Orthodontic tooth division accurate measurement method and system based on digital scanning

Abstract

The invention relates to an orthodontic tooth division accurate measurement method and system based on digital scanning, and the method comprises the steps: collecting the initial three-dimensional scanning data of teeth of a patient, and obtaining the dynamic change data of the teeth; performing association mapping processing on the tooth dynamic change data to obtain a tooth space position file; based on the tooth space position file, adopting an iterative comparison algorithm to be matched with subsequent scanning data, and judging deviation values of tooth gap width and tooth root inclination in an orthodontic process; the deviation value is analyzed, and real-time dental arch curve adjustment parameters are obtained; based on the real-time dental arch curve adjustment parameters, determining an optimization direction of an individualized tooth division scheme; and performing feedback circulation on the overall tooth space position file by adopting the optimization direction to obtain a final dynamic integration measurement result. According to the invention, through fusion of deep learning and time sequence analysis, accurate tracking of dynamic changes of teeth is realized.

Description

Technical Field

[0001] This invention relates to the field of digital scanning technology, and in particular to a method and system for accurate measurement of orthodontic tooth segmentation based on digital scanning. Background Technology

[0002] Orthodontic treatment, as an important branch of dentistry, aims to improve occlusal function and aesthetics by adjusting tooth position. Its precision directly impacts treatment outcomes and patient experience. With the widespread adoption of digital technology, orthodontic tooth separation measurement based on 3D scanning has become a research hotspot because it provides more accurate spatial data of teeth and reduces errors associated with traditional manual measurements. However, current methods are limited in that measurement systems often struggle to comprehensively integrate multidimensional tooth parameters and effectively track dynamic changes during treatment. This results in a lack of continuous and individualized data support for dentists when developing tooth separation plans. This not only increases the complexity of treatment design but may also affect the stability of long-term efficacy. Existing orthodontic measurement methods typically rely on tooth scan data at a single point in time, lacking continuous recording of dynamic changes in teeth. For example, traditional methods may only record the initial interdental width, ignoring the changing trends of interdental spacing, angles, and arch curves after tooth movement during treatment. This limitation of static data makes it difficult for dentists to accurately assess the progress and effectiveness of tooth separation treatment. Furthermore, existing systems often fail to effectively integrate key information such as tooth spatial location, interproximal contact points, and root inclination when processing multidimensional parameters, resulting in fragmented data and difficulty in forming a systematic patient dental profile. This lack of data integration further exacerbates the difficulties doctors face in diagnosis and treatment planning. The core technical challenge lies in how to achieve dynamic integration and continuous tracking of multidimensional tooth parameters. First, parameters such as tooth spatial location and interproximal contact points are complex in three-dimensional space, making it difficult for existing systems to uniformly store and correlate these parameters. For example, in a patient undergoing tooth separation treatment, a slight tilt of the tooth may alter the force distribution at the interproximal contact points, but current technology struggles to capture this change in real time and correlate it with other parameters such as the arch curve. Second, the lack of dynamic tracking prevents the precise quantification of changes in interdental spacing before and after treatment. For instance, doctors need to know whether the interdental width of a tooth has reached the expected level three months after treatment, but existing systems often only provide static initial data and cannot offer continuous comparative analysis. Therefore, how to construct a digital measurement method and system that can integrate multidimensional tooth parameters and support dynamic comparison at multiple time points has become a key issue in the accurate measurement of orthodontic tooth segmentation. Summary of the Invention

[0003] The purpose of this invention is to propose a method and system for accurate measurement of orthodontic tooth segmentation based on digital scanning, so as to solve the problems existing in the prior art.

[0004] To achieve the above objectives, the present invention provides the following solution:

[0005] A precise measurement method for orthodontic tooth segmentation based on digital scanning, comprising:

[0006] Collect initial three-dimensional scan data of the patient's teeth to obtain data on dynamic changes in the teeth;

[0007] The dynamic change data of the teeth are correlated and mapped to obtain a spatial position profile of the teeth;

[0008] Based on the aforementioned tooth spatial position file, an iterative comparison algorithm is used to match the data with subsequent scans to determine the deviation values ​​of interdental width and root inclination during orthodontic treatment.

[0009] Analyze the deviation values ​​to obtain real-time dental arch curve adjustment parameters;

[0010] Based on the real-time dental arch curve adjustment parameters, the optimization direction of the individualized tooth separation scheme is determined;

[0011] The optimized direction is used to perform a feedback loop on the overall tooth spatial position file to obtain the final dynamic integrated measurement result.

[0012] Optionally, initial three-dimensional scan data of the patient's teeth are collected to obtain dynamic change data of the teeth, including:

[0013] Initial three-dimensional scan data of the patient's teeth were collected, and features were extracted from the data using a convolutional neural network to obtain multidimensional parameters; wherein, the multidimensional parameters include: interdental width and root inclination;

[0014] Based on the multidimensional parameters, a multi-time-point data sequence is established, and the multi-time-point data sequence is processed by a time series analysis model to obtain data on dynamic changes in teeth.

[0015] Optionally, obtaining the multidimensional parameters includes:

[0016] Three-dimensional scanning data of the patient's teeth are collected using a three-dimensional scanning device to generate initial point cloud data;

[0017] Based on the initial point cloud data, a point cloud registration algorithm is used to automatically align the multi-angle scanning results, ensure data integrity, and calculate surface curvature to form a high-resolution three-dimensional mesh model.

[0018] The three-dimensional mesh model is converted into an image tensor and input into a convolutional neural network to obtain feature vectors.

[0019] The gradient of the feature vector is calculated to quantify changes in tooth morphology, and a multidimensional parameter vector is finally output. The interdental width is the average value obtained by measuring the minimum distance between adjacent teeth using the Euclidean distance algorithm, and the root inclination is the quantified value of the angle between the root axis and the crown axis calculated by principal component analysis.

[0020] Optionally, the three-dimensional mesh model is converted into an image tensor and input into the convolutional neural network, including:

[0021] First, preliminary feature extraction is performed using 7x7 convolutional kernels to extract edge and texture information. Then, the feature map size is gradually reduced through multiple residual block processing, while max pooling layers are applied to reduce noise.

[0022] Optionally, the dynamic tooth change data is correlated and mapped to obtain a tooth spatial location profile, including:

[0023] The multidimensional parameters in the dynamic tooth change data are correlated and mapped to obtain preliminary integration results.

[0024] Based on the preliminary integration results, a clustering algorithm was used to group the tooth spatial data and determine the set of location features;

[0025] Spatial dynamic indicators are obtained from the set of location features to determine whether they meet the archival standards;

[0026] If the spatial dynamic indicators meet the archival standards, then the trend changes are predicted using a linear regression model to obtain the optimized parameter set;

[0027] For the optimized parameter set, the associated mapping integrates the file elements to determine the final tooth spatial position file.

[0028] Optionally, based on the tooth spatial position file, an iterative comparison algorithm is used to match the data with subsequent scans to determine the deviation values ​​of interdental width and root inclination during orthodontic treatment, including:

[0029] Initial tooth spatial coordinates are extracted from the tooth spatial position file to obtain the tooth position reference;

[0030] The set of spatial matching points for teeth is determined by processing subsequent scan data and tooth position benchmarks through an iterative comparison algorithm.

[0031] The variation in interdental width is calculated based on the set of matching points in the tooth space, and the deviation value of the interdental width is determined.

[0032] If the width deviation exceeds the preset threshold, the root axis is extracted from the set of matching points in the tooth space to obtain the root tilt angle.

[0033] Principal component analysis was used to analyze the root inclination angle and the initial tooth spatial coordinates to determine the deviation value of the root inclination.

[0034] Optionally, analyzing the deviation value to obtain real-time dental arch curve adjustment parameters includes:

[0035] Based on the deviation value, continuous tracking records are obtained from the data sequence to obtain the deviation distribution;

[0036] Based on the deviation distribution, a linear regression model was used to fit the continuous tracking records, and the fitting coefficients were determined.

[0037] If the fitting coefficient exceeds the preset threshold, outliers are extracted from the continuous tracking records to obtain an anomaly sequence.

[0038] By filtering out abnormal sequences, the continuously tracked records in the data sequence are updated to obtain the updated sequence;

[0039] Based on the updated sequence, calculate the dental arch curve offset and determine the offset vector;

[0040] k-means clustering is used to group the offset vectors, determine the cluster centers, and obtain the center coordinates.

[0041] By interpolating the center coordinates, the dental arch curve is generated and the parameters are adjusted in real time.

[0042] Optionally, based on the real-time dental arch curve adjustment parameters, determining the optimization direction of the individualized tooth separation scheme includes:

[0043] Real-time dental arch curve data is acquired, and key point coordinates are extracted through image processing to obtain a set of curve parameters.

[0044] Based on the set of curve parameters, a linear regression model is used to fit the trend line to determine the parameter adjustment range;

[0045] If the parameter adjustment exceeds the preset threshold, cluster analysis is used to group similar curve segments to obtain feature groups at multiple time points.

[0046] For feature groups at multiple time points, a comparison matrix framework is constructed to determine the differences between matrix elements and obtain time comparison results.

[0047] By comparing the results over time and integrating individual scheme data, a decision tree model is used to classify and optimize the path, and the variation points of the tooth-splitting scheme are determined.

[0048] Obtain the variation points of the tooth splitting scheme, calculate the scheme optimization vector, determine the consistency of the vector direction, and obtain the optimization direction index.

[0049] A precision measurement system for orthodontic tooth segmentation based on digital scanning, comprising:

[0050] The data acquisition module is used to collect initial three-dimensional scan data of the patient's teeth and obtain data on dynamic changes in the teeth;

[0051] The mapping processing module is used to perform correlation mapping processing on the dynamic change data of the teeth to obtain a spatial location profile of the teeth.

[0052] The matching module is used to match the tooth spatial position file with subsequent scan data using an iterative comparison algorithm to determine the deviation values ​​of interdental width and root inclination during orthodontic treatment.

[0053] The analysis module is used to analyze the deviation value and obtain real-time dental arch curve adjustment parameters;

[0054] The determination module is used to determine the optimization direction of the individualized tooth separation scheme based on the adjustment parameters of the real-time dental arch curve.

[0055] The feedback module is used to perform feedback loops on the overall tooth spatial position file using the optimized direction to obtain the final dynamic integrated measurement result.

[0056] The beneficial effects of this invention are as follows:

[0057] This invention acquires initial 3D scan data of the patient's teeth using a 3D scanner and extracts features from the data using a convolutional neural network (CNN), enabling precise quantification of multidimensional parameters such as interdental width and root inclination angle. Compared to traditional manual measurement methods, the combination of 3D scanning and deep learning technology significantly improves measurement accuracy, controlling errors within the millimeter or even sub-millimeter range, thus providing more reliable basic data for orthodontic treatment.

[0058] This invention introduces a time-series analysis model to process dental data at multiple time points, enabling dynamic tracking of spatial positional changes in teeth during treatment. By comparing initial and subsequent scan data in real time, deviations in interdental width and root tilt can be accurately determined, and real-time adjustment parameters can be generated. This dynamic tracking capability overcomes the limitations of traditional methods that rely solely on data from a single time point, allowing for a more comprehensive understanding of the dynamic changes in teeth.

[0059] This invention integrates multidimensional parameters of teeth (such as interdental width, tilt angle, and occlusal force) into a tooth spatial location profile through association mapping and clustering algorithms, forming a systematic patient dental data profile. This integration method avoids the problems of data dispersion and difficulty in correlation in traditional methods.

[0060] This invention utilizes deviation value analysis and iterative comparison algorithms to accurately calculate the deviation values ​​of interdental width and root tilt during treatment, and updates the continuous tracking records in the data sequence accordingly to generate real-time dental arch curve adjustment parameters. This deviation analysis and adjustment mechanism ensures that every adjustment during treatment is based on accurate measurement data. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0062] Figure 1 This is a schematic diagram of a method for precise measurement of orthodontic tooth segmentation based on digital scanning, according to an embodiment of the present invention. Detailed Implementation

[0063] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0064] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0065] like Figure 1 As shown, this embodiment proposes a precise measurement method for orthodontic tooth segmentation based on digital scanning, including:

[0066] Collect initial three-dimensional scan data of the patient's teeth to obtain data on dynamic changes in the teeth;

[0067] The dynamic change data of the teeth are correlated and mapped to obtain a spatial position profile of the teeth;

[0068] Based on the aforementioned tooth spatial position file, an iterative comparison algorithm is used to match the data with subsequent scans to determine the deviation values ​​of interdental width and root inclination during orthodontic treatment.

[0069] Analyze the deviation values ​​to obtain real-time dental arch curve adjustment parameters;

[0070] Based on the real-time dental arch curve adjustment parameters, the optimization direction of the individualized tooth separation scheme is determined;

[0071] The optimized direction is used to perform a feedback loop on the overall tooth spatial position file to obtain the final dynamic integrated measurement result.

[0072] Furthermore, initial three-dimensional scan data of the patient's teeth were collected to obtain dynamic change data of the teeth, including:

[0073] Initial three-dimensional scan data of the patient's teeth were collected, and features were extracted from the data using a convolutional neural network to obtain multidimensional parameters; wherein, the multidimensional parameters include: interdental width and root inclination;

[0074] Based on the multidimensional parameters, a multi-time-point data sequence is established, and the multi-time-point data sequence is processed by a time series analysis model to obtain data on dynamic changes in teeth.

[0075] Furthermore, obtaining the multidimensional parameters includes:

[0076] Three-dimensional scanning data of the patient's teeth are collected using a three-dimensional scanning device to generate initial point cloud data;

[0077] Based on the initial point cloud data, a point cloud registration algorithm is used to automatically align the multi-angle scanning results, ensure data integrity, and calculate surface curvature to form a high-resolution three-dimensional mesh model.

[0078] The three-dimensional mesh model is converted into an image tensor and input into a convolutional neural network to obtain feature vectors.

[0079] The gradient of the feature vector is calculated to quantify changes in tooth morphology, and a multidimensional parameter vector is finally output. The interdental width is the average value obtained by measuring the minimum distance between adjacent teeth using the Euclidean distance algorithm, and the root inclination is the quantified value of the angle between the root axis and the crown axis calculated by principal component analysis.

[0080] Furthermore, the three-dimensional mesh model is converted into an image tensor, which is then input into the convolutional neural network, including:

[0081] First, preliminary feature extraction is performed using 7x7 convolutional kernels to extract edge and texture information. Then, the feature map size is gradually reduced through multiple residual block processing, while max pooling layers are applied to reduce noise.

[0082] Specifically, in this embodiment, a dental-specific 3D scanner, such as the 3Shape TRIOS device, is first used to scan the patient's tooth surface with a precision of 0.01mm to generate initial point cloud data containing 5 million points. The multi-angle scan results are automatically aligned using a point cloud registration algorithm, such as the iterative nearest point method, to ensure data integrity and calculate the average surface curvature as 2.5, forming a high-resolution 3D mesh model as subsequent input. The 3D mesh model was then converted into a 256x256x3 image tensor and input into a ResNet-50-based convolutional neural network. The network first performed preliminary feature extraction using 7x7 convolutional kernels to extract edge and texture information. Then, after processing through multiple residual blocks, the feature map size was gradually reduced to 8x8. Simultaneously, max pooling layers were applied to reduce noise. During network training, the Adam optimizer was used to iterate for 100 epochs at a learning rate of 0.001. During the analysis process, the gradient of the feature vector was calculated to quantify changes in tooth morphology. Finally, a multi-dimensional parameter vector was output. The interdental width was measured using the Euclidean distance algorithm to obtain an average value of 1.8 mm between adjacent teeth. The root tilt was calculated using principal component analysis to quantify the angle between the root axis and the crown axis as 12.5 degrees.

[0083] Quantitative data is collected from multiple time points, and an autoregressive integral moving average model is used to process the observations, resulting in a preliminary trend vector. A stationarity index is extracted from this preliminary trend vector. If the stationarity index is below a preset threshold, the vector is differencing to determine a stationary sequence. The autocorrelation function is calculated from the stationary sequence, and the parameters are fitted using the least squares method to obtain the prediction model coefficients. Future time points are simulated based on the prediction model coefficients, and it is determined whether the simulated output exceeds the abnormal range, identifying potential risk areas. Local fluctuation amplitudes are obtained from these potential risk areas. If the fluctuation amplitude exceeds a preset threshold, a smoothing filter is applied to process the amplitude, determining an adjustment vector. Based on the adjustment vector and historical quantitative values, a vector-weighted average is used to obtain an optimized dynamic trend curve for tooth changes.

[0084] Furthermore, the dynamic tooth change data is correlated and mapped to obtain a tooth spatial location profile, including:

[0085] The multidimensional parameters in the dynamic tooth change data are correlated and mapped to obtain preliminary integration results.

[0086] Based on the preliminary integration results, a clustering algorithm was used to group the tooth spatial data and determine the set of location features;

[0087] Spatial dynamic indicators are obtained from the set of location features to determine whether they meet the archival standards;

[0088] If the spatial dynamic indicators meet the archival standards, then the trend changes are predicted using a linear regression model to obtain the optimized parameter set;

[0089] For the optimized parameter set, the associated mapping integrates the file elements to determine the final tooth spatial position file.

[0090] Specifically, in this implementation, multidimensional parameters such as tooth angle of 30 degrees, interdental space of 2.0 mm, and occlusal pressure of 50 Newtons are correlated and mapped. These parameters are input into a neural network model, and the correlation matrix is ​​calculated through a fully connected layer. For example, matrix A = [[1,0.5,0.2], [0.3,1,0.4], [0.1,0.6,1]]. After mapping, the integrated vector v = A*[30,2.0,50]^T is output, resulting in [41.0,82.0,124.0]. This vector represents the spatial coordinate adjustment value. Furthermore, the original scan data is combined with Fourier transform analysis of frequency components to optimize the stability of the file. For example, low-frequency filtering is used to remove noise, and the final tooth spatial position file is generated. This file contains 10,000 nodes in a three-dimensional mesh model, with each node's coordinates accurate to 0.01 mm, and is stored in a database in association with the patient's historical records.

[0091] Furthermore, based on the aforementioned tooth spatial position profile, an iterative comparison algorithm is used to match the data with subsequent scans to determine the deviations in interdental width and root tilt during orthodontic treatment, including:

[0092] Initial tooth spatial coordinates are extracted from the tooth spatial position file to obtain the tooth position reference;

[0093] The set of spatial matching points for teeth is determined by processing subsequent scan data and tooth position benchmarks through an iterative comparison algorithm.

[0094] The variation in interdental width is calculated based on the set of matching points in the tooth space, and the deviation value of the interdental width is determined.

[0095] If the width deviation exceeds the preset threshold, the root axis is extracted from the set of matching points in the tooth space to obtain the root tilt angle.

[0096] Principal component analysis was used to analyze the root inclination angle and the initial tooth spatial coordinates to determine the deviation value of the root inclination.

[0097] Specifically, in this embodiment, for example, the position of each tooth in the maxillary dentition is represented by (x, y, z) coordinates, such as the center point of the crown of the first molar being (10.2, 15.4, 20.1), and historical treatment records are integrated to form a unified database file to ensure the consistency of the benchmark for subsequent comparisons. This file is automatically stored and indexed through a cloud computing platform, forming a logical chain from data collection to file construction, which facilitates the association with the patient's personalized treatment plan. Next, an iterative comparison algorithm, such as the ICP algorithm, is used for matching. Subsequent scan data (such as the 3D model from week 4 of treatment, with the same resolution of 0.05mm) are aligned with the initial file. The algorithm process includes initial coarse registration calculation of the transformation matrix T, and iterative optimization by minimizing the point-to-point distance error. For example, after 10 iterations, the error is reduced from the initial 1.2mm to 0.1mm. The Euclidean distance formula d=√[(x2-x1)²+(y2-y1)²+(z2-z1)²] is introduced in the analysis to evaluate the matching accuracy. If the deviation exceeds the threshold of 0.2mm, the matrix is ​​automatically adjusted and iterated again. This matching logic is extended to business associations, such as automatically generating visual overlay maps, and supporting deviation trend prediction. Finally, the deviation values ​​of interdental width and root tilt during treatment are determined. Key parameters are extracted through the matched model. For example, the interdental width is calculated by using a boundary detection algorithm to identify the edges of adjacent teeth. The initial width is 2.0 mm, and the subsequent width is 2.3 mm, with a deviation value of 0.3 mm. The root tilt is calculated by using vector analysis to calculate the root axis angle. The initial tilt is 5.0 degrees, and the subsequent tilt is 6.5 degrees, with a deviation of 1.5 degrees.

[0098] Furthermore, analyzing the deviation value to obtain real-time dental arch curve adjustment parameters includes:

[0099] Based on the deviation value, continuous tracking records are obtained from the data sequence to obtain the deviation distribution;

[0100] Based on the deviation distribution, a linear regression model was used to fit the continuous tracking records, and the fitting coefficients were determined.

[0101] If the fitting coefficient exceeds the preset threshold, outliers are extracted from the continuous tracking records to obtain an anomaly sequence.

[0102] By filtering out abnormal sequences, the continuously tracked records in the data sequence are updated to obtain the updated sequence;

[0103] Based on the updated sequence, calculate the dental arch curve offset and determine the offset vector;

[0104] k-means clustering is used to group the offset vectors, determine the cluster centers, and obtain the center coordinates.

[0105] By interpolating the center coordinates, the dental arch curve is generated and the parameters are adjusted in real time.

[0106] Specifically, in this embodiment, the dental arch scan data sequence is first processed by deviation value analysis. For example, a set of dental arch coordinate points are input, such as (x1=2.5, y1=3.0; x2=2.7, y2=3.1; x3=2.4, y3=2.9). The deviation value of each point from the standard dental arch model is calculated. The algorithm is used: deviation = actual value - standard value, to obtain a deviation sequence such as (0.2, 0.1, -0.1). Then, the average deviation is analyzed to be (0.2+0.1-0.1) / 3=0.07. If the deviation exceeds the threshold of 0.1, anomalies are marked to ensure data accuracy. This process forms a logical chain with subsequent tracking record updates because anomalies will affect continuous tracking, thereby automatically triggering the data cleaning module in the system to optimize the sequence quality. Next, the continuous tracking records in the data sequence are updated. For example, based on the deviation analysis results, the Kalman filter algorithm is applied to the original sequence for smoothing updates. Assuming the state transition matrix A=1, the observation matrix H=1, the process noise Q=0.01, and the observation noise R=0.05, the predicted state x'=A*x, the covariance P'=A*P*A'+Q, the Kalman gain K=P'*H' / (H*P'*H'+R), and the updated state x=x'+K*(zH*x'). For example, if the initial x=2.5 and z=2.6, the updated x=2.55 is obtained. This update is associated with the deviation analysis to maintain tracking continuity and is linked to the curve adjustment parameter calculation. Because the updated records provide real-time input for parameter optimization, adjustment suggestions are automatically generated in orthodontic services to avoid manual intervention. Finally, the real-time dental arch curve adjustment parameters are obtained. For example, the updated sequence is used to fit a quadratic curve y=ax^2+bx+c, and the least squares method is used to solve for a=0.1, b=0.5, c=2.0. The adjustment parameters such as curvature deviation Δk=0.05 and angle correction θ=1.2 degrees are calculated.

[0107] Furthermore, based on the real-time dental arch curve adjustment parameters, the optimization direction of the individualized tooth separation scheme is determined, including:

[0108] Real-time dental arch curve data is acquired, and key point coordinates are extracted through image processing to obtain a set of curve parameters.

[0109] Based on the set of curve parameters, a linear regression model is used to fit the trend line to determine the parameter adjustment range;

[0110] If the parameter adjustment exceeds the preset threshold, cluster analysis is used to group similar curve segments to obtain feature groups at multiple time points.

[0111] For feature groups at multiple time points, a comparison matrix framework is constructed to determine the differences between matrix elements and obtain time comparison results.

[0112] By comparing the results over time and integrating individual scheme data, a decision tree model is used to classify and optimize the path, and the variation points of the tooth-splitting scheme are determined.

[0113] Obtain the variation points of the tooth splitting scheme, calculate the scheme optimization vector, determine the consistency of the vector direction, and obtain the optimization direction index.

[0114] Specifically, in this embodiment, a linear regression model is used to fit the Bézier curve for parameter adjustment, optimizing the curvature to 15.5mm to match the accuracy requirement of less than 0.2mm deviation in the scanned point cloud data. This adjustment process involves calculating the sum of squared residuals and minimizing the formula ∑(yi - f(xi))^2, where yi is the actual point coordinate and f(xi) is the fitting function, thereby ensuring that the parameters are highly consistent with the real-time curve. Then, a multi-time-point comparison matrix is ​​generated based on the adjusted parameters. Data with curvature of 15.0mm at time T1, 15.3mm at time T2, and 15.6mm at time T3 are collected to construct a 3x3 matrix [[15.0,0.3,0.6], [-0.3,15.3,0.3], [-0.6, -0.3,15.6]], where the diagonal is the absolute value and the off-diagonal is the difference value. The Euclidean distance algorithm is used to calculate the distance between each time point, such as dist(T1,T2)=√(0.3^2)=0.3mm to analyze the trend of change. If the trend shows a gradual increase, it is associated with the tooth movement business model predicting that the future curvature will reach 16.0mm. Finally, the optimization direction of the individualized tooth separation scheme is determined. The optimization process is simulated by a genetic algorithm with a population size of 50 and 100 iterations. The fitness function is calculated to minimize the tooth spacing deviation, such as the difference between the target spacing of 1.2mm and the actual spacing of 1.0mm. The optimization direction is to adjust the spacing of the maxillary anterior teeth forward by 0.5mm and reduce the lateral force by 20% by combining the matrix trend.

[0115] Specifically, in this embodiment, the optimization direction is used to perform a feedback loop on the overall data archive to obtain the final dynamic integrated measurement result, including:

[0116] The process begins by acquiring data archives and determining the optimization direction based on preset thresholds, resulting in initial integrated measurement results. For these initial results, a support vector machine (SVM) algorithm is used to classify dynamically integrated attributes and determine the feedback loop path. Overall data features are extracted from the feedback loop path; if these features exceed a preset threshold, the optimization parameters are adjusted to obtain updated dynamic results. This updated dynamic result is used to construct a loop feedback model to capture the changing trends of the archive data. Based on these trends, a random forest algorithm is applied to predict potential biases in the integrated measurement; if the potential bias is below a preset threshold, the fusion results are dynamically adjusted to determine the final result.

[0117] This embodiment also proposes a precise orthodontic tooth segmentation measurement system based on digital scanning, including:

[0118] The data acquisition module is used to collect initial three-dimensional scan data of the patient's teeth and obtain data on dynamic changes in the teeth;

[0119] The mapping processing module is used to perform correlation mapping processing on the dynamic change data of the teeth to obtain a spatial location profile of the teeth.

[0120] The matching module is used to match the tooth spatial position file with subsequent scan data using an iterative comparison algorithm to determine the deviation values ​​of interdental width and root inclination during orthodontic treatment.

[0121] The analysis module is used to analyze the deviation value and obtain real-time dental arch curve adjustment parameters;

[0122] The determination module is used to determine the optimization direction of the individualized tooth separation scheme based on the adjustment parameters of the real-time dental arch curve.

[0123] The feedback module is used to perform feedback loops on the overall tooth spatial position file using the optimized direction to obtain the final dynamic integrated measurement result.

[0124] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A precision measurement method for orthodontic tooth separation based on digitized scanning, characterized in that, Comprise: Collecting initial three-dimensional scanning data of patient's teeth to obtain dynamic change data of teeth; Correlation mapping processing is performed on the dynamic change data of the teeth to obtain a tooth spatial position archive; Based on the tooth spatial position archive, an iterative comparison algorithm is used to match subsequent scanning data to judge the deviation value of the interdental space width and the root inclination in the orthodontic process; Analyzing the deviation value to obtain real-time arch curve adjustment parameters; Based on the real-time arch curve adjustment parameters, the optimization direction of the individualized tooth separation scheme is determined; Using the optimization direction to perform feedback cycle on the overall tooth spatial position archive to obtain the final dynamic integration measurement result.

2. The precision measurement method of orthodontic tooth separation based on digitized scanning according to claim 1, characterized in that, Collecting initial three-dimensional scanning data of patient's teeth to obtain dynamic change data of teeth comprises: Collecting initial three-dimensional scanning data of patient's teeth, using convolutional neural network to extract features of the data, and obtaining multi-dimensional parameters; wherein the multi-dimensional parameters include: interdental space width and root inclination; According to the multi-dimensional parameters, a multi-time point data sequence is established, and the multi-time point data sequence is processed by a time series analysis model to obtain the dynamic change data of the teeth.

3. The precision measurement method of orthodontic tooth separation based on digitized scanning according to claim 2, characterized in that, Obtaining the multi-dimensional parameters comprises: Collecting three-dimensional scanning data of patient's teeth through a three-dimensional scanning device to generate initial point cloud data; Based on the initial point cloud data, a point cloud registration algorithm is used to automatically align multi-angle scanning results, ensure data integrity and calculate surface curvature, and form a high-resolution three-dimensional grid model; The three-dimensional grid model is converted into an image tensor and input into a convolutional neural network to obtain a feature vector; Calculate the gradient of the feature vector to quantify the change of the tooth shape, and finally output a multi-dimensional parameter vector; wherein the interdental space width is the average value measured by the Euclidean distance algorithm between the adjacent teeth, and the root inclination is the quantitative value of the angle between the root axis and the crown axis calculated by principal component analysis.

4. The precision measurement method of orthodontic tooth separation based on digitized scanning according to claim 3, characterized in that, Converting the three-dimensional grid model into an image tensor and inputting it into a convolutional neural network comprises: First, a 7x7 convolution kernel is used for preliminary feature extraction to extract edge and texture information, then a plurality of residual block processing is performed to gradually reduce the feature map size, and a maximum pooling layer is applied to reduce noise.

5. The precision measurement method of orthodontic tooth separation based on digitized scanning according to claim 1, wherein, Correlation mapping processing is performed on the dynamic change data of the teeth to obtain a tooth spatial position archive comprises: Correlation mapping processing is performed on the multi-dimensional parameters in the dynamic change data of the teeth to obtain a preliminary integration result; According to the preliminary integration result, a clustering algorithm is used to group tooth spatial data to determine a position feature set; Obtain the space dynamic index from the position feature set to determine whether it meets the archive standard; If the space dynamic index meets the archive standard, an optimization parameter group is obtained by predicting the trend change through a linear regression model; For the optimization parameter group, correlation mapping is performed on the archive elements to determine the final tooth spatial position archive.

6. The precision measurement method of orthodontic tooth separation based on digitized scanning according to claim 1, wherein, Based on the tooth spatial position archive, an iterative comparison algorithm is used to match subsequent scanning data to judge the deviation value of the interdental space width and the root inclination in the orthodontic process comprises: Extracting initial tooth spatial coordinates from the tooth spatial position archive to obtain a tooth position reference; Determine a tooth spatial matching point set by processing subsequent scanning data and a tooth position reference through an iterative comparison algorithm; Calculate an interproximal width change according to the tooth spatial matching point set, and determine a deviation value of the interproximal width; If the width deviation exceeds a preset threshold, extract a tooth root axis from the tooth spatial matching point set to obtain a tooth root inclination angle; Analyze the tooth root inclination angle and the initial tooth spatial coordinates by using a principal component analysis algorithm to determine a deviation value of the tooth root inclination.

7. The precision measurement method of orthodontic tooth separation based on digitized scanning according to claim 1, wherein, The analysis of the deviation value includes: Based on the deviation value, obtain continuous tracking records from a data sequence to obtain a deviation distribution; According to the deviation distribution, use a linear regression model to fit the continuous tracking records to determine a fitting coefficient; If the fitting coefficient exceeds a preset threshold, extract abnormal points from the continuous tracking records to obtain an abnormal sequence; Filter the abnormal sequence to update the continuous tracking records in the data sequence to obtain an updated sequence; According to the updated sequence, calculate a dental arch curve offset, and determine an offset vector; Group the offset vectors by using a k-means clustering, determine a cluster center, and obtain a center coordinate; Generate real-time adjustment parameters of the dental arch curve by interpolation of the center coordinate.

8. The precision measurement method of orthodontic tooth separation based on digitized scanning according to claim 1, wherein, Based on the real-time dental arch curve adjustment parameters, determine the optimization direction of the individual tooth separation scheme includes: Obtain real-time dental arch curve data, extract key point coordinates through an image processing link to obtain a curve parameter set; According to the curve parameter set, use a linear regression model to fit a trend line to determine a parameter adjustment amplitude; If the parameter adjustment amplitude exceeds a preset threshold, group similar curve segments by clustering analysis to obtain a multi-time point feature group; For the multi-time point feature group, construct a comparison matrix framework, judge the difference between the matrix elements, and obtain a time comparison result; Through the time comparison result, fuse individual scheme data, use a decision tree model to classify and optimize the path, and determine the tooth separation scheme variation point; Obtain the tooth separation scheme variation point, calculate the scheme optimization vector, judge the consistency of the vector direction, and obtain the optimization direction index.

9. A digital scanning-based orthodontic tooth segmentation precision measurement system, characterized in that, The system is used to implement the digital scanning-based orthodontic tooth separation precision measurement method according to any one of claims 1-8, and the system comprises: A data acquisition module is configured to acquire initial three-dimensional scanning data of patient teeth and obtain tooth dynamic change data; A mapping processing module is configured to perform correlation mapping processing on the tooth dynamic change data to obtain a tooth spatial position archive; A matching module is configured to match subsequent scanning data with the tooth spatial position archive based on an iterative comparison algorithm to determine deviation values of interproximal width and tooth root inclination in an orthodontic process; An analysis module is configured to analyze the deviation values to obtain real-time dental arch curve adjustment parameters; A determination module is configured to determine an optimization direction of an individual tooth separation scheme based on the real-time dental arch curve adjustment parameters; A feedback module is configured to perform feedback cycles on the overall tooth spatial position archive by using the optimization direction to obtain a final dynamic integration measurement result.

Images