Dental caries visual identification method and system for oral implantation

Abstract

The invention relates to the technical field of image data processing, in particular to a decayed tooth visual identification method and system for oral implantation, and the method comprises the steps: screening out a target tooth region from an obtained root tip piece to be identified, carrying out gray distribution analysis processing and brightness analysis processing on a preset neighborhood corresponding to each pixel point in each target tooth region; screening out a suspected caries lesion region; drawing a target central axis of a target tooth area to which each suspected caries area belongs, determining a vertical line of the target central axis as a target vertical line, translating the target vertical line, and forming an intersection point group by intersection points of the target vertical line translated each time and the edge of the suspected caries area; screening out a suspected medullary cavity point group from the intersection point group set corresponding to each suspected caries lesion region; and screening out the target caries damage region according to all the determined medullary cavity possibilities. According to the invention, image data processing is carried out on the to-be-identified root tip piece, so that the accuracy of identifying the caries lesion region is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and particularly to a visual recognition method and system for dental caries for oral implantation. Background Art

[0002] With the development of technology, the application of image processing is becoming more and more extensive. For example, it can be applied to the recognition of dental caries. At present, when recognizing an object, the commonly used method is: according to the collected object image, the object is recognized by the difference in gray values.

[0003] However, when recognizing dental caries according to the collected periapical film by the difference in gray values, the following technical problems often exist: During the collection process of the periapical film, due to various factors such as improper fixation of the film position or improper vertical angle of the X-ray, the periapical film may appear blurred or have artifacts, so that the imaging gray situation of the tooth structure itself such as the pulp cavity is similar to that of the carious lesion area. Therefore, when recognizing dental caries containing the carious lesion area, if only considering the difference in gray values, it is often possible to misjudge the carious lesion pixel points, resulting in poor accuracy in recognizing the carious lesion area, and further possibly poor accuracy in recognizing dental caries containing the carious lesion area. Summary of the Invention

[0004] In order to solve the technical problem of poor accuracy in recognizing the carious lesion area, the present invention proposes a visual recognition method and system for dental caries for oral implantation.

[0005] In a first aspect, the present invention provides a visual recognition method for dental caries for oral implantation, the method comprising: Screening out target tooth regions from the obtained periapical films to be recognized, and performing gray distribution analysis processing and brightness analysis processing on the preset neighborhood corresponding to each pixel point in each target tooth region to obtain the gray distribution feature and the target brightness feature corresponding to each pixel point in each target tooth region; Screening out suspected carious lesion regions from the target tooth regions according to all the gray distribution features and the target brightness features; Constructing the target central axis of the target tooth region to which each suspected carious lesion region belongs, determining the perpendicular line of the target central axis as the target perpendicular line, translating the target perpendicular line, and forming an intersection point group with the intersection points of the translated target perpendicular line and the edge of the suspected carious lesion region each time, to obtain a set of intersection point groups corresponding to each suspected carious lesion region; Screening out suspected pulp cavity point groups from the set of intersection point groups corresponding to each suspected carious lesion region to obtain a set of suspected pulp cavity point groups corresponding to each suspected carious lesion region; Determine the pulp cavity possibility corresponding to each suspected caries lesion area according to the contour chain code corresponding to each suspected caries lesion area, the distance between each suspected caries lesion area and the root apex endpoint within its target tooth area, and the set of suspected pulp cavity point groups corresponding to each suspected caries lesion area; Screen out the target caries lesion areas from all the suspected caries lesion areas according to all the pulp cavity possibilities.

[0006] Combined with the first aspect above, in a possible implementation manner, the screening out of the target tooth areas from the obtained root tip radiographs to be recognized includes: Perform edge detection on the root tip radiograph to be recognized, and perform morphological closing operation to obtain a target image; Determine the area enclosed by each closed contour in the target image as a suspected tooth area; Determine the maximum value among all the continuous recurrence times corresponding to all the chain code values in the contour chain code of each suspected tooth area as the target continuous recurrence times corresponding to each suspected tooth area; Determine the ratio of the area of each suspected tooth area to the area of its minimum circumscribed ellipse as the target shape feature corresponding to each suspected tooth area; Determine the aspect ratio of the length to the width of the minimum circumscribed rectangle corresponding to each suspected tooth area as the shape extension feature corresponding to each suspected tooth area; Determine the tooth shape factor corresponding to each suspected tooth area according to the target continuous recurrence times, target shape feature and shape extension feature corresponding to each suspected tooth area, wherein the target continuous recurrence times, target shape feature and shape extension feature are all positively correlated with the tooth shape factor; Normalize the product value of the target continuous recurrence times, target shape feature and shape extension feature corresponding to each suspected tooth area to obtain the tooth shape factor corresponding to each suspected tooth area; If the tooth shape factor corresponding to the suspected tooth area is greater than the preset tooth shape threshold, then determine the suspected tooth area as the target tooth area.

[0007] Combined with the first aspect above, in a possible implementation manner, the gray distribution analysis processing and brightness analysis processing of the preset neighborhood corresponding to each pixel point in each target tooth area to obtain the gray distribution feature and target brightness feature corresponding to each pixel point in each target tooth area includes: Determine any one of the target tooth areas as the marked tooth area, and determine any one pixel point in the marked tooth area as the marked pixel point; Determine the mean value of the gray values of all the pixel points in the preset neighborhood corresponding to the marked pixel point as the neighborhood representative gray factor corresponding to the marked pixel point; Determine the absolute value of the difference between the neighborhood representative gray factor corresponding to the marked pixel point and the gray value corresponding to each pixel point within its corresponding preset neighborhood as the target difference corresponding to each pixel point within the preset neighborhood of the marked pixel point; Determine the cumulative value of the target differences corresponding to all pixel points within the preset neighborhood of the marked pixel point as the gray distribution feature corresponding to the marked pixel point; Determine the average value of the gray values corresponding to all pixel points within the marked tooth area as the tooth representative gray factor corresponding to the marked tooth area; Determine the ratio of the neighborhood representative gray factor to the tooth representative gray factor as the target brightness feature corresponding to the marked pixel point.

[0008] Combined with the first aspect above, in a possible implementation manner, the screening of the suspected caries area from the target tooth area according to all gray distribution features and target brightness features includes: Determine the suspected caries factor corresponding to each pixel point within each target tooth area according to the gray distribution feature and the target brightness feature corresponding to each pixel point within each target tooth area, where the gray distribution feature is positively correlated with the suspected caries factor, and the target brightness feature is negatively correlated with the suspected caries factor; If the suspected caries factor corresponding to a pixel point is greater than the preset caries threshold, then determine the pixel point as a suspected caries pixel point; Perform connected component extraction on the area composed of all suspected caries pixel points, and determine the extracted connected component as the suspected caries area.

[0009] Combined with the first aspect above, in a possible implementation manner, the screening of the suspected pulp cavity point group from the set of intersection point groups corresponding to each suspected caries area includes: Screen out the intersection point group with the number of intersection points being 2 and the intersection points being located on both sides of the target central axis from the set of intersection point groups corresponding to each suspected caries area as the suspected pulp cavity point group.

[0010] Combined with the first aspect above, in a possible implementation manner, the formula for the pulp cavity possibility corresponding to the suspected caries area is: ; ; ; ; Among them, is the pulp cavity possibility corresponding to the th suspected caries area; is the serial number of the suspected caries area; is a normalization function; is the maximum value among the consecutive repeated occurrences corresponding to all the chain code values in the contour chain code of the th suspected caries lesion area; is the function value when the independent variable takes the value of , and is negatively correlated with ; is a factor preset to be greater than 0; is the average value of the suspected caries factors corresponding to all the pixel points in the th suspected caries lesion area; is the distance between the th suspected caries lesion area and a random root endpoint within its target tooth area; is an activation function for normalization; is the number of suspected pulp cavity point groups in the set of suspected pulp cavity point groups corresponding to the th suspected caries lesion area; is the number of pixel points on the edge of the th suspected caries lesion area; is the function value when the independent variable takes the value of , and is negatively correlated with ; is a factor preset to be greater than 0; characterizes the average distance difference between the two intersection points and the target central axis in all the suspected pulp cavity point groups corresponding to the th suspected caries lesion area; is the serial number of the suspected pulp cavity point group in the set of suspected pulp cavity point groups corresponding to the th suspected caries lesion area; is an absolute value function; is the distance between the first intersection point and the target central axis of the th suspected caries lesion area corresponding to the th suspected pulp cavity point group in the set of suspected pulp cavity point groups, where the th suspected caries lesion area belongs to the target tooth area; is the distance between the second intersection point and the target central axis of the th suspected caries lesion area corresponding to the th suspected pulp cavity point group in the set of suspected pulp cavity point groups, where the th suspected caries lesion area belongs to the target tooth area.

[0011] Combined with the above first aspect, in a possible implementation manner, the method further includes: Determine the caries severity index corresponding to each target caries lesion area according to the caries extension direction of each target caries lesion area, the tooth extension direction of the target tooth area to which it belongs, and the area of each target caries lesion area; Determine the target tooth area to which the target caries lesion area belongs as a candidate tooth area, and screen out the pixel points adjacent to the target gingival area from the contour of each candidate tooth area as target contour points, where the target gingival area is the gingival area in the periapical film to be recognized; Screen out gingival inflammation points from the target gingival area, screen out inflammation contour points from all target contour points, screen out the sub-gingival area corresponding to each candidate tooth area from the target gingival area, and screen out effective caries lesion areas from all target caries lesion areas according to the caries severity index corresponding to the target caries lesion area, where the effective caries lesion area represents the target caries lesion area with a changing trend of caries severity; Determine the caries severity level corresponding to each candidate tooth area according to the number of gingival inflammation points in the sub-gingival area corresponding to each candidate tooth area, the number of inflammation contour points in each candidate tooth area, and the number of effective caries lesion areas, where the number of gingival inflammation points, the number of inflammation contour points, and the number of effective caries lesion areas are all positively correlated with the caries severity level.

[0012] Combined with the first aspect above, in a possible implementation manner, the determining the caries severity index corresponding to each target caries lesion area according to the caries extension direction of each target caries lesion area, the tooth extension direction of the target tooth area to which it belongs, and the area of each target caries lesion area includes: Determine the included angle between the caries extension direction of each target caries lesion area and the tooth extension direction of the target tooth area to which it belongs as the caries extension feature corresponding to each target caries lesion area; Determine the caries severity index corresponding to each target caries lesion area according to the pulp cavity possibility, caries extension feature, and area corresponding to each target caries lesion area, where the pulp cavity possibility and the caries extension feature are both negatively correlated with the caries severity index, and the area of the target caries lesion area is positively correlated with its caries extension feature.

[0013] Combined with the first aspect above, in a possible implementation manner, the screening out effective caries lesion areas from all target caries lesion areas according to the caries severity index corresponding to the target caries lesion area includes: Sort the target caries lesion areas in the target tooth area in sequence along the tooth extension direction of the target tooth area, determine any target caries lesion area in the target tooth area as a marked caries lesion area, and if the caries severity index corresponding to the marked caries lesion area is less than the caries severity index corresponding to the next target caries lesion area, then determine the marked caries lesion area as an effective caries lesion area.

[0014] In a second aspect, the present invention provides a dental caries visual recognition system for oral implantation, comprising a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement the method in the first aspect or any possible implementation of the first aspect. A dental caries visual recognition system for oral implantation may specifically include: A screening processing module is used to screen out the target tooth area from the acquired periapical film to be identified, and perform grayscale distribution analysis and brightness analysis on the preset neighborhood corresponding to each pixel point in each target tooth area to obtain the grayscale distribution characteristics and target brightness characteristics corresponding to each pixel point in each target tooth area; A suspected caries area screening module is used to screen out suspected caries areas from the target tooth area based on all grayscale distribution characteristics and target brightness characteristics; A data straight line construction module is used to make a target central axis of a target tooth region to which each suspected caries region belongs, determine a perpendicular line to the target central axis as a target perpendicular line, translate the target perpendicular line, and form an intersection point group with the intersection points of the target perpendicular line after each translation and the edge of the suspected caries region to obtain a set of intersection point groups corresponding to each suspected caries region; A suspected pulp cavity point screening module is used to screen out a suspected pulp cavity point group from the intersection point group set corresponding to each suspected caries area, so as to obtain a suspected pulp cavity point group set corresponding to each suspected caries area; A possibility determination module is used to determine the pulp cavity possibility corresponding to each suspected caries area according to the contour chain code corresponding to each suspected caries area, the distance between each suspected caries area and the root end point in the target tooth area to which it belongs, and the suspected pulp cavity point group set corresponding to each suspected caries area; The target caries area screening module is used to screen out the target caries area from all suspected caries areas according to all pulp cavity possibilities.

[0015] In a third aspect, a server is provided, comprising a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, so that the device executes the method in the first aspect or any possible implementation of the first aspect.

[0016] In a fourth aspect, a computer program product is provided, comprising: a computer program code, which, when executed on a computer, enables the computer to execute the method in the first aspect or any possible implementation of the first aspect.

[0017] In a fifth aspect, a computer-readable storage medium is provided, which stores computer program code. When the computer program code runs on a computer, the computer is caused to execute the method in the first aspect or any possible implementation manner of the first aspect described above.

[0018] The present invention has the following beneficial effects: A method for visual recognition of dental caries for oral implant in the present invention realizes the recognition of caries-affected areas by performing image data processing on the root tip films to be recognized, thereby realizing the recognition of dental caries, solving the technical problem of poor accuracy in recognizing caries-affected areas, and improving the accuracy of recognizing caries-affected areas. Compared with recognizing dental caries only by considering the differences in gray values, the present invention comprehensively considers multiple indexes related to the dental caries situation of teeth, such as gray distribution characteristics and target brightness characteristics, so as to preliminarily screen out suspected caries-affected areas. However, in actual situations, due to the influence of various factors, there may be a certain similarity between the dental pulp cavity of teeth and the gray value of dental caries. Therefore, there may be dental pulp cavity areas among all the screened suspected caries-affected areas. Therefore, by analyzing the pulp cavity characteristics of the suspected caries-affected areas and quantifying multiple indexes related to the pulp cavity characteristics, such as suspected pulp cavity point groups, contour chain codes, and pulp cavity possibility, the target caries-affected areas representing the true caries-affected areas can be distinguished from the suspected caries-affected areas, thereby further improving the accuracy of recognizing caries-affected areas and thus improving the accuracy of dental caries recognition. Secondly, compared with the dental caries recognition achieved by relying on the subjective observation of dentists, the present invention quantifies multiple indexes related to the dental caries situation of teeth when performing dental caries recognition, and can objectively recognize caries-affected areas to a certain extent, thereby improving the accuracy of dental caries recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0020] Figure 1 is a flowchart of a method for visual recognition of dental caries for oral implant in the present invention; Figure 2 is a schematic diagram of the screening process of the suspected caries-affected areas in the present invention; Figure 3 is a schematic diagram of the determination process of the caries severity level in the present invention; Figure 4 is a schematic diagram of the composition structure of a system for visual recognition of dental caries for oral implant in the present invention; Figure 5 Schematic diagram of a computer device according to the present invention. Detailed implementation manners

[0021] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following describes in detail the specific implementation manners, structures, features, and effects of the technical solutions proposed according to the present invention in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0023] Reference Figure 1 , shows the flow of some embodiments of a method for visual recognition of dental caries for oral implant according to the present invention. The method for visual recognition of dental caries for oral implant includes the following steps: Step S1, screen out the target tooth regions from the obtained root tip films to be recognized, and perform gray-scale distribution analysis processing and brightness analysis processing on the preset neighborhoods corresponding to each pixel point in each target tooth region, so as to obtain the gray-scale distribution characteristics and target brightness characteristics corresponding to each pixel point in each target tooth region.

[0024] Among them, the root tip film to be recognized can be a root tip film to be recognized for dental caries. The root tip film, also known as an X-ray film, is the most common dental radiograph. Usually, only 2-3 teeth can be seen on one root tip film. Through the root tip film, it is often possible to observe whether there are lesions at the root of the tooth and whether the tooth is loose, etc. Therefore, the root tip film can be used to identify the caries situation of the tooth, so as to achieve the recognition of dental caries. Among them, the tooth with caries is the dental caries. The target tooth region can be the tooth region in the root tip film to be recognized. The preset neighborhood can be a pre-set neighborhood. For example, the preset neighborhood can be an eight-neighborhood.

[0025] It should be noted that dental caries is a common oral bacterial disease. As the disease progresses, the tooth changes in color and forms a substantial lesion. If not treated in time, with the demineralization of inorganic substances and the decomposition of organic substances, a caries cavity will form in the lesion area, damaging the tooth crown until the tooth is lost.

[0026] As an example, this step may include the following steps: The first step is to obtain the root tip film to be recognized.

[0027] For example, the periapical film of a patient can be collected by means of bisecting angle projection, and the collected periapical film can be preprocessed. The preprocessed periapical film is used as the periapical film to be recognized. Among them, the preprocessing may include, but is not limited to, filtering and denoising and image enhancement.

[0028] It should be noted that during the process of collecting the periapical film of a patient by means of bisecting angle projection, a film can be placed on the lingual side of the teeth in the patient's mouth, and the X-ray tube is perpendicular to the film and then exposed. The obtained X-ray image is the periapical film.

[0029] In the second step, edge detection is performed on the above-mentioned periapical film to be recognized, and morphological closing operation is performed to obtain a target image.

[0030] For example, the Canny operator can be used to perform edge detection on the above-mentioned periapical film to be recognized, and morphological closing operation is performed on the obtained edge detection image to bridge the narrow discontinuities in the figure and fill the breaks in the contour line. The image after the morphological closing operation is denoted as the target image.

[0031] In the third step, the area enclosed by each closed contour in the above-mentioned target image is determined as a suspected tooth area.

[0032] Among them, the contour is also called the edge.

[0033] It should be noted that in actual situations, the tooth contour often presents a closed contour. Therefore, the closed contour in the target image may be the tooth contour.

[0034] In the fourth step, the maximum value among the continuous recurrence times corresponding to all the chain code values in the contour chain code of each suspected tooth area is determined as the target continuous recurrence times corresponding to each suspected tooth area.

[0035] Among them, the contour chain code in the embodiments of the present invention can be an 8-neighborhood chain code.

[0036] For example, any one of the suspected tooth areas can be determined as a reference tooth area. According to the contour chain code of the reference tooth area, the chain code value corresponding to each pixel point on the contour of the reference tooth area is obtained, and based on the chain code values corresponding to all the pixel points on the contour of the reference tooth area, region growing is performed on all the pixel points on the contour of the reference tooth area. The number of pixel points in the region obtained by region growing can represent the continuous recurrence times of the corresponding chain code value. The number of pixel points in the largest region obtained by region growing at this time can be denoted as the target continuous recurrence times corresponding to the reference tooth area. Among them, the rule of region growing in the embodiments of the present invention can be that the chain code values are the same.

[0037] It should be noted that, generally, the more times the same chain code value of a certain area contour continuously repeats, the smoother the edge of the area is often indicated. Therefore, the larger the number of consecutive repetitions of the target corresponding to the suspected tooth area, the smoother the edge of the suspected tooth area is often indicated.

[0038] Step 5: Determine the ratio of the area of each suspected tooth area to the area of its minimum circumscribed ellipse as the target shape feature corresponding to each suspected tooth area.

[0039] It should be noted that in actual situations, the shape of teeth often approximates an ellipse. Therefore, when the ratio of the area of a suspected tooth area to the area of its minimum circumscribed ellipse is larger, it often indicates that the shape of the suspected tooth area is closer to an ellipse, often indicates that the shape of the suspected tooth area is closer to the shape of a tooth, and often indicates that the suspected tooth area is more likely to represent a tooth.

[0040] Step 6: Determine the aspect ratio of the length to the width of the minimum circumscribed rectangle corresponding to each suspected tooth area as the shape extension feature corresponding to each suspected tooth area.

[0041] It should be noted that in actual situations, the aspect ratio of the tooth area is often relatively large. Therefore, when the aspect ratio of the minimum circumscribed rectangle corresponding to a suspected tooth area is large, it often indicates that the suspected tooth area is more likely to represent a tooth.

[0042] Step 7: Determine the tooth shape factor corresponding to each suspected tooth area according to the target consecutive repetition times, target shape feature, and shape extension feature corresponding to each suspected tooth area.

[0043] Among them, the target consecutive repetition times, target shape feature, and shape extension feature can all be positively correlated with the tooth shape factor.

[0044] For example, the product value of the target consecutive repetition times, target shape feature, and shape extension feature corresponding to each suspected tooth area can be normalized to obtain the tooth shape factor corresponding to each suspected tooth area.

[0045] It should be noted that when the tooth shape factor corresponding to a suspected tooth area is larger, it often indicates that the edge of the suspected tooth area is smoother and that the suspected tooth area is more likely to represent a tooth.

[0046] Step 8: If the tooth shape factor corresponding to a suspected tooth area is greater than the preset tooth shape threshold, determine the suspected tooth area as the target tooth area.

[0047] Among them, the preset tooth shape threshold can be a threshold set in advance. For example, the preset tooth shape threshold can be 0.5.

[0048] It should be noted that the target tooth area can represent real teeth.

[0049] Step 9: Designate any one of the target tooth areas as the marked tooth area, and designate any one pixel point within the above-mentioned marked tooth area as the marked pixel point.

[0050] Step 10: Determine the mean value of the gray values of all pixel points within the preset neighborhood corresponding to the above-mentioned marked pixel point as the neighborhood representative gray factor corresponding to the above-mentioned marked pixel point.

[0051] Step 11: Determine the absolute value of the difference between the neighborhood representative gray factor corresponding to the above-mentioned marked pixel point and the gray value of each pixel point within its corresponding preset neighborhood as the target difference corresponding to each pixel point within the preset neighborhood corresponding to the above-mentioned marked pixel point.

[0052] Step 12: Determine the cumulative value of the target differences of all pixel points within the preset neighborhood corresponding to the above-mentioned marked pixel point as the gray distribution feature corresponding to the above-mentioned marked pixel point.

[0053] For example, the formula for determining the gray distribution feature corresponding to the pixel point within the target tooth area can be: ; where is the gray distribution feature corresponding to the th pixel point within the th target tooth area. is the serial number of the target tooth area. is the th serial number of the pixel point within the th target tooth area. is the number of pixel points within the preset neighborhood. is the absolute value function. is the th gray value of the th pixel point within the preset neighborhood corresponding to the th pixel point within the th target tooth area. is the th neighborhood representative gray factor corresponding to the th pixel point within the

[0054] It should be noted that since teeth are radiopaque to X-rays and the X-ray blocking capabilities of different regions inside teeth are different, the X-ray imaging results of different regions vary. For example, high-density tooth structures such as enamel, cementum, and dentin appear as bright and relatively evenly distributed white images under X-rays. However, carious lesions often damage the tooth structure, causing changes in the density of the normal tooth structure. For example, demineralization in the carious lesion area shows a shadow with increased transmittance in the hard tissue of the tooth. Therefore, carious lesion areas often appear as regions with darker brightness and uneven gray-scale distribution in X-ray films. When is larger, it often indicates that the gray-scale distribution in the preset neighborhood corresponding to the th pixel point is more uneven, and it often indicates that the gray-scale distribution around the th pixel point is more uneven, and it often indicates that the th pixel point is more likely to be a carious lesion pixel point.

[0055] Step 13: Determine the average value of the gray-scale values corresponding to all pixel points within the marked tooth region as the tooth representative gray-scale factor corresponding to the marked tooth region.

[0056] Step 14: Determine the ratio of the neighborhood representative gray-scale factor to the tooth representative gray-scale factor as the target brightness feature corresponding to the marked pixel point.

[0057] For example, the formula for determining the target brightness feature corresponding to the pixel point within the target tooth region can be: ; where is the target brightness feature corresponding to the th pixel point within the th target tooth region. is the serial number of the target tooth region. is the serial number of the pixel point. is the serial number of the pixel point within the th target tooth region. is the neighborhood representative gray-scale factor corresponding to the th pixel point within the th target tooth region. is the tooth representative gray-scale factor corresponding to the th target tooth region, that is, the average value of the gray-scale values corresponding to all pixel points within the

[0058] It should be noted that carious lesion areas often appear as regions with darker brightness and uneven gray-scale distribution in X-ray films. When is larger, it often indicates that the gray-scale of the preset neighborhood corresponding to the th pixel point is relatively higher, and it often indicates that the The higher the relative gray level around a pixel, the more likely it is that the pixel is not a carious pixel.

[0059] Step S2: According to all the gray level distribution features and the target brightness feature, screen out the suspected carious regions from the target tooth regions.

[0060] As an example, the process of screening the suspected carious regions can be as Figure 2 shown, and specifically can include the following steps: Step 201: According to the gray level distribution feature and the target brightness feature corresponding to each pixel in each target tooth region, determine the suspected carious factor corresponding to each pixel in each target tooth region.

[0061] Among them, the gray level distribution feature can have a positive correlation with the suspected carious factor. The target brightness feature can have a negative correlation with the suspected carious factor.

[0062] For example, the formula for determining the suspected carious factor corresponding to the pixel in the target tooth region can be: ; where is the suspected carious factor corresponding to the th pixel in the th target tooth region. is the serial number of the target tooth region. is the serial number of the pixel in the th target tooth region. is the normalization function. is the gray level distribution feature corresponding to the th pixel in the th target tooth region. is the target brightness feature corresponding to the th pixel in the th target tooth region.

[0063] It should be noted that when is larger, it often indicates that the gray level distribution in the preset neighborhood corresponding to the th pixel is more uneven. When is smaller, it often indicates that the gray level of the preset neighborhood corresponding to the th pixel is relatively darker. Therefore, when is larger, it often indicates that the th pixel is more likely to be a carious pixel.

[0064] Step 202: If the suspected carious factor corresponding to the pixel is greater than the preset carious threshold, then determine the pixel as a suspected carious pixel.

[0065] Among them, the preset caries lesion threshold can be a pre-set threshold. For example, the preset caries lesion threshold can be 0.5.

[0066] Step 203: Extract the connected components of the area composed of all suspected caries lesion pixel points, and determine the extracted connected components as suspected caries lesion areas.

[0067] It should be noted that the target tooth area without a suspected caries lesion area can often represent a tooth without caries. That is to say, the tooth represented by the target tooth area without a suspected caries lesion area is often not a carious tooth.

[0068] Step S3: Draw the target central axis of the target tooth area to which each suspected caries lesion area belongs, determine the perpendicular line of the target central axis as the target perpendicular line, translate the target perpendicular line, and form an intersection group with the intersection points of the translated target perpendicular line and the edge of the suspected caries lesion area each time, so as to obtain a set of intersection groups corresponding to each suspected caries lesion area.

[0069] Among them, the target central axis of the target tooth area can be a straight line that roughly divides the target tooth area into two equal parts on the left and right.

[0070] It should be noted that due to factors such as improper shooting angle or exposure conditions, there may be artifacts and blurs in the X-ray imaging results, which may lead to an uneven gray-scale distribution in the X-ray imaging results of the pulp cavity area, and further may lead to misjudging the pulp cavity area as a caries lesion area. Therefore, the suspected caries lesion area may represent a real caries lesion area or a pulp cavity area. Secondly, since the pulp cavity area is usually symmetric about the tooth central axis, based on the target central axis and the set of intersection groups, it is convenient to distinguish the real caries lesion area and the pulp cavity area in the subsequent process.

[0071] As an example, this step may include the following steps: The first step: Connect the center point of the upper edge of the tooth crown of the target tooth area with the center of gravity of the target tooth area, and use the obtained connection line at this time as the target central axis of the target tooth area.

[0072] Among them, the upper edge of the tooth crown can be the edge of the tooth surface for chewing food. For example, the upper edge of the tooth crown of the target tooth area can be obtained through technologies such as neural networks or threshold segmentation.

[0073] In the second step, any suspected caries lesion area is determined as a reference caries lesion area, a perpendicular line to the target central axis of the target tooth area to which the reference caries lesion area belongs is determined as a reference perpendicular line, the reference perpendicular line is translated within the target tooth area to which the reference caries lesion area belongs, and the intersection points between the reference perpendicular line and the edge of the reference caries lesion area during each translation process form a set of intersection points of the reference caries lesion area, and all sets of intersection points of the reference caries lesion area form a set of intersection point sets corresponding to the reference caries lesion area.

[0074] Step S4, filter out the suspected pulp point groups from the set of intersection point sets corresponding to each suspected caries lesion area, and obtain the set of suspected pulp point groups corresponding to each suspected caries lesion area.

[0075] As an example, intersection point groups with 2 intersection points and the intersection points located on both sides of the target central axis can be filtered out from the set of intersection point sets corresponding to each suspected caries lesion area as the suspected pulp point groups.

[0076] For example, any suspected caries lesion area is determined as a reference caries lesion area, the target central axis of the target tooth area to which the reference caries lesion area belongs is determined as a reference central axis, and any suspected pulp point group in the set of suspected pulp point groups corresponding to the reference caries lesion area is determined as a reference suspected pulp point group. Then, the number of intersection points in the reference suspected pulp point group can be 2, and the 2 intersection points in the reference suspected pulp point group are respectively located on both sides of the reference central axis.

[0077] It should be noted that since the pulp cavity area is usually symmetric about the tooth central axis, the intersection points in the suspected pulp point group may be pulp cavity edge pixel points.

[0078] Step S5, determine the pulp cavity possibility corresponding to each suspected caries lesion area according to the contour chain code corresponding to each suspected caries lesion area, the distance between each suspected caries lesion area and the root apex point within the target tooth area to which it belongs, and the set of suspected pulp point groups corresponding to each suspected caries lesion area.

[0079] Among them, the root apex point is also called the root tip point. For example, the root apex point can be obtained through technologies such as neural networks or threshold segmentation.

[0080] As an example, the formula for determining the pulp cavity possibility corresponding to the suspected caries lesion area can be: ; ; where, is the pulp cavity possibility corresponding to the th suspected caries lesion area. is the serial number of the suspected caries lesion area. is the normalization function. is the The maximum value among the consecutive recurrence counts corresponding to all chain code values within the contour chain code of a suspected carious lesion area. is the function value when the independent variable takes the value of and is negatively correlated with For example, ; is a factor preset to be greater than 0, mainly used to prevent the denominator from being 0. For example, can be 0.001. is the mean value of the suspected carious lesion factors corresponding to all pixel points within the th suspected carious lesion area. is the distance between the th suspected carious lesion area and a random root tip point within its target tooth area. In actual situations, there are often teeth containing one or two roots. is an activation function used for normalization. is the number of suspected pulp cavity point groups in the set of suspected pulp cavity point groups corresponding to the th suspected carious lesion area. is the number of pixel points on the edge of the th suspected carious lesion area. is the function value when the independent variable takes the value of and is negatively correlated with For example, ; is a factor preset to be greater than 0, mainly used to prevent the denominator from being 0. For example, can be 0.001. characterizes the average distance difference between the two intersection points and the target central axis among all suspected pulp cavity point groups corresponding to the th suspected carious lesion area. is the serial number of the suspected pulp cavity point group in the set of suspected pulp cavity point groups corresponding to the th suspected carious lesion area. is the absolute value function. is the distance between the first intersection point and the target central axis of the target tooth area to which the th suspected carious lesion area belongs, in the th suspected pulp cavity point group in the set of suspected pulp cavity point groups corresponding to the th suspected carious lesion area. is the distance between the second intersection point and the target central axis of the target tooth area to which the th suspected carious lesion area belongs, in the th suspected pulp cavity point group in the set of suspected pulp cavity point groups corresponding to the th suspected carious lesion area.

[0081] It should be noted that in actual situations, carious lesions usually form in the dental crown, damaging the tooth structure. For example, demineralization of the hard tissue of the tooth, decomposition of organic matter, and dissolution of inorganic substances, etc., make the shape of the carious lesion area irregular, the edges uneven, and the position of the carious lesion area is mostly on the tooth surface, such as the occlusal surface and the proximal surface, etc., far from the tooth apex. The pulp cavity area is often located in the center of the tooth and extends towards the root, usually symmetric about the tooth axis, and its edges are smooth. When is larger, it often indicates that the th suspected carious lesion area has smoother edges, and it often indicates that the th suspected carious lesion area is more likely to be the pulp cavity area. When is larger, it often indicates that the th suspected carious lesion area is more likely to be a real carious lesion area. When is larger, the th suspected carious lesion area is farther from the tooth apex, and it often indicates that the th suspected carious lesion area is more likely to be a real carious lesion area. When is larger, it often indicates that the th suspected carious lesion area has a larger number of intersection points in the set of suspected pulp cavity point groups corresponding to it, and it often indicates that the th suspected carious lesion area has a larger number of possible pulp cavity pixel points, and it often indicates that the th suspected carious lesion area is more likely to be the pulp cavity area. When is smaller, it often indicates that the intersection points in the suspected pulp cavity point group corresponding to the th suspected carious lesion area are more likely to be symmetric about the target axis, and it often indicates that the th suspected carious lesion area is more likely to be the pulp cavity area. Therefore, when is larger, it often indicates that the th suspected carious lesion area is more likely to be the pulp cavity area rather than a real carious lesion area.

[0082] Step S6, screen out the target carious lesion area from all suspected carious lesion areas according to all the pulp cavity possibilities.

[0083] Among them, the target tooth area to which the target carious lesion area belongs can characterize dental caries.

[0084] As an example, if the pulp cavity possibility corresponding to the suspected carious lesion area is less than the preset carious lesion discrimination threshold, the suspected carious lesion area is determined as the target carious lesion area. Among them, the preset carious lesion discrimination threshold can be a threshold set in advance. For example, the preset carious lesion discrimination threshold can be 0.5.

[0085] Optionally, as Figure 3 shown, the present invention may further include the following steps: Step 301: Determine the caries severity index corresponding to each target caries lesion area according to the caries extension direction of each target caries lesion area, the tooth extension direction of its corresponding target tooth area, and the area of each target caries lesion area.

[0086] Among them, the caries extension direction can represent the caries diffusion direction. For example, the method for obtaining the caries extension direction of a target caries lesion area can be as follows: Determine any pixel point on the contour of the target caries lesion area as a marked caries pixel point, and determine the two pixel points adjacent to the marked caries pixel point on the contour of the target caries lesion area as the first pixel point and the second pixel point respectively; Determine the absolute value of the slope of the line connecting the first pixel point and the marked caries pixel point as the first change representative factor; Determine the absolute value of the slope of the line connecting the second pixel point and the marked caries pixel point as the second change representative factor; Determine the sum value of the first change representative factor and the second change representative factor as the target change index corresponding to the marked caries pixel point; Screen out a pixel point with the largest target change index from the contour of the target caries lesion area as the target extension endpoint; Determine the direction from the center of gravity of the target caries lesion area to the target extension endpoint as the caries extension direction of the target caries lesion area.

[0087] The tooth extension direction can represent the direction from the center of gravity of the tooth to the root. For example, if the tooth represented by the target tooth area has one root, the tooth extension direction of the target tooth area can be the direction from the center of gravity of the target tooth area to the root endpoint. If the tooth represented by the target tooth area has two roots, the tooth extension direction of the target tooth area can be: the direction of the resultant vector of the target direction vectors from the center of gravity of the target tooth area to the two root endpoints. The direction of the target direction vector from the center of gravity of the target tooth area to the root endpoint can be the direction from the center of gravity of the target tooth area to the root endpoint. The modulus of the target direction vector from the center of gravity of the target tooth area to the root endpoint can be equal to the distance between the center of gravity of the target tooth area and the root endpoint.

[0088] For example, determining the caries severity index corresponding to each target caries lesion area may include the following steps: First step: Determine the caries extension feature corresponding to each target caries lesion area as the included angle between the caries extension direction of each target caries lesion area and the tooth extension direction of its corresponding target tooth area.

[0089] It should be noted that when the caries extension feature corresponding to the target caries lesion area is smaller, it often means that the caries extension direction of the target caries lesion area is more similar to the tooth extension direction of its corresponding target tooth area, often means that the target caries lesion area is more likely to extend deep into the tooth, may affect the pulp cavity or even the root apex, and its caries degree may be more serious.

[0090] Step 2: Determine the caries severity index corresponding to each target caries lesion area according to the pulp cavity possibility, caries lesion extension characteristics, and area corresponding to each target caries lesion area.

[0091] Among them, both the pulp cavity possibility and the caries lesion extension characteristics can have a negative correlation with the caries severity index. The area of the target caries lesion area can have a positive correlation with its caries lesion extension characteristics.

[0092] For example, the formula for determining the caries severity index corresponding to the target caries lesion area can be: ; Where is the caries severity index corresponding to the th target caries lesion area. is the serial number of the target caries lesion area. is the th pulp cavity possibility corresponding to the target caries lesion area. is the th caries lesion extension characteristic corresponding to the target caries lesion area. is a preset factor greater than 0, mainly used to prevent the denominator from being 0. For example, can be 0.001. is the area of the th target caries lesion area. is the th area of the target tooth area to which the target caries lesion area belongs.

[0093] It should be noted that when is larger, it often means that the th suspected caries lesion area is more likely to be a pulp cavity area rather than a real caries lesion area, and it often means that there is less likely to be a more serious caries situation. When is smaller, it often means that the caries lesion extension direction of the th target caries lesion area is more similar to the tooth extension direction of the target tooth area to which it belongs, and it often means that the th target caries lesion area is more likely to extend deep into the tooth, and it often means that its caries degree may be relatively severe. When is larger, it often means that the area ratio of the th target caries lesion area is relatively larger. Therefore, when is larger, it often means that the th target caries lesion area is more likely to extend deep into the tooth, and its caries degree may be relatively severe, and it often means that the caries situation of the th target caries lesion area is more serious.

[0094] Step 302: Determine the target tooth area to which the target carious lesion area belongs as the candidate tooth area, and screen out the pixel points adjacent to the target gingival area from the contour of each candidate tooth area as the target contour points.

[0095] Among them, the target gingival area can be the gingival area in the apical film to be recognized.

[0096] It should be noted that in actual situations, teeth and gums are often connected. Therefore, the pixel points adjacent to the gums on the tooth contour are often the pixel points at the junction of the tooth and the gum. Therefore, the target contour points can represent the position points on the junction line between the tooth and the gum.

[0097] Step 303: Screen out the gingival inflammation points from the target gingival area, screen out the inflamed contour points from all the target contour points, screen out the sub-gingival areas corresponding to each candidate tooth area from the target gingival area, and screen out the effective carious lesion areas from all the target carious lesion areas according to the carious lesion severity index corresponding to the target carious lesion area.

[0098] Among them, the effective carious lesion area can represent the target carious lesion area with a trend of significant change in carious lesion severity.

[0099] It should be noted that areas with relatively severe carious lesions often affect the tooth apex, resulting in the occurrence of apical periodontitis. Therefore, screening out the gingival inflammation points can facilitate the subsequent analysis of the carious lesion situation in the target carious lesion area.

[0100] For example, screening out the gingival inflammation points from the target gingival area may include the following steps: First step, according to the gray values of the pixel points corresponding to the preset neighborhood of each pixel point in the target gingival area, the formula for determining the gingival inflammation possibility corresponding to each pixel point in the target gingival area can be: ; Among them, is the gingival inflammation possibility corresponding to the th pixel point in the target gingival area. is the serial number of the pixel point in the target gingival area. is the normalization function. is the maximum value of the gray values of all the pixel points corresponding to the preset neighborhood of the th pixel point in the target gingival area. is the minimum value of the gray values of all the pixel points corresponding to the preset neighborhood of the th pixel point in the target gingival area. is the natural exponential function. is the The mean of the grayscale values corresponding to all the pixel points within the preset neighborhood corresponding to a pixel point. Is the mean of the grayscale values corresponding to all the pixel points in the root tip film to be recognized.

[0101] It should be noted that the degree of inflammation in different positions of the gum is often different, and the corresponding grayscale is often different. Therefore, the grayscale change around the inflamed pixel points is often more unstable, and the pixel points in the inflamed area outside the root tip are often darker as a whole. When is larger, it often indicates that the larger the grayscale change within the preset neighborhood corresponding to a pixel point, it often indicates that the more likely the pixel point is an inflamed pixel point. When is smaller, it often indicates that the more likely the grayscale of the preset neighborhood corresponding to a pixel point is darker as a whole, it often indicates that the more likely the pixel point is an inflamed pixel point. Therefore, when is larger, it often indicates that the more likely the pixel point is a gum inflamed pixel point.

[0102] Second step, determine any pixel point within the target gum area as a reference gum point. If the gum inflammation possibility corresponding to the reference gum point is greater than the preset gum inflammation threshold, then determine the reference gum point as a gum inflamed point.

[0103] Among them, the preset gum inflammation threshold can be a threshold set in advance. For example, the preset gum inflammation threshold can be 0.5.

[0104] Another example, screening out the inflamed contour points from all the target contour points may include the following steps: First step, determine the contour inflammation possibility corresponding to each target contour point according to the gradient magnitude corresponding to each target contour point.

[0105] Among them, the gradient magnitude is also called the gradient size.

[0106] For example, the mean of the gradient magnitudes corresponding to all the target contour points can be determined as the gradient representative value, and the ratio of the gradient magnitude corresponding to each target contour point to the gradient representative value is normalized to obtain the contour inflammation possibility corresponding to each target contour point.

[0107] It should be noted that apical periodontitis makes the apical edge of the tooth area in the X-ray imaging form a relatively obvious light-dark boundary with the gum area. Therefore, the gradient magnitude corresponding to the pixel points on the inflamed root contour is often larger. Therefore, when the contour inflammation possibility corresponding to the target contour point is greater, it often indicates that the target contour point is more likely to be an inflamed root contour point.

[0108] Step 2: If the likelihood of inflammation of the contour corresponding to the target contour point is greater than the preset contour inflammation threshold, then determine the target contour point as an inflamed contour point.

[0109] Among them, the preset contour inflammation threshold can be a threshold set in advance. For example, the preset contour inflammation threshold can be 0.5.

[0110] It should be noted that the inflamed contour point can be an inflamed root contour point.

[0111] For another example, the method of screening out the sub-gingival area corresponding to each candidate tooth area from the target gingival area can be: using the intersection curve segment between each candidate tooth area and the target gingival area as the segmentation step length, and segmenting out the sub-gingival area corresponding to each candidate tooth area from the target gingival area. Among them, the sub-gingival area corresponding to the candidate tooth area can represent the part of the gingival area connected to the candidate tooth area.

[0112] For example, the method of screening the effective carious lesion area can be: along the tooth extension direction of the target tooth area, sequentially sort the target carious lesion areas in the target tooth area, that is, the closer the target carious lesion area is to the root, the larger its serial number. Determine any target carious lesion area in the target tooth area as the marked carious lesion area. If the carious lesion severity index corresponding to the above marked carious lesion area is less than the carious lesion severity index corresponding to the next target carious lesion area, then determine the above marked carious lesion area as the effective carious lesion area.

[0113] Step 304: Determine the carious lesion severity level corresponding to each candidate tooth area according to the number of gingival inflammation points in the sub-gingival area corresponding to each candidate tooth area, and the number of inflamed contour points and the number of effective carious lesion areas in each candidate tooth area.

[0114] Among them, the number of gingival inflammation points, the number of inflamed contour points, and the number of effective carious lesion areas can all have a positive correlation with the carious lesion severity level.

[0115] For example, the formula for determining the carious lesion severity level corresponding to the candidate tooth area can be: ; among them, is the carious lesion severity level corresponding to the th candidate tooth area. is the serial number of the candidate tooth area. is the normalization function. is the number of inflamed contour points in the th candidate tooth area. is the th candidate tooth area. is the The number of gingival inflammation points in the subgingival area corresponding to the candidate tooth area. is the number of effective caries lesions in the th candidate tooth area.

[0116] It should be noted that the inflammation contour points can represent the inflamed root contour points. When is larger, it often indicates that there are relatively more inflammation contour points in the th candidate tooth area, which often means that there are more inflamed root contour points on the tooth represented by the th candidate tooth area, which often means that the candidate tooth area is more likely to be damaged and inflamed, and often means that the caries condition of the tooth represented by the th candidate tooth area is more serious. When is larger, it often indicates that there are more gingival inflammation points in the subgingival area corresponding to the th candidate tooth area, which often means that the inflammation caused by the tooth represented by the th candidate tooth area is more serious. When is larger, it often indicates that there are relatively more caries lesions with a serious change trend in the th candidate tooth area, which often means that the caries condition of the tooth represented by the th candidate tooth area may be more serious. Therefore, when is larger, it often indicates that the caries condition of the tooth represented by the th candidate tooth area is more serious.

[0117] Secondly, the doctor can refer to the caries severity level corresponding to the candidate tooth area to judge the caries condition of the tooth represented by the reference candidate tooth area. Therefore, quantifying the caries severity level corresponding to the reference candidate tooth area can facilitate the doctor to judge the caries condition of the tooth, so as to facilitate the doctor to design the treatment plan for the caries of the tooth represented by the reference candidate tooth area. It is also possible to mark the reference candidate tooth area with different colors according to the different caries severity levels corresponding to different reference candidate tooth areas, so as to facilitate the doctor to observe.

[0118] Reference Figure 4 , based on the same inventive concept as the above method embodiment, the present invention provides a visual recognition system for dental caries for oral implant, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the above computer program is executed by the processor, it implements the steps of a visual recognition method for dental caries for oral implant, which may specifically include: The screening and processing module 401 is configured to screen out the target tooth regions from the acquired root tip films to be recognized, and perform gray-scale distribution analysis processing and brightness analysis processing on the preset neighborhoods corresponding to each pixel point in each target tooth region, so as to obtain the gray-scale distribution features and target brightness features corresponding to each pixel point in each target tooth region.

[0119] The suspected caries lesion region screening module 402 is configured to screen out the suspected caries lesion regions from the target tooth regions according to all the gray-scale distribution features and target brightness features.

[0120] The data straight line construction module 403 is configured to make the target central axis of the target tooth region to which each suspected caries lesion region belongs, determine the target perpendicular line as the perpendicular line of the target central axis, translate the target perpendicular line, and form an intersection point group with the intersection points of the translated target perpendicular line and the edge of the suspected caries lesion region each time, so as to obtain the set of intersection point groups corresponding to each suspected caries lesion region.

[0121] The suspected pulp cavity point screening module 404 is configured to screen out the suspected pulp cavity point groups from the set of intersection point groups corresponding to each suspected caries lesion region, so as to obtain the set of suspected pulp cavity point groups corresponding to each suspected caries lesion region.

[0122] The possibility determination module 405 is configured to determine the pulp cavity possibility corresponding to each suspected caries lesion region according to the contour chain code corresponding to each suspected caries lesion region, the distance between each suspected caries lesion region and the root tip endpoints in its target tooth region, and the set of suspected pulp cavity point groups corresponding to each suspected caries lesion region.

[0123] The target caries lesion region screening module 406 is configured to screen out the target caries lesion regions from all the suspected caries lesion regions according to all the pulp cavity possibilities.

[0124] Figure 5 It is a schematic structural diagram of a computer device provided by an embodiment of the present invention. Exemplarily, as Figure 5 shown, the computer device 500 includes: a memory 501, a processor 502, and a computer program 503 stored in the memory 501 and running on the processor 502. Wherein, when the processor 502 executes the computer program 503, the computer device can execute any one of the foregoing methods for visual recognition of dental caries for oral implantology.

[0125] Based on the same inventive concept as the above method embodiment, the present invention provides a server, including a memory and a processor. The memory is used to store executable program codes, and the processor is used to call and run the executable program codes from the memory, so that the device executes any one of the foregoing methods for visual recognition of dental caries for oral implantology.

[0126] Based on the same inventive concept as the above method embodiments, the present invention provides a computer program product, which includes: computer program code that, when running on a computer, causes the computer to execute any of the above methods for visual recognition of dental caries for oral implant.

[0127] Based on the same inventive concept as the above method embodiments, the present invention provides a computer-readable storage medium that stores computer program code, which, when running on a computer, causes the computer to execute any of the above methods for visual recognition of dental caries for oral implant.

[0128] In summary, compared with identifying dental caries only by considering the differences in gray values, the present invention comprehensively considers multiple indicators related to the dental caries condition of teeth, such as gray distribution characteristics and target brightness characteristics, so as to preliminarily screen out suspected caries regions. However, in actual situations, due to the influence of various factors, there may be a certain similarity between the dental pulp cavity and the dental caries gray value of teeth. Therefore, there may be dental pulp cavity regions among all the screened suspected caries regions. Thus, by analyzing the pulp cavity characteristics of the suspected caries regions and quantifying multiple indicators related to the pulp cavity characteristics, such as suspected pulp cavity point groups, contour chain codes, and pulp cavity possibilities, the target caries regions representing the true caries regions can be distinguished from the suspected caries regions, thereby improving the accuracy of identifying the caries regions and thus the accuracy of dental caries recognition. Secondly, compared with the dental caries recognition achieved by relying on the subjective observation of dentists, the present invention quantifies multiple indicators related to the dental caries condition of teeth when performing dental caries recognition, and to a certain extent, can more objectively identify the caries regions, thereby improving the accuracy of dental caries recognition.

[0129] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention and should all be included in the protection scope of the present invention.

Claims

1. A visual recognition method for dental caries in oral implant, characterized in that, Including the following steps: Screen out the target tooth regions from the obtained root tip films to be recognized, and perform gray-scale distribution analysis processing and brightness analysis processing on the preset neighborhoods corresponding to each pixel point in each target tooth region to obtain the gray-scale distribution characteristics and target brightness characteristics corresponding to each pixel point in each target tooth region; According to all the gray-scale distribution characteristics and target brightness characteristics, screen out the suspected caries regions from the target tooth regions; Construct the target central axis of the target tooth region to which each suspected caries region belongs, determine the target perpendicular line as the perpendicular line of the target central axis, translate the target perpendicular line, and form an intersection group with the intersection points of the translated target perpendicular line and the edge of the suspected caries region each time, to obtain the set of intersection groups corresponding to each suspected caries region; Screen out the suspected pulp point groups from the set of intersection groups corresponding to each suspected caries region to obtain the set of suspected pulp point groups corresponding to each suspected caries region; According to the contour chain code corresponding to each suspected caries region, the distance between each suspected caries region and the root tip endpoint in its target tooth region, and the set of suspected pulp point groups corresponding to each suspected caries region, determine the pulp possibility corresponding to each suspected caries region; According to all the pulp possibilities, screen out the target caries regions from all the suspected caries regions.

2. The visual recognition method for dental caries used in oral implant according to claim 1, characterized in that, The screening out of the target tooth regions from the obtained root tip films to be recognized includes: Perform edge detection on the root tip film to be recognized and perform morphological closing operation to obtain a target image; Determine the region enclosed by each closed contour in the target image as a suspected tooth region; Determine the maximum value among all the continuous recurrence times corresponding to all the chain code values in the contour chain code of each suspected tooth region as the target continuous recurrence times corresponding to each suspected tooth region; Determine the ratio of the area of each suspected tooth region to the area of its minimum circumscribed ellipse as the target shape feature corresponding to each suspected tooth region; Determine the aspect ratio of the length and width of the minimum circumscribed rectangle corresponding to each suspected tooth region as the shape extension feature corresponding to each suspected tooth region; Normalize the product value of the target continuous recurrence times, target shape feature, and shape extension feature corresponding to each suspected tooth region to obtain the tooth shape factor corresponding to each suspected tooth region; If the tooth shape factor corresponding to the suspected tooth region is greater than the preset tooth shape threshold, then determine the suspected tooth region as the target tooth region.

3. A visual recognition method for dental caries in oral implant according to claim 1, characterized in that, The performing of gray-scale distribution analysis processing and brightness analysis processing on the preset neighborhoods corresponding to each pixel point in each target tooth region to obtain the gray-scale distribution characteristics and target brightness characteristics corresponding to each pixel point in each target tooth region includes: Determine any one target tooth region as the marked tooth region and determine any one pixel point in the marked tooth region as the marked pixel point; Determine the mean value of the gray-scale values of all the pixel points in the preset neighborhood corresponding to the marked pixel point as the neighborhood representative gray-scale factor corresponding to the marked pixel point; Determine the absolute value of the difference between the neighborhood representative gray factor corresponding to the marked pixel point and the gray value corresponding to each pixel point within its corresponding preset neighborhood as the target difference corresponding to each pixel point within the preset neighborhood of the marked pixel point; Determine the cumulative value of the target differences corresponding to all pixel points within the preset neighborhood of the marked pixel point as the gray distribution feature corresponding to the marked pixel point; Determine the average value of the gray values of all pixel points within the marked tooth area as the tooth representative gray factor corresponding to the marked tooth area; Determine the ratio of the neighborhood representative gray factor to the tooth representative gray factor as the target brightness feature corresponding to the marked pixel point.

4. A visual recognition method for dental caries in oral implant according to claim 1, characterized in that, The screening of the suspected caries lesion area from the target tooth area according to all the gray distribution features and the target brightness features includes: Determine the suspected caries lesion factor corresponding to each pixel point within each target tooth area according to the gray distribution feature and the target brightness feature corresponding to each pixel point within each target tooth area, wherein the gray distribution feature is positively correlated with the suspected caries lesion factor, and the target brightness feature is negatively correlated with the suspected caries lesion factor; If the suspected caries lesion factor corresponding to a pixel point is greater than the preset caries lesion threshold, then determine the pixel point as a suspected caries lesion pixel point; Perform connected component extraction on the area composed of all suspected caries lesion pixel points, and determine the extracted connected component as the suspected caries lesion area.

5. A visual recognition method for dental caries in oral implant according to claim 4, characterized in that, The screening of the suspected pulp cavity point group from the set of intersection point groups corresponding to each suspected caries lesion area includes: Screen out the intersection point group with the number of intersection points being 2 and the intersection points being located on both sides of the target central axis from the set of intersection point groups corresponding to each suspected caries lesion area as the suspected pulp cavity point group.

6. The visual recognition method for dental caries used in oral implantation according to claim 5, characterized in that, The formula for the pulp cavity possibility corresponding to the suspected caries lesion area is: ； ； ； ； in, It is The probability of the pulp cavity corresponding to the suspected carious area; is the serial number of the suspected carious lesion area; is the normalization function; It is The maximum value of the number of consecutive repetitions of all chain code values ​​in the contour chain code of the suspected caries area; The independent variable takes the value The function value when , and and Negative correlation; is a factor that is preset to be greater than 0; It is The mean value of the suspected caries factors corresponding to all pixels in the suspected caries area; It is The distance between a suspected carious area and a random root endpoint in the target tooth area to which it belongs; is the activation function, used for normalization; It is The number of suspected pulp cavity point groups in the set of suspected pulp cavity point groups corresponding to the suspected caries area; It is The number of pixels on the edge of the suspected caries area; The independent variable takes the value The function value when , and and Negative correlation; is a factor that is preset to be greater than 0; Characterization The average distance difference between the two intersection points and the target central axis in all suspected pulp cavity point groups corresponding to the suspected caries area; It is The serial number of the suspected pulp cavity point group in the set of suspected pulp cavity point groups corresponding to the suspected caries area; It is the absolute value function; It is The first In the group of suspected medullary cavity points, the first intersection point and the The distance between the target central axes of the target tooth areas to which the suspected caries areas belong; It is The first Among the suspected medullary cavity point groups, the second intersection point The distance between the target central axes of the target tooth areas to which the suspected caries areas belong.

7. A visual recognition method for dental caries in oral implant according to claim 1, characterized in that, The method further includes: Determine the caries lesion severity index corresponding to each target caries lesion area according to the caries lesion extension direction of each target caries lesion area, the tooth extension direction of the target tooth area to which it belongs, and the area of each target caries lesion area; Determine the target tooth area to which the target caries lesion area belongs as the candidate tooth area, and screen out the pixel points adjacent to the target gingival area from the contour of each candidate tooth area as the target contour points, wherein the target gingival area is the gingival area in the to-be-recognized periapical radiograph; Screen out the gingival inflammation points from the target gingival area, screen out the inflamed contour points from all the target contour points, screen out the sub-gingival areas corresponding to each candidate tooth area from the target gingival area, and screen out the effective caries lesion areas from all the target caries lesion areas according to the caries lesion severity index corresponding to the target caries lesion area, wherein the effective caries lesion area represents the target caries lesion area with a caries lesion severity change trend; Determine the caries lesion severity level corresponding to each candidate tooth area according to the number of gingival inflammation points within the sub-gingival area corresponding to each candidate tooth area, the number of inflamed contour points within each candidate tooth area, and the number of effective caries lesion areas, wherein the number of gingival inflammation points, the number of inflamed contour points, and the number of effective caries lesion areas are all positively correlated with the caries lesion severity level.

8. A method for visual recognition of dental caries for oral implantation according to claim 7, characterized in that Determining a caries severity index corresponding to each target caries lesion area according to the caries lesion extension direction of each target caries lesion area, the tooth extension direction of the target tooth area to which it belongs, and the area of each target caries lesion area, including: Determining the included angle between the caries lesion extension direction of each target caries lesion area and the tooth extension direction of the target tooth area to which it belongs as the caries lesion extension feature corresponding to each target caries lesion area; Determining the caries severity index corresponding to each target caries lesion area according to the pulp cavity possibility, caries lesion extension feature, and area corresponding to each target caries lesion area, wherein both the pulp cavity possibility and the caries lesion extension feature are negatively correlated with the caries severity index, and the area of the target caries lesion area is positively correlated with its caries lesion extension feature.

9. A visual recognition method for dental caries in oral implant according to claim 7, characterized in that, Screening out effective caries lesion areas from all target caries lesion areas according to the caries severity index corresponding to the target caries lesion area, including: Sequentially sorting the target caries lesion areas within the target tooth area along the tooth extension direction of the target tooth area, determining any one target caries lesion area within the target tooth area as a marked caries lesion area, and if the caries severity index corresponding to the marked caries lesion area is less than the caries severity index corresponding to the next target caries lesion area, determining the marked caries lesion area as an effective caries lesion area.

10. A visual recognition system for dental caries in oral implant, characterized in that, Including a processor and a memory, the processor is configured to process instructions stored in the memory to implement a method for visual recognition of dental caries for oral implant as described in any one of claims 1-9.

Images