adjusted design draft to simulate its performance in actual use, collect feedback information during the test, and continue to adjust the design to improve adaptability and comfort, so as to obtain an iterative optimized design solution;

[0108] S603: Based on the iterative optimization design plan, the design plan is presented to patients, user feedback is collected, fine-tuned according to the feedback, and the design plan is verified to meet the patient's needs and expectations to obtain the best customized design plan.

[0109] S601: Based on the parameter impact analysis records, identify the design parameters that need to be adjusted, use the design software SolidWorks to manually adjust the parameters, ensure that each adjustment is based on patient feedback and test results, use the parametric design function in the software to input the size and shape parameters that need to be adjusted, generate a draft of the adjusted design plan, and use 3D modeling tools to accurately adjust the design parameters.

[0110] S602: Conduct iterative testing on the adjusted draft design, use the finite element analysis (FEA) software ANSYS to simulate its performance in actual use, collect feedback during the testing process, and continue to adjust the design to improve fit and comfort. By setting appropriate material properties and boundary conditions, evaluate the structural strength and durability of the design, obtain an iteratively optimized design, and use the simulation results to guide further design adjustments.

[0111] S603: Based on iterative optimization of the design plan, the plan is demonstrated to the patients, and user feedback is collected using virtual reality (VR) technology. The design plan is displayed through virtual reality devices such as 0culus Rift, allowing patients to preview the customized design plan in a virtual environment, and user feelings and suggestions during use are collected. Fine-tuning is performed based on feedback, and it is verified whether the design plan meets the patient's needs and expectations. The collected feedback information is used to refine the design adjustments to obtain the optimal customized design plan, ensuring that the design plan meets the patient's specific needs to the greatest extent.

[0112] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for personalized denture design, characterized in that: The following steps are involved: Based on the oral 3D scan image, it can capture and analyze details, automatically learn and extract oral structure features, including the key structures of teeth and gum lines, determine the structure identifiers, and generate oral structure feature data; Based on the oral structure feature data, the difference structure components in the oral cavity are identified and segmented, the boundary between the teeth and the gums is determined through edge detection and area division, and the component outline is drawn to obtain the structure component identification data; Based on the structural component identification data, a U-Net network is used to measure and analyze geometric parameters, and each component is quantitatively evaluated through size measurement and shape analysis to obtain geometric parameter data; Based on the geometric parameter data, a gradient lifting machine is used to construct a customized denture design, and the shape and size of the denture are adjusted and optimized through the design software to match the patient's oral structure, and a preliminary customized design plan is established; Based on the preliminary customized design plan, combined with historical design data and patient feedback, the influence analysis of design parameters is performed to evaluate the contribution of multiple parameters to design quality and obtain parameter influence analysis records; According to the parameter impact analysis records, adjust and optimize the design parameters, refine the design plan through iterative testing and patient feedback information, and obtain the optimal customized design plan; Based on the structural component identification data, the U-Net network is used to measure and analyze the geometric parameters. Through size measurement and shape analysis, each component is quantitatively evaluated. The steps for obtaining the geometric parameter data are as follows: Based on the structural component identification data, using straight line and curve measurement tools to measure the dimensions of each tooth and gum component, record the length and width, and generate component dimension data; Based on the component size data, a U-Net network is used to perform shape analysis on the components, the curvature and morphological characteristics of each component are evaluated, and shape descriptors are used to quantify the shape details of the components to obtain shape analysis data; Based on the shape analysis data and component size data, uniformly encode the length, width and curvature of the component, convert them into standardized expressions, and obtain geometric parameter data; The U-Net network adopts the Hu invariant moment formula: ; Obtain shape analysis data; in, is the invariant moment value of Hu, is the horizontal coordinate of the pixel on the image, is the vertical coordinate of the pixel on the image, is the horizontal coordinate of the image centroid, is the ordinate of the image centroid, and is the order of the moment, and is the fine-tuning parameter for the centroid coordinates, and is the adjustment factor added to the original order, is the weight coefficient that changes according to the pixel position; According to the geometric parameter data, a gradient lifting machine is used to construct a customized denture design. The shape and size of the denture are adjusted and optimized through the design software to match the patient's oral structure. The specific steps for establishing a preliminary customized design plan are as follows: Based on the geometric parameter data, the oral structure model is imported using the design software, and the size and shape of the model in the software are adjusted to match the patient's oral structure to generate a preliminarily adjusted design model; Based on the preliminarily adjusted design model, a gradient boosting machine is used to optimize the model design details, the shape and size of the denture are adjusted according to the geometric parameters, and the matching of the design is iteratively optimized to obtain an optimized design model; Based on the optimized design model, a genetic algorithm is used to compare and analyze the patient's oral structure to verify the feasibility of the design solution, and the design solution is adjusted to establish a preliminary customized design solution; The gradient boosting machine, according to the formula: ; Denture design model for computational geometry parameter optimization; in, represents the loss function, Design geometric parameter values ​​for dentures, are the geometric parameter values ​​predicted by the model, is the geometry adjustment factor, is the size adjustment factor, is the adaptability adjustment coefficient, is the elastic coefficient of the material, is the regularization coefficient.

2. The method for personalized denture design according to claim 1, characterized in that: The oral structure feature data includes the number of teeth, the three-dimensional shape description of the teeth, the outline of the gum line, and the spatial position relationship between the teeth. The structural component identification data includes the segmented tooth edge image, the gum boundary line image, and the annotation information of the teeth and gums. The geometric parameter data includes the identified tooth length, width, curvature measurement values ​​and shape parameters of the gum line. The preliminary customized design plan includes the basic shape model of the design, the adjusted size parameters, and the optimized shape characteristics. The parameter impact analysis record includes the parameter ranking affecting the design quality, the key optimization parameters, and the patient preference adjustment proposal. The optimal customized design plan includes the optimal denture model, size specifications, and customized shape adjustment details.

3. The method for personalized denture design according to claim 1, characterized in that: Based on the oral structure feature data, the steps of identifying and segmenting the difference structure components in the oral cavity, determining the boundary between the teeth and the gums through edge detection and area division, and drawing the component outline to obtain the structure component identification data are as follows: Based on the oral structure feature data, image resolution is adjusted to optimize visibility of teeth and gum margins, highlight boundaries between differentiated oral structures, and obtain an image set with clear boundaries; Based on the boundary-defined image set, a structure segmentation operation is performed to divide the contours of teeth and gums, optimize the visual separation between structures, and generate contour refinement data; Based on the contour refinement data, the identified tooth and gum contours are converted into an operable data format, and the teeth and gums are classified and encoded to obtain structural component identification data.

4. The method for personalized denture design according to claim 1, characterized in that: Based on the preliminary customized design plan, combined with historical design data and patient feedback, the influence analysis of design parameters is carried out to evaluate the contribution of multiple parameters to design quality. The specific steps for obtaining parameter influence analysis records are as follows: Based on the preliminary customized design plan, collect and organize relevant historical design data and current patient feedback information, mark frequently mentioned improvement points in the design, and generate a summary record of improvement points; Using the summary records of the improvement points, weights are assigned to each improvement point to evaluate the potential impact on the design quality, and the priority of the optimization strategy is determined through comparative analysis to obtain a priority list of the optimization strategies; According to the optimization strategy priority list, optimization measures and adjustment plans are formulated, and the influence of design parameters is analyzed to obtain parameter influence analysis records.