Standard evaluation method and system for scanning accuracy of intraoral three-dimensional scanner

Abstract

The present application relates to the technical field of medical devices, in particular to a standard evaluation method and system for scanning accuracy of an intraoral three-dimensional scanner, comprising the following steps: S1, constructing a standardized oral dynamic model: fixing a standardized three-dimensional entity model; S2, scanning data acquisition: recording the motion trajectory and motion speed data of the scanning probe; S3, registration and alignment of scanning data and reference model: outputting a registration error distribution map; S4, multi-dimensional accuracy analysis: calculating the profile coincidence degree of the tooth preparation edge region, the occlusal surface curvature matching degree and the adjacent area point cloud density deviation value; S5, dynamic error compensation evaluation: generating a dynamic error compensation coefficient; S6, generating a comprehensive evaluation index: outputting a comprehensive accuracy score. The present application can improve the matching accuracy of intraoral three-dimensional scanning data and the stability of the measurement system, and provide a higher precision and more stable standard evaluation system for digital diagnosis and treatment of the oral cavity and restoration body manufacturing.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, and in particular to a standard evaluation method and system for scanning accuracy of an intraoral 3D scanner. Background Technology

[0002] With the rapid development of digital oral diagnosis and treatment technology, intraoral 3D scanners have become an important tool in dental restoration, orthodontics, and implantology. Compared with traditional impression methods, 3D scanning technology can acquire high-precision 3D data of intraoral structures in real time, improving patient comfort and optimizing the prosthesis fabrication process. The accuracy of intraoral scanners directly affects the fit and occlusal relationship of prostheses. Therefore, how to scientifically and objectively evaluate the measurement accuracy of scanners and ensure their stability and consistency in complex oral environments has become an important research direction in the field of digital oral medicine. Currently, the evaluation of scanning accuracy usually relies on static models or clinical comparative experiments, lacking a standardized and quantifiable accuracy evaluation system. This leads to significant deviations in measurement results under different equipment and experimental environments, which is not conducive to the optimization and improvement of scanning technology.

[0003] Currently, the accuracy evaluation of intraoral 3D scanners mainly relies on static standard model testing, but it does not fully consider the biomechanical characteristics and dynamic humidity changes inside the oral cavity, resulting in a significant gap between the measurement environment and actual clinical applications. In addition, existing accuracy analysis methods usually only focus on the overall point cloud error, while ignoring the error distribution of key areas such as tooth preparation margins, occlusal surface curvature matching, and point cloud density in adjacent areas, making it difficult to meet the requirements for high-precision restoration fabrication. At the same time, the impact of probe scanning speed and its fluctuations on error is not adequately compensated, and the stability of measurement results is highly uncertain under different scanning operation conditions. Furthermore, most current evaluation methods use simple error mean calculations, failing to comprehensively consider the interaction between multi-dimensional indicators, resulting in insufficient reliability and clinical applicability of the evaluation results.

[0004] To address the aforementioned issues, this invention proposes a standard evaluation method and system for the scanning accuracy of intraoral 3D scanners, ensuring the objectivity, repeatability, and clinical applicability of scanner accuracy evaluation, and providing a standardized reference for digital oral diagnosis and treatment and prosthesis fabrication. Summary of the Invention

[0005] This invention provides a standard evaluation method and system for scanning accuracy of intraoral 3D scanners.

[0006] A standard method for evaluating the scanning accuracy of an intraoral 3D scanner includes the following steps: S1, Construct a standardized oral dynamic model: On a biomimetic gingival base with biomechanical properties, fix a standardized three-dimensional solid model including tooth preparation body, adjacent teeth and opposing teeth, and set a programmable humidity control layer on the surface of the standardized three-dimensional solid model. S2, Scanning data acquisition: The standardized three-dimensional solid model is scanned from multiple angles using the intraoral scanner under test, and the motion trajectory and speed data of the scanning probe are recorded simultaneously. S3, Registration and Alignment of Scanned Data with Reference Model: An adaptive registration algorithm based on feature point cloud is used to align the obtained scanned data with the reference oral digital model in spatial coordinates and output a registration error distribution map; S4, Multi-dimensional accuracy analysis: Based on the output registration error distribution map, calculate the contour overlap of the tooth preparation edge area, the occlusal surface curvature matching degree, and the point cloud density deviation value of the adjacent area. S5, Dynamic Error Compensation Assessment: Based on the recorded probe movement speed data, establish a correlation model between scanning speed and contour overlap, and generate dynamic error compensation coefficients. S6, Comprehensive evaluation index generation: Weighted fusion of contour overlap, curvature matching degree, density deviation value and dynamic error compensation coefficient to output a comprehensive accuracy score.

[0007] Optionally, constructing a standardized oral dynamic model in S1 includes: S11, Biomimetic Gingival Substrate Preparation: A biomimetic gingival substrate with biomechanical properties was prepared using a silicone rubber and polyurethane composite material, and its elastic modulus was calculated. ; S12, 3D solid model fixation: The standardized 3D solid models of the prepared tooth, adjacent teeth, and opposing teeth are arranged according to the clinical dental arch morphology and embedded into the biomimetic gingival base using positioning pins. The positioning error meets the following conditions: ,in, This represents the maximum displacement deviation between the model and the base. S13, Humidity Control Layer Construction: A programmable humidity control layer is set on the surface of the 3D solid model, with a surface contact angle of... Adjusted via microfluidic channels.

[0008] Optionally, the scan data acquisition in S2 includes: S21, Multi-angle scanning path planning: Set the starting angle of the scanning probe and adjust it sequentially according to fixed angle increments; S22, Synchronous recording of motion trajectory: Real-time acquisition of three-dimensional spatial coordinates via an inertial measurement unit (IMU) mounted on the scanning probe. and record timestamps The sampling frequency meets the condition as follows: ; S23, Motion speed calculation: based on continuous timestamps and Calculate instantaneous velocity from coordinate data. ; S24, Abnormal Data Marker: For instantaneous velocity Exceeding the threshold range The scanned segments were marked, among which , ; S25, Data Synchronization Encapsulation: Align scan data, motion trajectory coordinates, and velocity data according to timestamps to generate structured data packets. .

[0009] Optionally, the registration and alignment of the scan data with the reference model in S3 includes: S31, Feature point cloud extraction: Extract tooth anatomical feature point clouds from the scan data and the benchmark oral digital model respectively; S32, Coarse Registration Alignment: Initial registration is performed based on principal component analysis (PCA) of the feature point cloud, and the transformation matrix is ​​calculated; S33, Fine Registration Optimization: Iterative Closest Point (ICP) algorithm is used for precise registration; S34, Error Distribution Map Generation: Calculate the Euclidean distance error between each registration point and the matching point of the baseline oral digital model after fine registration optimization. and the error Mapped to the HSV color space, a two-dimensional error heatmap and a three-dimensional point cloud error distribution map are generated, with blue representing... Red indicates .

[0010] Optionally, the feature point cloud extraction in S31 includes: S311, Tooth Region Segmentation: Using a semantic segmentation network, the scanned data and a baseline digital oral model are used to segment teeth by position, outputting a set of independent points including anterior teeth, premolars, and molars. The segmentation accuracy meets the requirements. ,in, To determine the overlap of the segmented regions; S312, Local curvature calculation: For points in each partitioned point cloud... Based on its neighborhood radius Calculating curvature values ​​from the eigenvalues ​​of the covariance matrix of interior points ; S313, Curvature Threshold Filtering: Set Differentiated Curvature Thresholds Based on Tooth Position , retain satisfaction The feature points constitute the initial feature point cloud. ; S314, Morphological optimization: Initial feature point set Perform 3D morphological closing operations to eliminate discrete noise points and obtain a morphologically optimized feature point cloud. ; S315, Curvature Gradient Enhancement: Optimization of Morphologically Optimized Feature Point Clouds Curvature gradient constraints are applied to eliminate gradient abrupt changes and outliers, ultimately outputting a set of anatomical feature points. .

[0011] Optionally, the tooth region segmentation in S311 includes: S3111, 3D point cloud data preprocessing: normalizing the scanned data and the benchmark oral digital model; S3112, Semantic Segmentation Network Inference: The PointNet++ semantic segmentation network is used to classify tooth positions on the normalized point cloud data and output the category probability. S3113, Post-processing of segmentation results: Perform 3D morphological closing operations on the segmentation results output by the semantic segmentation network to obtain morphologically optimized tooth position segmentation labels. ; S3114, Tooth Position Mapping and Point Cloud Extraction: Based on Morphologically Optimized Tooth Position Segmentation Labels Extracting the independent point clusters of each tooth position .

[0012] Optionally, the multi-dimensional precision analysis in S4 includes: S41, Tooth preparation edge contour extraction: In the registration error distribution map, a 0.5mm wide strip area is extracted along the normal direction of the tooth preparation edge, and B-spline curve fitting is performed on the reference oral digital model and the scanning data respectively. S42, Contour overlap calculation: Calculate the mean Hausdorff distance between the B-spline curves of the baseline digital oral model and the tooth preparation margins in the scanned data. ; S43, Calculation of occlusal surface curvature matching degree: Randomly select occlusal surface area Using a sampling point, the relative curvature error between the baseline oral digital model and the scan data is calculated. ; S44, Calculation of point cloud density deviation in the adjacent region: Within the adjacent region (within ±1 mm of the contact point), calculate the ratio of the point cloud density of the scanned data to that of the reference oral digital model. ; S45, Three-dimensional error vector synthesis: The overlap of the contour of the tooth preparation edge area ( ), Occlusal surface curvature matching degree ( ) and the point cloud density deviation value of the adjacent area ( Convert to normalized error vector .

[0013] Optionally, the dynamic error compensation assessment in S5 includes: S51, Speed ​​Data Segmentation Processing: The probe motion speed data is segmented according to the scanning path, with each segment lasting for a specified duration. Calculate the average speed of each segment. and speed fluctuation standard deviation ; S52, Speed-Error Correlation Modeling: Establishing an exponential decay model for the relationship between scanning speed and contour overlap error; S53, Dynamic Compensation Coefficient Generation: Calculate the compensation coefficient based on the correlation model. ; S54, Velocity Distribution Weight Allocation: The compensation coefficients are weighted and fused according to the velocity range to generate global compensation coefficients. ; S55, Compensation Effect Verification: Verify the global compensation coefficient. The effect of compensation was verified by applying it to the original contour overlap.

[0014] Optionally, the generation of comprehensive evaluation indicators in S6 includes: S61, Index Normalization Processing: The contour overlap, curvature matching degree, density deviation value and dynamic error compensation coefficient will be normalized. S62, Dynamic allocation of tooth position weights: Differentiated weights are set according to tooth regions. ; S63, Nonlinear Weighted Fusion: An exponential decay function is used for weighted calculation to obtain the comprehensive evaluation score. ; S64, Boundary Constraint Correction: Corrects scores for special scenarios to obtain the final comprehensive evaluation score. .

[0015] A standard evaluation system for scanning accuracy of an intraoral 3D scanner, used to implement the aforementioned standard evaluation method for scanning accuracy of an intraoral 3D scanner, includes the following modules: Standardized oral dynamic model module: Based on a biomimetic gingival base, a standardized three-dimensional solid model including tooth preparation, adjacent teeth and opposing teeth is constructed, and a programmable humidity control layer is set; Scanning data acquisition module: Performs multi-angle scanning through the scanner inside the test port, and simultaneously records the motion trajectory and speed data of the scanning probe; Data registration and alignment module: It adopts an adaptive registration algorithm based on feature point cloud to align the scanned data with the reference oral digital model in spatial coordinates and outputs a registration error distribution map; Multi-dimensional accuracy analysis module: Calculates the contour overlap, occlusal surface curvature matching degree, and point cloud density deviation value of the tooth preparation edge area based on the registration error distribution map; Dynamic error compensation evaluation module: Based on probe motion speed data, establish a correlation model between scanning speed and contour overlap, and generate dynamic error compensation coefficients; The comprehensive evaluation index generation module performs weighted fusion of contour overlap, curvature matching, density deviation, and dynamic error compensation coefficient, and outputs a comprehensive accuracy score.

[0016] The beneficial effects of this invention are: This invention effectively improves the reliability and consistency of intraoral three-dimensional scanning accuracy evaluation by constructing a standardized oral dynamic model. It adopts a biomimetic gingival base and adjusts the elastic modulus by using silicone rubber and polyurethane composite materials to make its physical properties close to those of real oral tissue, simulating a real biomechanical environment. In addition, the embedded fixing method of the positioning pin can minimize the displacement deviation between the model and the base, improving the stability of experimental repeatability. At the same time, the use of microfluidic channels to achieve programmable humidity control allows for precise simulation of dynamic humidity changes in the oral environment, optimizing the realism and clinical applicability of scanning accuracy evaluation.

[0017] This invention ensures uniform coverage of the scanning spot by planning the scanning path from multiple angles, reducing data omissions and improving measurement integrity. Simultaneously, through feature point cloud extraction, adaptive registration algorithms, and multi-dimensional accuracy analysis, it achieves high-precision matching and error visualization between the scanning data and the benchmark model. Feature point cloud extraction utilizes semantic segmentation and curvature filtering mechanisms to improve matching stability. Coarse registration combined with principal component analysis achieves global alignment. The fine-matching criterion employs an iterative nearest-point algorithm, combined with point cloud density weights and normal vector constraints, to further improve local matching accuracy and reduce point cloud deformation errors. Finally, the error distribution analysis is presented intuitively as a heatmap, making accuracy evaluation more quantitative, visual, and intuitively controllable.

[0018] This invention improves the scientific rigor of scanning accuracy evaluation by introducing a dynamic error compensation and comprehensive scoring system. Through real-time acquisition of probe motion trajectory and speed data, a correlation model between scanning speed and contour overlap error is established. Dynamic compensation coefficients are generated based on an exponential decay function, effectively reducing measurement errors caused by changes in scanning speed. Combined with nonlinear weighted fusion and boundary constraint correction, accurate comprehensive accuracy scoring is achieved, making the scoring system more aligned with clinical needs and ensuring that low-quality scanning results do not affect the reliability of the overall evaluation. This improves the matching accuracy of intraoral 3D scanning data and the stability of the measurement system, providing a more precise and stable standard evaluation system for digital oral diagnosis and treatment and prosthesis manufacturing. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the evaluation method flow according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the system functional modules according to an embodiment of the present invention. Detailed Implementation

[0021] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0022] It should be noted that the use of terms such as "an embodiment," "an embodiment," "an exemplary embodiment," and "some embodiments" in the specification indicates that the described embodiment may include a specific feature, structure, or characteristic, but not every embodiment necessarily includes that specific feature, structure, or characteristic. Furthermore, when a specific feature, structure, or characteristic is described in connection with an embodiment, implementing such a feature, structure, or characteristic in conjunction with other embodiments (whether explicitly described or not) should be within the knowledge of those skilled in the art.

[0023] Generally, terms can be understood at least partly from their use in context. For example, depending at least partly on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in a singular sense, or a combination of features, structures, or characteristics in a plural sense. Additionally, the term "based on" can be understood not necessarily to convey an exclusive set of factors, but rather, alternatively, depending at least partly on the context, to allow for the presence of other factors that are not necessarily explicitly described.

[0024] like Figure 1 As shown, a standard evaluation method for scanning accuracy using an intraoral 3D scanner includes the following steps: S1, Construct a standardized oral dynamic model: On a biomimetic gingival base with biomechanical properties, fix a standardized three-dimensional solid model including tooth preparation body, adjacent teeth and opposing teeth, and set a programmable humidity control layer on the surface of the standardized three-dimensional solid model. S2, Scanning data acquisition: The standardized three-dimensional solid model is scanned from multiple angles using the intraoral scanner under test, and the motion trajectory and speed data of the scanning probe are recorded simultaneously. S3, Registration and Alignment of Scanned Data with Reference Model: An adaptive registration algorithm based on feature point cloud is used to align the obtained scanned data with the reference oral digital model in spatial coordinates and output a registration error distribution map; S4, Multi-dimensional accuracy analysis: Based on the output registration error distribution map, calculate the contour overlap of the tooth preparation edge area, the occlusal surface curvature matching degree, and the point cloud density deviation value of the adjacent area. S5, Dynamic Error Compensation Assessment: Based on the recorded probe movement speed data, establish a correlation model between scanning speed and contour overlap, and generate dynamic error compensation coefficients. S6, Comprehensive evaluation index generation: Weighted fusion of contour overlap, curvature matching degree, density deviation value and dynamic error compensation coefficient to output a comprehensive accuracy score.

[0025] The construction of a standardized oral dynamic model in S1 includes: S11, Biomimetic Gingival Substrate Preparation: A biomimetic gingival substrate with biomechanical properties was prepared using a silicone rubber and polyurethane composite material, and its elastic modulus was calculated. , is represented as: ; in, The elastic modulus (MPa) is controlled within Within a range to simulate real teeth, For the applied vertical load , The base area under stress , For compression deformation (mm). The original thickness of the substrate (mm); S12, 3D solid model fixation: The standardized 3D solid models of the prepared tooth, adjacent teeth, and opposing teeth are arranged according to the clinical dental arch morphology and embedded into the biomimetic gingival base using positioning pins. The positioning error meets the following conditions: ,in, This represents the maximum displacement deviation between the model and the base. S13, Humidity Control Layer Construction: A programmable humidity control layer is set on the surface of the 3D solid model, with a surface contact angle of... Adjustment via microfluidic channels is represented as follows: ; in, For microfluidic channel inlet flow , Microfluidic channel outlet flow rate , The material permeability coefficient ( ), For control time (min). For the contact angle, control it within To simulate a moist oral environment; The above methods effectively improve the reliability and consistency of intraoral 3D scanner accuracy evaluation. The biomimetic gingival base uses a composite material of silicone rubber and polyurethane, making its elastic modulus close to that of real oral tissue, thereby simulating the clinical scanning environment and ensuring the realism of measurement conditions. The embedded fixing method of the positioning pin can minimize the displacement deviation between the model and the base, and improve the stability of experimental repeatability. At the same time, the use of microfluidic channels to achieve programmable humidity control allows for accurate simulation of dynamic humidity changes in the oral environment, optimizing the realism of scanning accuracy evaluation.

[0026] The scan data acquisition in S2 includes: S21, Multi-angle scanning path planning: Set the starting angle of the scanning probe and adjust it sequentially according to fixed angle increments to ensure multiple scans are performed within a reasonable range, resulting in a set of scanning angles. Represented as: ; in, As the starting scanning angle, take , As the angle increment, take , The number of scans, satisfying ; S22, Synchronous recording of motion trajectory: Real-time acquisition of three-dimensional spatial coordinates via an inertial measurement unit (IMU) mounted on the scanning probe. and record timestamps The sampling frequency meets the condition as follows: ; S23, Motion speed calculation: based on continuous timestamps and Calculate instantaneous velocity from coordinate data. , is represented as: ; in, For a moment probe movement speed , for The three-dimensional coordinates at time [time]. for The three-dimensional coordinates at any given time; S24, Abnormal Data Marker: For instantaneous velocity Exceeding the threshold range The scanned segments were marked, among which , ; S25, Data Synchronization Encapsulation: Align scan data, motion trajectory coordinates, and velocity data according to timestamps to generate structured data packets. , is represented as: ; Through the above methods, we ensure uniform coverage of the scanning spot, reduce data omissions, improve measurement integrity, record the motion trajectory of the scanning probe in real time, and combine high-frequency sampling to give the data higher spatiotemporal accuracy, providing a reliable basis for subsequent error analysis. The motion speed calculation and abnormal data marking mechanism can effectively identify and eliminate unstable scanning segments, reduce measurement errors, and improve data consistency. In addition, structured data encapsulation ensures the integrity and traceability of the data, making the evaluation of scanning accuracy more scientific and standardized, which helps to improve the measurement reliability and application value of intraoral 3D scanners.

[0027] The registration and alignment of scan data with the baseline model in S3 includes: S31, Feature point cloud extraction: Extract tooth anatomical feature point clouds from the scan data and the benchmark oral digital model respectively; S32, Coarse Registration and Alignment: Initial registration is performed based on Principal Component Analysis (PCA) of the feature point cloud, and the transformation matrix is ​​calculated, as follows: ; in, For rotation matrix, It is a translation vector. To scan the feature points of the data, For the baseline model feature points, To match the number of feature point pairs, , This is the coarse registration transformation matrix; S33, Fine Registration Optimization: The Iterative Closest Point (ICP) algorithm is used for precise registration. The objective function is: ; in, It is a rigid transformation matrix. To scan data points, For the corresponding point of the baseline model, These are weighting coefficients, dynamically adjusted based on the point cloud density. , , These are the normal vectors of the scan point and the reference point, respectively. For the normal vector constraint weights, take , The number of point pairs that match the scanned data points with the baseline model. To optimize the objective function value for precise registration; S34, Error Distribution Map Generation: Calculate the Euclidean distance error between each registration point and the matching point of the baseline oral digital model after fine registration optimization. and the error Mapped to the HSV color space, a two-dimensional error heatmap and a three-dimensional point cloud error distribution map are generated, with blue representing... Red indicates ,error Represented as: ; Through the above, high-precision matching and error visualization of scanned data were achieved. Feature point cloud extraction utilizes a curvature screening mechanism to ensure that the selected matching points have significant anatomical features, thereby improving matching stability. Coarse registration uses principal component analysis to calculate the initial transformation matrix, enabling rapid global alignment of scanned data. Fine registration improves local matching accuracy and reduces point cloud deformation errors through an improved iterative nearest-point algorithm combined with point cloud density weights and normal vector constraints. Finally, error distribution analysis presents the registration error intuitively in the form of a heatmap, making accuracy evaluation more quantitative and visual.

[0028] Feature point cloud extraction in S31 includes: S311, Tooth Region Segmentation: Using a semantic segmentation network, the scanned data and a baseline digital oral model are used to segment teeth by position, outputting a set of independent points including anterior teeth, premolars, and molars. The segmentation accuracy meets the requirements. ,in, The overlap of the segmented regions is expressed as: ; S312, Local curvature calculation: For points in each partitioned point cloud... Based on its neighborhood radius Calculating curvature values ​​from the eigenvalues ​​of the covariance matrix of interior points , is represented as: ; in, The eigenvalues ​​of the covariance matrix reflect the flatness of the local surfaces in the point cloud. A larger value indicates a more drastic change in curvature; S313, Curvature Threshold Filtering: Set Differentiated Curvature Thresholds Based on Tooth Position , retain satisfaction The feature points constitute the initial feature point cloud. , Represented as: ; S314, Morphological optimization: Initial feature point set Perform 3D morphological closing operations to eliminate discrete noise points and obtain a morphologically optimized feature point cloud. , is represented as: ; in, For spherical structural elements, radius , , These represent expansion and erosion operations, respectively. S315, Curvature Gradient Enhancement: Optimization of Morphologically Optimized Feature Point Clouds Curvature gradient constraints are applied to eliminate gradient abrupt changes and outliers, ultimately outputting a set of anatomical feature points. , is represented as: ; in, The gradient is the curvature space gradient.

[0029] The tooth region segmentation in S311 includes: S3111, 3D point cloud data preprocessing: The scanned data and the baseline oral digital model are normalized and represented as follows: ; ; ; in, This is the original set of point cloud coordinates. For the normalized point cloud, Let the coordinates be the centroid coordinates of the point cloud. Let the standard deviation of the point cloud spatial distribution be . This represents the total number of points in the point cloud data. S3112, Semantic Segmentation Network Inference: The PointNet++ semantic segmentation network is used to classify tooth positions on normalized point cloud data, outputting class probabilities. The objective function is optimized as follows: ; ; in, Labels for tooth position categories (anterior teeth, premolars, molars). For category weight coefficients, For multi-task loss weights, The true label indicates that a point belongs to the category. True segmentation marks on (anterior teeth, premolars, or molars) To predict probabilities, we represent the category of the neural network output. probability fractions, The value of the Dice loss function; S3113, Post-processing of segmentation results: Perform 3D morphological closing operations on the segmentation results output by the semantic segmentation network to obtain morphologically optimized tooth position segmentation labels. , is represented as: ; in, For a closed operational structure, the radius is... , To enable the operation structure, radius , and These represent expansion and erosion operations, respectively. This is the initial tooth segmentation result; S3114, Tooth Position Mapping and Point Cloud Extraction: Based on Morphologically Optimized Tooth Position Segmentation Labels Extracting the independent point clusters of each tooth position , is represented as: ; in, A single point in point cloud data.

[0030] Multidimensional precision analysis in S4 includes: S41, Tooth preparation edge contour extraction: In the registration error distribution map, a 0.5mm wide strip region is extracted along the normal direction of the tooth preparation edge. B-spline curve fitting is performed on the reference oral digital model and the scan data respectively, as shown below: ; ; in, for B-spline basis functions , These are the baseline oral digital model and the coordinates of the control points in the scan data. For curve parameters, B-spline curves are used to prepare the margins of the tooth structure in the reference model. B-spline curves are prepared for the edges of the tooth structure in the scan data. , To control the number of points; S42, Contour overlap calculation: Calculate the mean Hausdorff distance between the B-spline curves of the baseline digital oral model and the tooth preparation margins in the scanned data. , is represented as: ; in, The average Hausdorff distance for preparing the edge profile of the tooth, if If so, it is considered qualified. The number of sampling points is uniform; S43, Calculation of occlusal surface curvature matching degree: Randomly select occlusal surface area Using a sampling point, the relative curvature error between the baseline oral digital model and the scan data is calculated. , is represented as: ; in, The number of sampling points. , Sampling points The maximum principal curvature of the baseline oral digital model and scan data; S44, Calculation of point cloud density deviation in the adjacent region: Within the adjacent region (within ±1 mm of the contact point), calculate the ratio of the point cloud density of the scanned data to that of the reference oral digital model. , is represented as: ; in, The density ratio of point clouds in adjacent regions, if If so, it is considered qualified. , These are the point cloud densities of the reference oral digital model and the scan data for the adjacent area, respectively. S45, Three-dimensional error vector synthesis: The overlap of the contour of the tooth preparation edge area ( ), Occlusal surface curvature matching degree ( ) and the point cloud density deviation value of the adjacent area ( Convert to normalized error vector , is represented as: ; in, To normalize the three-dimensional error vector, This is the transpose operation for a matrix or vector.

[0031] The dynamic error compensation assessment in S5 includes: S51, Speed ​​Data Segmentation Processing: The probe motion speed data is segmented according to the scanning path, with each segment lasting for a specified duration. Calculate the average speed of each segment. and speed fluctuation standard deviation , is represented as: ; ; in, For the first Number of segment velocity data points Instantaneous velocity; S52, Speed-Error Correlation Modeling: Establish an exponential decay model for the relationship between scanning speed and contour overlap error, expressed as: ; in, As the reference error, The velocity attenuation coefficient, This is the sensitivity coefficient for velocity fluctuations; S53, Dynamic Compensation Coefficient Generation: Calculate the compensation coefficient based on the correlation model. , is represented as: ; S54, Velocity Distribution Weight Allocation: The compensation coefficients are weighted and fused according to the velocity range to generate global compensation coefficients. , is represented as: ; ; in, The total number of speed segments; S55, Compensation Effect Verification: Verify the global compensation coefficient. The effect of compensation is verified by applying it to the original contour overlap, and is expressed as follows: ; Verification conditions: ; And the error reduction rate before and after compensation satisfies: ; in, The compensated contour error, The error reduction rate; The above methods effectively reduce measurement errors caused by variations in scanning speed. By processing speed data in segments, the motion characteristics at different stages of the scanning process can be captured. Combined with an exponential decay model and a fluctuation sensitivity factor, fine-tuning of the error is achieved. Furthermore, a weighted fusion method is used to allocate compensation weights, resulting in a more balanced overall compensation effect and ensuring the accuracy of error compensation under different scanning paths. Finally, the compensation effect is verified by the error reduction rate, ensuring that the error reduction reaches more than 40%, thereby improving the matching accuracy of intraoral 3D scanning data and enhancing the stability and reliability of the measurement system.

[0032] The comprehensive evaluation indicators generated in S6 include: S61, Index Normalization Processing: The contour overlap, curvature matching degree, density deviation value, and dynamic error compensation coefficient are normalized and expressed as follows: ; in, , , These are the maximum error thresholds for contour overlap, curvature matching, and density deviation, respectively. , , , These are scores for contour overlap, curvature matching, density deviation, and dynamic error compensation coefficient, respectively. S62, Dynamic allocation of tooth position weights: Differentiated weights are set according to tooth regions. , is represented as: ; S63, Nonlinear Weighted Fusion: An exponential decay function is used for weighted calculation to obtain the comprehensive evaluation score. , is represented as: ; in, To evaluate the overall score, For standardized scores, These are the weighting coefficients; S64, Boundary Constraint Correction: Corrects scores for special scenarios to obtain the final comprehensive evaluation score. , is represented as: ; in, This is the final comprehensive evaluation score; Through the above, a comprehensive quantitative assessment of intraoral scanning accuracy was achieved. Normalization processing converted error data of different dimensions into standardized scores, ensuring the comparability of various indicators. Based on dynamic weight allocation of tooth position regions, the differences in accuracy requirements of different tooth locations were fully considered, making the evaluation system more in line with clinical needs. At the same time, exponential decay weighted calculation was adopted to avoid the impact of excessively low individual scores on the overall score, improving the stability and robustness of the assessment. Boundary constraint correction ensured that key indicators met the minimum clinical standards; otherwise, they were directly judged as unqualified, thereby strengthening the safety requirements of scanning accuracy.

[0033] like Figure 2 As shown, a standard evaluation system for scanning accuracy of an intraoral 3D scanner is used to implement the aforementioned standard evaluation method for scanning accuracy of an intraoral 3D scanner, and includes the following modules: Standardized oral dynamic model module: Based on a biomimetic gingival base, a standardized three-dimensional solid model including tooth preparation, adjacent teeth and opposing teeth is constructed, and a programmable humidity control layer is set; Scanning data acquisition module: Performs multi-angle scanning through the scanner inside the test port, and simultaneously records the motion trajectory and speed data of the scanning probe; Data registration and alignment module: It adopts an adaptive registration algorithm based on feature point cloud to align the scanned data with the reference oral digital model in spatial coordinates and outputs a registration error distribution map; Multi-dimensional accuracy analysis module: Calculates the contour overlap, occlusal surface curvature matching degree, and point cloud density deviation value of the tooth preparation edge area based on the registration error distribution map; Dynamic error compensation evaluation module: Based on probe motion speed data, establish a correlation model between scanning speed and contour overlap, and generate dynamic error compensation coefficients; The comprehensive evaluation index generation module performs weighted fusion of contour overlap, curvature matching, density deviation, and dynamic error compensation coefficient, and outputs a comprehensive accuracy score.

[0034] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0035] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A standard evaluation method for scanning accuracy using an intraoral 3D scanner, characterized in that, Includes the following steps: S1, Construct a standardized oral dynamic model: On a biomimetic gingival base with biomechanical properties, fix a standardized three-dimensional solid model including tooth preparation body, adjacent teeth and opposing teeth, and set a programmable humidity control layer on the surface of the standardized three-dimensional solid model. S2, Scanning data acquisition: The standardized three-dimensional solid model is scanned from multiple angles using the intraoral scanner under test, and the motion trajectory and speed data of the scanning probe are recorded simultaneously. S3, Registration and Alignment of Scanned Data with Reference Model: An adaptive registration algorithm based on feature point cloud is used to align the obtained scanned data with the reference oral digital model in spatial coordinates and output a registration error distribution map; S4, Multi-dimensional accuracy analysis: Based on the output registration error distribution map, calculate the contour overlap of the tooth preparation edge area, the occlusal surface curvature matching degree, and the point cloud density deviation value of the adjacent area. S5, Dynamic Error Compensation Assessment: Based on the recorded probe movement speed data, establish a correlation model between scanning speed and contour overlap, and generate dynamic error compensation coefficients. S6, Comprehensive evaluation index generation: Weighted fusion of contour overlap, curvature matching degree, density deviation value and dynamic error compensation coefficient to output a comprehensive accuracy score.

2. The standard evaluation method for scanning accuracy of an intraoral 3D scanner according to claim 1, characterized in that, The construction of the standardized oral dynamic model in S1 includes: S11, Biomimetic Gingival Substrate Preparation: A biomimetic gingival substrate with biomechanical properties was prepared using a silicone rubber and polyurethane composite material, and its elastic modulus was calculated. ; S12, 3D solid model fixation: The standardized 3D solid models of the prepared tooth, adjacent teeth, and opposing teeth are arranged according to the clinical dental arch morphology and embedded into the biomimetic gingival base using positioning pins. The positioning error meets the following conditions: ,in, This represents the maximum displacement deviation between the model and the base. S13, Humidity Control Layer Construction: A programmable humidity control layer is set on the surface of the 3D solid model, with a surface contact angle of... Adjusted via microfluidic channels.

3. The standard evaluation method for scanning accuracy of an intraoral 3D scanner according to claim 1, characterized in that, The scanning data acquisition in S2 includes: S21, Multi-angle scanning path planning: Set the starting angle of the scanning probe and adjust it sequentially according to fixed angle increments; S22, Synchronous recording of motion trajectory: Real-time acquisition of three-dimensional spatial coordinates via an inertial measurement unit mounted on the scanning probe. and record timestamps The sampling frequency meets the condition as follows: ; S23, Motion speed calculation: based on continuous timestamps and Calculate instantaneous velocity from coordinate data. ; S24, Abnormal Data Marker: For instantaneous velocity Exceeding the threshold range The scanned segments were marked, among which , ; S25, Data Synchronization Encapsulation: Align scan data, motion trajectory coordinates, and velocity data according to timestamps to generate structured data packets. .

4. The standard evaluation method for scanning accuracy of an intraoral 3D scanner according to claim 1, characterized in that, The registration and alignment of the scan data with the reference model in S3 includes: S31, Feature point cloud extraction: Extract tooth anatomical feature point clouds from the scan data and the benchmark oral digital model respectively; S32, Coarse Registration Alignment: Initial registration is performed based on principal component analysis of the feature point cloud, and the transformation matrix is ​​calculated; S33, Fine Registration Optimization: Iterative nearest point algorithm is used for precise registration; S34, Error Distribution Map Generation: Calculate the Euclidean distance error between each registration point and the matching point of the baseline oral digital model after fine registration optimization. and the error Mapped to the HSV color space, a two-dimensional error heatmap and a three-dimensional point cloud error distribution map are generated, with blue representing... Red indicates .

5. The standard evaluation method for scanning accuracy of an intraoral 3D scanner according to claim 4, characterized in that, The feature point cloud extraction in S31 includes: S311, Tooth Region Segmentation: Using a semantic segmentation network, the scanned data and a baseline digital oral model are used to segment teeth by position, outputting a set of independent points for anterior teeth, premolars, and molars. The segmentation accuracy meets the requirements. ,in, To determine the overlap of the segmented regions; S312, Local curvature calculation: For points in each partitioned point cloud... Based on its neighborhood radius Calculating curvature values ​​from the eigenvalues ​​of the covariance matrix of interior points ; S313, Curvature Threshold Filtering: Set Differentiated Curvature Thresholds Based on Tooth Position , retain satisfaction The feature points constitute the initial feature point cloud. ; S314, Morphological optimization: Initial feature point set Perform 3D morphological closing operations to eliminate discrete noise points and obtain a morphologically optimized feature point cloud. ; S315, Curvature Gradient Enhancement: Optimization of Morphologically Optimized Feature Point Clouds Curvature gradient constraints are applied to eliminate gradient abrupt changes and outliers, ultimately outputting a set of anatomical feature points. .

6. The standard evaluation method for scanning accuracy of an intraoral 3D scanner according to claim 5, characterized in that, The tooth region segmentation in S311 includes: S3111, 3D point cloud data preprocessing: normalizing the scanned data and the benchmark oral digital model; S3112, Semantic Segmentation Network Inference: The PointNet++ semantic segmentation network is used to classify tooth positions on the normalized point cloud data and output the category probability. S3113, Post-processing of segmentation results: Perform 3D morphological closing operations on the segmentation results output by the semantic segmentation network to obtain morphologically optimized tooth position segmentation labels. ; S3114, Tooth Position Mapping and Point Cloud Extraction: Based on Morphologically Optimized Tooth Position Segmentation Labels Extracting the independent point clusters of each tooth position .

7. The standard evaluation method for scanning accuracy of an intraoral 3D scanner according to claim 1, characterized in that, The multi-dimensional precision analysis in S4 includes: S41, Tooth preparation edge contour extraction: In the registration error distribution map, a 0.5mm wide strip region is extracted along the normal direction of the tooth preparation edge, and B-spline curve fitting is performed on the reference oral digital model and the scanning data respectively. S42, Contour overlap calculation: Calculate the mean Hausdorff distance between the B-spline curves of the baseline digital oral model and the tooth preparation margins in the scanned data. ; S43, Calculation of occlusal surface curvature matching degree: Randomly select occlusal surface area Using a sampling point, the relative curvature error between the baseline oral digital model and the scan data is calculated. ; S44, Calculation of point cloud density deviation in adjacent regions: Within the adjacent region, calculate the ratio of the point cloud density of the scanned data to that of the baseline oral digital model. ; S45, 3D Error Vector Synthesis: The contour overlap of the tooth preparation margin area, the occlusal surface curvature matching degree, and the point cloud density deviation of the adjacent area are converted into a normalized error vector. .

8. The standard evaluation method for scanning accuracy of an intraoral 3D scanner according to claim 7, characterized in that, The dynamic error compensation evaluation in S5 includes: S51, Speed ​​Data Segmentation Processing: The probe motion speed data is segmented according to the scanning path, with each segment lasting for a specific duration. Calculate the average speed of each segment. and speed fluctuation standard deviation ; S52, Speed-Error Correlation Modeling: Establishing an exponential decay model for the relationship between scanning speed and contour overlap error; S53, Dynamic Compensation Coefficient Generation: Calculate the compensation coefficient based on the correlation model. ; S54, Velocity Distribution Weight Allocation: The compensation coefficients are weighted and fused according to the velocity range to generate global compensation coefficients. ; S55, Compensation Effect Verification: Verify the global compensation coefficient. The effect of compensation was verified by applying it to the original contour overlap.

9. The standard evaluation method for scanning accuracy of an intraoral 3D scanner according to claim 8, characterized in that, The generation of the comprehensive evaluation index in S6 includes: S61, Index Normalization Processing: The contour overlap, curvature matching degree, density deviation value and dynamic error compensation coefficient will be normalized. S62, Dynamic allocation of tooth position weights: Differentiated weights are set according to tooth regions. ; S63, Nonlinear Weighted Fusion: An exponential decay function is used for weighted calculation to obtain the comprehensive evaluation score. ; S64, Boundary Constraint Correction: Corrects scores for special scenarios to obtain the final comprehensive evaluation score. .

10. A standard evaluation system for scanning accuracy of an intraoral 3D scanner, used to implement the standard evaluation method for scanning accuracy of an intraoral 3D scanner as described in any one of claims 1-9, characterized in that, Includes the following modules: Standardized oral dynamic model module: Based on a biomimetic gingival base, a standardized three-dimensional solid model including tooth preparation, adjacent teeth and opposing teeth is constructed, and a programmable humidity control layer is set; Scanning data acquisition module: Performs multi-angle scanning through the scanner inside the test port, and simultaneously records the motion trajectory and speed data of the scanning probe; Data registration and alignment module: It adopts an adaptive registration algorithm based on feature point cloud to align the scanned data with the reference oral digital model in spatial coordinates and outputs a registration error distribution map; Multi-dimensional accuracy analysis module: Calculates the contour overlap, occlusal surface curvature matching degree, and point cloud density deviation value of the tooth preparation edge area based on the registration error distribution map; Dynamic error compensation evaluation module: Based on probe motion speed data, establish a correlation model between scanning speed and contour overlap, and generate dynamic error compensation coefficients; The comprehensive evaluation index generation module performs weighted fusion of contour overlap, curvature matching, density deviation, and dynamic error compensation coefficient, and outputs a comprehensive accuracy score.

Images