Segmentation and evaluation method for tooth preparation shoulder

Abstract

The invention discloses a segmentation and evaluation method for a tooth preparation shoulder, relates to the technical field of oral medicine digitization, and aims to solve the problems of low segmentation precision, neglect of anatomical distinguishing, single evaluation, unsmooth boundary and the like of an existing method. The method comprises the following steps: shoulder coarse segmentation: screening three-dimensional grid cube units based on an average curvature, and extracting a coarse segmentation shoulder of a maximum connected domain; carrying out shoulder fine segmentation, extracting inner and outer contours, and then carrying out optimization by using a greedy algorithm (ensuring smoothness through inner energy, and enabling outer energy to fit concave-convex characteristics); performing multi-quadrant division, dividing a mesial region, a distal region, a labial region and a lingual region by taking a long axis of a tooth body as a reference, and fitting boundaries by using a B spline; in the evaluation stage, the gingival height and shoulder width of each region are quantified, the weighted average height is calculated to obtain a basic score, and a final score is obtained by combining region balance deduction. The method is high in segmentation precision, achieves differentiated and comprehensive evaluation, is convenient to operate, and is suitable for oral cavity practical operation teaching and false tooth design scenes.

Description

Technical Field

[0001] This invention relates to the field of digital technology in oral medicine, specifically to a method for segmenting and evaluating the shoulder of a prepared tooth. Background Technology

[0002] During tooth preparation, the quality of the prepared shoulder directly affects the retention, stability, periodontal health, and aesthetics of the restoration. Common shoulder problems in clinical practice include poor preparation, inappropriate position and morphological design, discontinuous margins or burrs, and insufficient surface roughness.

[0003] In oral practice teaching, there are two main methods for evaluating the results of students' tooth preparation procedures: one relies on the teacher's experience and judgment, which is highly subjective; the other uses digital evaluation methods, analyzing results through computer algorithms. Existing digital evaluation methods, such as threshold-based segmentation or contour extraction algorithms, are susceptible to data noise and blurred gingival boundaries, making it difficult to accurately and continuously extract the three-dimensional contour of the shoulder, resulting in inaccurate analysis results for shoulder width and height. Furthermore, existing methods have the following drawbacks: Low segmentation accuracy: Traditional threshold-based segmentation or contour extraction algorithms are easily affected by data noise and blurred gingival boundaries, and cannot accurately and continuously extract the complete three-dimensional contour of the shoulder. Ignoring anatomical distinctions: Existing methods typically treat the shoulder as a whole without independently evaluating the functional and aesthetic importance of different anatomical regions, such as the labial (cheeky), lingual (palatal), mesial, and distal sides. Evaluation is too simplistic: Existing methods mainly score based on a single height or single point measurement, lacking comprehensive analysis such as weighting of the overall height distribution of the shoulder platform and regional balance. Unsmooth boundaries: Using multiple straight lines or simple curves to fit the boundaries cannot guarantee the continuity of curvature, affecting the stability and accuracy of the evaluation results.

[0004] Therefore, there is an urgent need for a high-precision, differentiated, and comprehensive method for shoulder segmentation and evaluation to improve the reliability and accuracy of digital evaluation results. Summary of the Invention

[0005] The purpose of this invention is to provide a method for segmenting and evaluating the shoulder of a prepared tooth, comprising the following steps: S1: Coarse segmentation of the shoulder: Calculate the average curvature of the tooth preparation model, traverse the cubic elements in each 3D mesh of the model, delete the cubic elements in the 3D mesh whose average curvature is less than the preset negative value, extract the maximum connected component and subtract it from the original model to obtain the coarsely segmented shoulder. S2: Shoulder refinement segmentation, extracting the inner and outer contours of the coarsely segmented shoulder, and optimizing the inner and outer contours using a greedy algorithm. The greedy algorithm adjusts the points on the contour by minimizing the inner and outer energies, where the inner energy is used to ensure the curve is smooth and the outer energy is used to make the curve fit the concave and convex features of the contour. Then, the optimized inner and outer contours are used to segment the initial tooth model to obtain the finely segmented shoulder. S3: Multi-quadrant division of the shoulder. Based on the long axis of the tooth, the finely divided shoulder is divided into four independent regions: mesial, distal, labial, and lingual. The boundaries of each region are smoothly connected by fitting B-spline curves. S4: Shoulder evaluation, which quantifies the gingival height and shoulder width of each independent region, calculates the weighted average height based on preset weights, calculates the base score by combining the ideal height, and then calculates the regional balance deduction based on the difference between the standard deviation of the height of each region and the threshold. The final score is the base score minus the regional balance deduction.

[0006] Furthermore, in the coarse segmentation of the shoulder, the preset negative value is set according to the material and morphological characteristics of the tooth preparation model.

[0007] Furthermore, the expression for the internal energy is: Where a is the curve length control parameter and b is the smoothness control parameter, set a=0.5, b=0.5.

[0008] Furthermore, the expression for the external energy is: , where c is the average curvature of vertex Vi.

[0009] Furthermore, the greedy algorithm iterates 10 times, traversing each point on the contour, searching for the point with the minimum energy within its local search range, and moving to that position.

[0010] Furthermore, in the multi-quadrant division, the lip side is the aesthetic zone with a weighting coefficient of 1.2; the lingual side is the functional zone with a weighting coefficient of 1.0; the mesial region has a weighting coefficient of 1.1; and the distal region has a weighting coefficient of 1.0.

[0011] Furthermore, the method for quantifying gingival height is as follows: along the long axis of the tooth, samples are taken at equal intervals of 0.5 mm in each independent region, the vertical distance from the gingival apex to the edge of the shoulder is measured at the sampling point, and the average value is taken as the gingival height of that region.

[0012] Furthermore, the formula for calculating the basic score is as follows: ,in For weighted average height, The ideal gingival height is represented by k, which is a proportionality coefficient.

[0013] Furthermore, the formula for calculating the regional balance deduction is as follows: ,in Let be the standard deviation of the gingival height in the i-th region. The allowable standard deviation threshold, This is the deduction intensity coefficient.

[0014] Furthermore, the aforementioned The tolerance is 0.3mm, and 2 points will be deducted for every 0.1mm exceeding this tolerance. Beneficial effects

[0015] 1. High segmentation accuracy: By combining coarse and fine segmentation and optimizing the contour with energy function, it effectively resists the influence of data noise and blurred gingival boundaries, and extracts complete and continuous three-dimensional contours of the shoulder. 2. Divide the area into quadrants according to anatomical regions and assign different weights to achieve targeted evaluation of aesthetic and functional areas; 3. A scoring rule combining weighted average height and regional balance is adopted to comprehensively reflect the overall quality and local balance of the shoulder platform; 4. Easy to operate, requiring no prior conditions, highly automated, and convenient for teachers and students to use; 5. The overall process is adapted to dental teaching and denture design scenarios, with strong practicality and scalability. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the tooth preparation model structure of the present invention; Figure 2 This is a schematic diagram of the algorithm flow of the present invention; Figure 3 This is a schematic diagram of the coarsely segmented inner and outer contour structure of the present invention; Figure 4 This is a schematic diagram of the multi-quadrant division of the shoulder platform according to the present invention; Figure 5 This is a schematic diagram illustrating the quantitative structure of gingival height and width according to the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the following description, in conjunction with the appendix, is provided. Figure 1-5 The present invention will be further described in detail below with reference to specific embodiments.

[0018] I. Overall Process Overview This invention relates to a method for segmenting and evaluating the shoulder of a prepared tooth, comprising four core steps: coarse segmentation of the shoulder (S1), fine segmentation of the shoulder (S2), multi-quadrant division of the shoulder (S3), and shoulder evaluation (S4). Through the process of "coarse segmentation - fine optimization - region division - comprehensive evaluation," it achieves precise segmentation and comprehensive evaluation of the shoulder, making it suitable for dental teaching and prosthesis design scenarios.

[0019] II. Specific Steps and Implementation Details (a) Coarse division of the shoulder (S1) This step aims to initially extract the shoulder area through average curvature screening, laying the foundation for subsequent fine segmentation.

[0020] Data input: Obtain a 3D model of the prepared tooth (e.g., Figure 1 As shown, the model is stored in the form of a cubic cell grid, with each cell containing multiple 3D point coordinates and curvature information.

[0021] Mean curvature calculation: Traverse each cell in the model, calculate the mean curvature of all points within that cell, and obtain the mean curvature value of each cell (denoted as ). ).

[0022] Cube cell filtering: Set a preset negative threshold (denoted as ). Adjust the settings according to the model's material and shape characteristics; for example, use -0.8 for resin models and -1.2 for plaster models. If a certain cell... If the cell is not a shoulder area, it will be determined and deleted.

[0023] Coarse segmentation results: Connectivity analysis is performed on the remaining cells after screening to extract the largest connected region (i.e. the most likely main shoulder area). This is then subtracted from the original tooth preparation model to obtain the coarsely segmented shoulder (the coarse segmentation results retain the general outline of the shoulder, and the edges may have burrs or discontinuities).

[0024] (ii) Detailed division of the shoulder platform (S2) This step extracts and optimizes the inner and outer contours to achieve precise segmentation of the shoulder platform, solving the problem of rough and discontinuous edges in coarse segmentation.

[0025] Inner and outer contour extraction: Extract the inner and outer contours from the coarsely segmented shoulder edge (e.g., Figure 3 As shown in the figure, the inner contour is the boundary of the shoulder near the inside of the tooth body, and the outer contour is the boundary of the shoulder near the gingiva, which serves as the initial contour.

[0026] Contour optimization (greedy algorithm): Energy function definition: Optimization is achieved by minimizing the total energy (internal energy + external energy) at each point on the contour.

[0027] Internal energy : Controls the smoothness and length of the contour, expressed as: ； Where a=0.5 is the length control parameter, the larger the value, the shorter the outline; b=0.5 is the smoothness control parameter, the larger the value, the fewer the jagged edges on the outline.

[0028] External energy E ext( vi) To make the contour conform to the actual concave and convex features of the shoulder platform, the expression is: (c is the average curvature of point Vi. The larger the absolute value of the curvature, the greater the external energy, which pushes the contour to fit into the high curvature area (the edge of the shoulder).

[0029] Iterative optimization process: ① Let the number of points on the contour be n, and perform the operation on each point Vi (i = 0 to n-1); ② Define the local search range of point Vi (a spherical region with a radius of 0.3 mm centered at Vi). ③ Traversal Calculate the total energy for all points Vi within the region. ; ④ Retain the point with the minimum energy Move Vi to Place; ⑤ Repeat the above steps for 10 iterations to obtain the optimized inner and outer contours (smooth and continuous edges).

[0030] Fine segmentation results: The initial tooth model is cut using the optimized inner and outer contours to obtain finely segmented shoulder surfaces (with smooth edges that fit the real shoulder surface shape).

[0031] (III) Multi-quadrant division of the shoulder platform (S3) This step divides the area according to anatomical function to achieve differentiated evaluation.

[0032] Reference point determination: The long axis of the tooth body of the prepared tooth body is used as the reference (the long axis of the tooth body passes through the geometric center of the model and points from the root apex to the crown apex).

[0033] Region division: Along the long axis of the tooth, the finely divided shoulder is divided into four independent regions (e.g., Figure 4 (as shown) The midline region (the side closest to the midline of the face); The distal region (the side furthest from the midline of the face); Labial region (anterior teeth) / buccal region (posterior teeth, aesthetic focus area); Lingual region (anterior teeth) / Palatal region (posterior teeth, functional focus area).

[0034] Boundary smoothing: B-spline curves are used to fit the boundaries of each region. By selecting feature points on the boundary (such as contour inflection points) as control points, the continuity of boundary curvature is ensured, avoiding the sharp corners caused by traditional straight line fitting.

[0035] (iv) Shoulder platform evaluation (S4) This step uses quantitative indicators and weighted analysis to achieve a comprehensive score for the shoulder quality.

[0036] Quantitative parameter measurement: Gingival height quantification: Along the long axis of the tooth, samples are taken at equal intervals of 0.5 mm in each independent region (e.g., 10-15 points in the labial region). The vertical distance from the gingival apex to the shoulder edge at each sampling point is measured (e.g., ...). Figure 5 As shown), the average value is taken as the gingival height H of this region. i (i=1,2,3,4 correspond to the mesial, distal, labial / buccal, and lingual / palatal sides, respectively).

[0037] Shoulder width quantification: Similar to the sampling method above, measure the horizontal width of the shoulder at each sampling point and take the average value as the shoulder width of the area (the calculation logic is the same as the height).

[0038] Weighted average calculation: Weighting coefficients W are set according to the functional importance of the region. i : Lip / cheek side (aesthetic zone): W3=1.2; Near-central region: W1=1.1; Far-middle region: W2=1.0; Tongue / palatal side (functional area): W4=1.0; Weighted average height calculation formula: .

[0039] Basic score calculation: Preset clinical ideal gingival height H ideal， For example, the front teeth H ideal =0.8mm, posterior H ideal =0.5mm, dynamically adjustable); Basic scoring formula: (k is a proportionality coefficient, for example, k=20, which controls the intensity of the impact of deviation on deductions).

[0040] Regional balance deduction calculation: Calculate the standard deviation of gingival height for each region. (Reflecting a high degree of consistency within the region); Set the allowable standard deviation threshold =0.3mm; Deduction formula: , =20, meaning 2 points are deducted for every 0.1mm exceeding the limit. For example... When the thickness is 0.4mm, the deduction is 20 × (0.4 - 0.3) = 2 points.

[0041] Final rating: (Score range: 0-100 points; higher scores indicate better shoulder quality).

[0042] III. Implementation Results Verification The above method was used to test 100 student tooth preparation models, and the results showed: The shoulder segmentation accuracy is improved by 35% compared with the traditional threshold method, and the contour continuity reaches 98%. The evaluation results showed 92% consistency with the manual scoring by senior physicians, significantly reducing subjectivity. The entire process is automated, with a single-case model analysis time of less than 5 seconds, making it suitable for efficient assessment needs in teaching scenarios.

[0043] In the description of this invention, it should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship when the device or element is in normal use. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation at any time, or be constructed and operated in a specific orientation, unless otherwise stated in the text.

[0044] While specific embodiments of the present invention have been described above, those skilled in the art should understand that these are merely illustrative examples, and the scope of protection of the present invention is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of the present invention, but all such changes and modifications fall within the scope of protection of the present invention.

Claims

1. A method for segmenting and evaluating the shoulder of a prepared tooth, characterized in that, Includes the following steps: S1: Coarse segmentation of the shoulder: Calculate the average curvature of the tooth preparation model, traverse the cubic elements in each 3D mesh of the model, delete the cubic elements in the 3D mesh whose average curvature is less than the preset negative value, extract the maximum connected component and subtract it from the original model to obtain the coarsely segmented shoulder. S2: Shoulder refinement segmentation, extracting the inner and outer contours of the coarsely segmented shoulder, and optimizing the inner and outer contours using a greedy algorithm. The greedy algorithm adjusts the points on the contour by minimizing the inner and outer energies, where the inner energy is used to ensure the curve is smooth and the outer energy is used to make the curve fit the concave and convex features of the contour. Then, the optimized inner and outer contours are used to segment the initial tooth model to obtain the finely segmented shoulder. S3: Multi-quadrant division of the shoulder. Based on the long axis of the tooth, the finely divided shoulder is divided into four independent regions: mesial, distal, labial, and lingual. The boundaries of each region are smoothly connected by fitting B-spline curves. S4: Shoulder evaluation, which quantifies the gingival height and shoulder width of each independent region, calculates the weighted average height based on preset weights, calculates the base score by combining the ideal height, and then calculates the regional balance deduction based on the difference between the standard deviation of the height of each region and the threshold. The final score is the base score minus the regional balance deduction.

2. The method according to claim 1, characterized in that, In the coarse segmentation of the shoulder, the preset negative value is set according to the material and morphological characteristics of the tooth preparation model.

3. The method according to claim 1, characterized in that, The expression for the internal energy is: Where a is the curve length control parameter and b is the smoothness control parameter, set a=0.5, b=0.

5.

4. The method according to claim 1, characterized in that, The expression for the external energy is: , where c is the average curvature of vertex Vi.

5. The method according to claim 1, characterized in that, The greedy algorithm iterates 10 times, traversing each point on the contour, searching for the point with the minimum energy within its local search range, and moving to that position.

6. The method according to claim 1, characterized in that, In the multi-quadrant division, the lip side is the aesthetic zone with a weighting coefficient of 1.2; the lingual side is the functional zone with a weighting coefficient of 1.0; the mesial region has a weighting coefficient of 1.1; and the distal region has a weighting coefficient of 1.

0.

7. The method according to claim 1, characterized in that, The method for quantifying gingival height is as follows: along the long axis of the tooth, samples are taken at equal intervals of 0.5 mm in each independent region, the vertical distance from the gingival apex to the edge of the shoulder is measured at the sampling point, and the average value is taken as the gingival height of that region.

8. The method according to claim 1, characterized in that, The formula for calculating the basic score is as follows: ,in For weighted average height, The ideal gingival height is represented by k, which is a proportionality coefficient.

9. The method according to claim 1, characterized in that, The formula for calculating the regional balance deduction is as follows: ,in Let be the standard deviation of the gingival height in the i-th region. The allowable standard deviation threshold, This is the deduction intensity coefficient.

10. The method according to claim 9, characterized in that, The The tolerance is 0.3mm, and 2 points will be deducted for every 0.1mm exceeding this tolerance.

Images