Eye chamber angle shape determination method and device, electronic equipment and storage medium

Abstract

This application provides a method, apparatus, electronic device, and storage medium for determining the morphology of the anterior chamber angle. The method first acquires UBM images of the left and right eyes and segments the cornea, iris, and lens. Then, it calculates the closest and furthest distances between the first iris and the corneal reference center point in the target direction, and determines whether an anterior chamber angle is closed or open based on the ratio of the two distances and a first threshold. This method transforms the angle problem into a distance problem for measurement, thus making the determination result more accurate, simple, and efficient. When the first anterior chamber angle is open, the first anterior chamber angle is corrected based on the second initial angle of the mirrored second anterior chamber angle and the mirror similarity between the left and right eye UBM images to obtain the target angle. Then, it determines whether it is a narrow angle or an open angle based on the target angle and the second threshold. Since the target angle integrates mirror and similarity factors, the error is effectively reduced, and the determination accuracy is significantly improved.

Description

Technical Field

[0001] This application relates to the field of medical auxiliary technology, and in particular to a method, device, electronic device and storage medium for determining the morphology of the anterior chamber angle. Background Technology

[0002] In the diagnosis of eye-related diseases, it is often necessary to determine the morphology of the anterior chamber angle. Anterior chamber angles are generally classified as closed or open angles, with open angles further divided into narrow and open angles. Different angle shapes often lead to different diagnostic results, making accurate determination of angle morphology crucial. Currently, the determination is primarily made by manually measuring the angle between the cornea and iris in a UBM image and then comparing this angle with the corresponding angle thresholds for each shape. However, manual measurement is susceptible to errors due to factors such as measurement techniques, easily leading to incorrect assessments and affecting subsequent diagnosis and treatment.

[0003] Therefore, the current method for determining the morphology of the anterior chamber angle suffers from a technical problem where large errors in manual measurement lead to misjudgments, and this needs to be improved. Summary of the Invention

[0004] This application provides a method, device, electronic device, and storage medium for determining the morphology of the anterior chamber angle, in order to alleviate the technical problem of misjudgment caused by large errors in manual measurement when determining the morphology of the anterior chamber angle.

[0005] To address the aforementioned technical problems, this application provides the following technical solutions:

[0006] This application provides a method for determining the morphology of the anterior chamber angle, including:

[0007] Obtain a pair of UBM images of the same eye, the pair of UBM images including a first UBM image and a second UBM image, the first UBM image being one of the left eye UBM image and the right eye UBM image, and the second UBM image being the other;

[0008] The trained image segmentation model is called to process the UBM image pairs to obtain the first segmentation boundary of the cornea, the second segmentation boundary of the iris, and the third segmentation boundary of the lens in each UBM image. In each UBM image, the iris includes a first iris and a second iris. The first iris is located on one side of the left and right sides of the eye, and the second iris is located on the other side.

[0009] A corneal reference center point is determined based on the first segmentation boundary of the cornea in the first UBM image. The closest distance and the farthest distance between the first iris and the corneal reference center point in the target direction are determined based on the second segmentation boundary of the first iris in the first UBM image and the corneal reference center point. When the ratio of the closest distance to the farthest distance is greater than a first threshold, the first anterior chamber angle on the side where the first iris is located is determined to be a closed angle. When the ratio is not greater than the first threshold, the first anterior chamber angle is determined to be a non-closed angle.

[0010] When the first anterior chamber angle is not closed, the first initial angle of the first anterior chamber angle is determined according to the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image. The second initial angle of the second anterior chamber angle on the side where the second iris is located in the second UBM image is determined according to the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image. The mirror similarity between the first UBM image and the second UBM image is obtained.

[0011] The first initial angle is corrected based on the second initial angle and the mirror similarity to obtain the target angle of the first ovarian angle. When the target angle is less than the second threshold, the first ovarian angle is determined to be a narrow angle. When the target angle is not less than the second threshold, the first ovarian angle is determined to be an open angle.

[0012] Meanwhile, this application also provides a device for determining the shape of the anterior chamber angle, including:

[0013] The acquisition module is used to acquire a pair of UBM images of the same eye in the same case. The pair of UBM images includes a first UBM image and a second UBM image. The first UBM image is one of the left eye UBM image and the right eye UBM image, and the second UBM image is the other one.

[0014] The processing module is used to call the trained image segmentation model to process the UBM image pairs and obtain the first segmentation boundary of the cornea, the second segmentation boundary of the iris and the third segmentation boundary of the lens in each UBM image. In each UBM image, the iris includes a first iris and a second iris. The first iris is located on one side of the left and right sides of the eye, and the second iris is located on the other side.

[0015] The first determination module is used to determine a corneal reference center point based on the first segmentation boundary of the cornea in the first UBM image, and to determine the closest distance and the farthest distance between the first iris and the corneal reference center point in the target direction based on the second segmentation boundary of the first iris in the first UBM image and the corneal reference center point. When the ratio of the closest distance to the farthest distance is greater than a first threshold, the first anterior chamber angle morphology on the side where the first iris is located is determined to be a closed angle. When the ratio is not greater than the first threshold, the first anterior chamber angle morphology is determined to be a non-closed angle.

[0016] The determination module is used to determine the first initial angle of the first anterior chamber angle based on the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image when the first anterior chamber angle is not closed; and to determine the second initial angle of the second anterior chamber angle on the side where the second iris is located in the second UBM image based on the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image; and to obtain the mirror similarity between the first UBM image and the second UBM image.

[0017] The second determination module is used to correct the first initial angle based on the second initial angle and the mirror similarity to obtain the target angle of the first ovarian angle. When the target angle is less than the second threshold, the first ovarian angle is determined to be a narrow angle. When the target angle is not less than the second threshold, the first ovarian angle is determined to be an open angle.

[0018] This application also provides an electronic device, including a memory and a processor; the memory stores an application program, and the processor is used to run the application program in the memory to perform the steps in the ocular angle morphology determination method described in any of the preceding claims.

[0019] This application provides a computer-readable storage medium storing multiple instructions adapted for loading by a processor to execute the steps in the above-described method for determining the morphology of the anterior chamber angle.

[0020] Beneficial Effects: This application provides a method, apparatus, electronic device, and storage medium for determining the morphology of the anterior chamber angle. The method first acquires UBM images of the left and right eyes of the same case, and then processes them using a trained image segmentation model. In each UBM image, the cornea, iris, and lens are segmented to obtain the segmentation boundaries of each object. Then, a corneal reference center point is determined using the corneal segmentation boundary as a reference, and the closest and furthest distances between the first iris and the corneal reference center point in the target direction are determined using the first iris segmentation boundary as a reference. Finally, the morphology of a certain anterior chamber angle is determined based on the relationship between the ratio of the closest to the furthest distance and a first threshold. This method transforms the angle problem into a distance problem for measurement, resulting in more accurate calculations and a simpler, more efficient determination method. When the first anterior chamber angle is not closed, it is corrected based on the second initial angle of the mirrored second anterior chamber angle and the mirror similarity between the left and right UBM images. The relationship between the corrected target angle and a second threshold is then used to determine whether it is a narrow or open angle. Since the final target angle incorporates the effects of mirror and similarity factors, the error between it and the true angle is effectively reduced, and the accuracy of determining whether it is an open or narrow angle is significantly improved. In other words, after obtaining the UBM image pair, this application does not require manual measurement but automatically determines the shape of each anterior chamber angle based on relevant algorithms, thus improving the accuracy of the determination. Attached Figure Description

[0021] The technical solution and other beneficial effects of this application will become apparent from the following detailed description of specific embodiments in conjunction with the accompanying drawings.

[0022] Figure 1 This is a schematic diagram illustrating an application scenario of the ocular angle morphology determination method provided in the embodiments of this application.

[0023] Figure 2 This is a schematic diagram of the first method for determining the anterior chamber angle morphology provided in an embodiment of this application.

[0024] Figure 3 This is a schematic diagram of an eye UBM image in an embodiment of this application.

[0025] Figure 4 This is a schematic diagram of three morphologies of the anterior chamber angle in the embodiments of this application.

[0026] Figure 5 This is a schematic diagram illustrating the calculation process of closed and open angles in the embodiments of this application.

[0027] Figure 6 This is a schematic diagram illustrating the calculation process of the first room angle in an embodiment of this application.

[0028] Figure 7This is a schematic diagram of a second method for determining the anterior chamber angle morphology provided in an embodiment of this application.

[0029] Figure 8 This is a schematic diagram of the structure of the ocular angle morphology determination device provided in the embodiments of this application.

[0030] Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0031] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0032] This application provides a method, apparatus, electronic device, and computer-readable storage medium for determining the morphology of the anterior chamber angle. The apparatus for determining the morphology of the anterior chamber angle can be integrated into an electronic device, which can be a server or a terminal or other similar device.

[0033] Please see Figure 1 , Figure 1 This is a schematic diagram of an application scenario for the ocular angle morphology determination method provided in this application embodiment. The scenario may include a terminal and a server. Terminals, servers, and terminals and servers are connected and communicate through the Internet or other means composed of various gateways. The application scenario includes an image receiving device 11 and a server 12. The image receiving device 11 may be a device with human-computer interaction function. The server 12 includes a local server and / or a remote server.

[0034] The image receiving device 11 and the server 12 are located in a wireless network or a wired network to enable data interaction between them, wherein:

[0035] The user inputs a pair of UBM images of the same eye (left and right eye UBM images) into the image receiving device 11. The image receiving device 11 sends the UBM image pair and an angle morphology determination request to the server 12. The server 12 calls a pre-trained model and a pre-set algorithm to automatically determine the morphology of any one of the left, right, left, and right angles of the left and right eyes in the UBM image pair, identifying whether each angle is open, narrow, or closed. The determination result can be something like "left eye left angle is narrow, left eye right angle is narrow, right eye left angle is open, right eye right angle is closed," etc. The determination result is then returned to the image receiving device 11, which displays the result to the user. This method eliminates the need for manual measurement and offers high accuracy.

[0036] It should be noted that, Figure 1 The system scenario diagram shown is merely an example. The servers and scenarios described in this application embodiment are for the purpose of more clearly illustrating the technical solutions of this application embodiment and do not constitute a limitation on the technical solutions provided in this application embodiment. As those skilled in the art will know, with the evolution of systems and the emergence of new business scenarios, the technical solutions provided in this application embodiment are also applicable to similar technical problems. Detailed descriptions are provided below. It should be noted that the order of description of the following embodiments is not intended to limit the preferred order of embodiments.

[0037] Please see Figure 2 , Figure 2 This is a schematic flowchart of the first method for determining the morphology of the anterior chamber angle provided in this application embodiment. The method specifically includes:

[0038] S1: Obtain a pair of UBM images of the same eye. The pair of UBM images includes a first UBM image and a second UBM image. The first UBM image is one of the left eye UBM image and the right eye UBM image, and the second UBM image is the other one.

[0039] UBM (Ultrasoud biomicroscope) is a commonly used basic examination instrument in ophthalmology. It uses high-frequency ultrasound imaging technology to perform contact examinations of the eye. During the examination, a topical anesthetic is first applied to the eye, and then the instrument's probe contacts the corneal surface to detect the location of the anterior chamber angle and obtain a UBM image of the eye. Figure 3 and Figure 4As shown, in a UBM image, the eye includes the corneal region, the iris region, and the lens region. The angle between the cornea and the iris is the anterior chamber angle θ. Based on the degree of opening or closing of the anterior chamber angle θ, its shape can be divided into closed and open angles. Open angles can be further subdivided into open and narrow angles. A closed angle can be less than 10 degrees, a narrow angle can be not less than 10 degrees and less than 49 degrees, and an open angle can be not less than 49 degrees. Of course, in different scenarios, the grading standards for each shape can also be other values. This application is mainly used to determine whether the anterior chamber angle shape is closed, open, or narrow.

[0040] In this embodiment, a case refers to the data acquisition object of UBM, such as a patient, volunteer, or other relevant personnel. For the same case, UBM is used to acquire images of the left and right eyes separately, resulting in a UBM image pair composed of a left-eye UBM image and a right-eye UBM image. In practical scenarios, depending on the needs, it is possible to determine the iridocorneal angle morphology of only the left eye or only the right eye. One of the left-eye UBM image and the right-eye UBM image is referred to as the first UBM image, and the other as the second UBM image. In the following embodiments, the first UBM image can refer to either the left-eye UBM image or the right-eye UBM image, and the second UBM image automatically refers to the other one, so that all operations performed on the left-eye UBM image are also applicable to the right-eye UBM image.

[0041] S2: Call the trained image segmentation model to process UBM image pairs and obtain the first segmentation boundary of the cornea, the second segmentation boundary of the iris, and the third segmentation boundary of the lens in each UBM image. In each UBM image, the iris includes a first iris and a second iris. The first iris is located on one side of the left and right sides of the eye, and the second iris is located on the other side.

[0042] A UNet++ image segmentation neural network model was constructed and trained. The trained image segmentation model can be used to segment the cornea, iris, and lens in UBM images. The first and second UBM images from a UBM image pair were input into the trained image segmentation model. For each UBM image, the first segmentation boundary of the cornea, the second segmentation boundary of the iris, and the third segmentation boundary of the lens can be obtained.

[0043] For both the left and right eyes, the iris includes a first iris located on the left side of the eye and a second iris located on the right side. Therefore, each UBM image contains two anterior chamber angles, one in the lower left corner and one in the lower right corner. In practical scenarios, depending on the requirements, the anterior chamber angle morphology can be determined only for the left or right side of the eye. The iris on one side of the eye is called the first iris, and the anterior chamber angle on the side containing the first iris is called the first anterior chamber angle. The iris on the other side is called the second iris, and the anterior chamber angle on the side containing the second iris is called the second anterior chamber angle. In the following embodiments, the first iris and the first anterior chamber angle can refer to either the iris and anterior chamber angle on the left or right side of the eye, while the second iris and the second anterior chamber angle automatically refer to the iris and anterior chamber angle on the other side. This ensures that all operations performed on the left side of the eye are also applicable to the right side.

[0044] S3: Determine the corneal reference center point based on the first segmentation boundary of the cornea in the first UBM image. Determine the closest distance and the farthest distance between the first iris and the corneal reference center point in the target direction based on the second segmentation boundary of the first iris in the first UBM image and the corneal reference center point. When the ratio of the closest distance to the farthest distance is greater than a first threshold, determine that the first anterior chamber angle on the side where the first iris is located is a closed angle. When the ratio is not greater than the first threshold, determine that the first anterior chamber angle is a non-closed angle.

[0045] In the first UBM image, the first segmentation boundary of the cornea reflects the overall morphology of the corneal region, and the second segmentation boundary of the first iris reflects the overall state of the first iris region. First, a corneal reference center point is determined based on the first segmentation boundary. This point approximates the centroid of the cornea and can be used to characterize the position and shape of the cornea. Then, the point on the second segmentation boundary that is closest to this point in the target direction is found, and this closest distance is denoted as d1. Simultaneously, the point that is furthest away is found, and this furthest distance is denoted as d2. The target direction is the direction perpendicular to the left or right border of the minimum bounding rectangle of the first iris. For example, when the lines containing the left and right borders of the minimum bounding rectangle are vertical, the target direction is horizontal. Then, the ratio of d1 and d2 is calculated and compared with a first threshold T1, which can be 0.4 or other values.

[0046] When the ratio of d1 to d2 is large, it indicates that the horizontal span of the first iris is small, and the upper segmentation boundary of the first iris is more convex; or it indicates that the horizontal span of the cornea is large, and the lower segmentation boundary of the cornea is less convex. In both cases, the upper segmentation boundary of the first iris and the lower segmentation boundary of the cornea are closer together, and the angle between them is small. If the ratio is greater than the first threshold T1, it can be determined as a closed angle. Conversely, when the ratio of d1 to d2 is small, it indicates that the horizontal span of the first iris is large, and the upper segmentation boundary of the first iris is less convex; or the horizontal span of the cornea is small, and the lower segmentation boundary of the cornea is more convex. In both cases, the upper segmentation boundary of the first iris and the lower segmentation boundary of the cornea are less close together, and the angle between them is large. If the ratio is not greater than the first threshold T1, it can be determined as a non-closed angle.

[0047] By using the above method, the angle problem can be cleverly transformed into a distance problem for measurement, making the judgment method simpler and more efficient.

[0048] In one embodiment, S3 specifically includes: determining a first minimum bounding rectangle of the cornea based on the first segmentation boundary of the cornea in the first UBM image, and determining the center of the first minimum bounding rectangle as the corneal reference center point; determining a second minimum bounding rectangle of the first iris based on the second segmentation boundary of the first iris in the first UBM image, and determining the closest distance and the farthest distance between the first iris and the corneal reference center point in the target direction based on the left and right borders of the second minimum bounding rectangle and the corneal reference center point.

[0049] like Figure 5 As shown, when calculating each minimum bounding rectangle, the segmented UBM image pairs are first binarized to make the area inside each segmentation boundary white and the area outside each segmentation boundary black, which facilitates the calculation of the first and second minimum bounding rectangles. The first minimum bounding rectangle is the smallest rectangle that can completely cover the cornea, and the second minimum bounding rectangle is the smallest rectangle that can completely cover the first iris. Taking the first iris as being located on the left side of the eye as an example, after calculating the first and second minimum bounding rectangles, the center of the first minimum bounding rectangle is used as the corneal reference center. Then, the distances between the right and left borders of the second minimum bounding rectangle and the corneal reference center in the target direction are the closest distance d1 and the farthest distance d2, respectively.

[0050] Since the dividing boundary is an irregular line while the minimum bounding rectangle is a regular line, the method of obtaining the regular minimum bounding rectangle based on the dividing boundary and measuring it by the distance ratio is more accurate and simpler and more efficient than the method of directly measuring the angle based on the irregular dividing boundary.

[0051] S4: When the first anterior chamber angle is not closed, determine the first initial angle of the first anterior chamber angle based on the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image. Determine the second initial angle of the second anterior chamber angle on the side where the second iris is located in the second UBM image based on the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image, and obtain the mirror similarity between the first UBM image and the second UBM image.

[0052] If the first anterior chamber angle was determined to be a non-closed angle in the previous step, it is necessary to further determine whether it is an open or narrow angle. In this step, the included angle between the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image is calculated as the first initial angle of the first anterior chamber angle. Simultaneously, the included angle between the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image is calculated as the second initial angle of the second anterior chamber angle. The first and second anterior chamber angles are mirror images of each other; that is, when the first anterior chamber angle is to the left of the left eye, the second anterior chamber angle is to the right of the right eye, and vice versa. Furthermore, it is necessary to obtain the mirror similarity between the first and second UBM images. Mirror similarity refers to the similarity between a mirror image of one of the first and second UBM images and the original image of the other.

[0053] After determining that the first anterior chamber angle is a non-closed angle, the process further includes: performing a second binarization on the segmented UBM image pairs so that the area inside each segmentation boundary is a black area and the area outside each segmentation boundary is a white area, which improves the automatic measurement effect when calculating the included angle later.

[0054] In one embodiment, S4 specifically includes: determining a first intersection point of the cornea and the first iris in the first UBM image based on the first segmentation boundary of the cornea and the second segmentation boundary of the first UBM image; determining a second intersection point of the cornea and the second iris in the second UBM image based on the first segmentation boundary of the cornea and the second segmentation boundary of the second UBM image; and sequentially determining the first intersection point and the second intersection point as target points; setting i to 1, obtaining the i-th radius, and generating an i-th reference circle centered on the target point based on the i-th radius; using the target point as the vertex of the iris angle, obtaining the i-th candidate iris angle based on the first intersection point formed by the i-th reference circle and the first segmentation boundary, the second intersection point formed by the i-th reference circle and the second segmentation boundary, and the target point; and obtaining the i-th candidate angle of the i-th candidate iris angle; and cyclically executing the (i+1)-th radius generation operation, the (i+1)-th reference circle generation operation, the (i+1)-th candidate angle acquisition operation, and the determination of whether the variance of the first (i+1)-th candidate angles is greater than a third threshold. The operation includes: the (i+1)th radius generation operation, which generates the (i+1)th radius based on the ith radius and a preset radius incrementing parameter; the (i+1)th reference circle generation operation, which generates the (i+1)th reference circle centered on the target point based on the (i+1)th radius; and the (i+1)th candidate angle acquisition operation, which obtains the (i+1)th candidate angle based on the target point as the corner vertex, the first intersection point formed by the (i+1)th reference circle and the first dividing boundary, the second intersection point formed by the (i+1)th reference circle and the second dividing boundary, and the target point, and acquires the (i+1)th candidate angle of the (i+1)th candidate angle. The judgment operation includes judging whether the variance of the first (i+1)th candidate angles is greater than a third threshold. In each loop, if the judgment result is yes, when the target point is the first intersection point, the average value of the first (i+1)th candidate angles is determined as the first initial angle of the first angle; when the target point is the second intersection point, the average value of the first (i+1)th candidate angles is determined as the second initial angle of the second angle. If the judgment result is no, i is incremented by 1 and the next loop is performed.

[0055] like Figure 6 As shown, assuming that in the first UBM image, the first anterior chamber angle is located on the left side of the eye and the second anterior chamber angle is located on the right side of the eye, then the first segmentation boundary of the cornea and the second segmentation boundary of the first iris intersect at point O, and the first segmentation boundary of the cornea and the second segmentation boundary of the second iris intersect at point P.

[0056] First, take point O as the target point and set i to 1 to obtain the first radius, which is represented by r1. Draw the first reference circle with point O as the center and r1 as the radius. The first reference circle intersects the first segmentation boundary of the cornea at point A1 and the second segmentation boundary of the first iris at point B1. With point O as the vertex, the angle between OA1 and OB1 forms the first candidate anterior chamber angle. The i-th candidate angle θ1 of the first candidate anterior chamber angle is automatically measured.

[0057] Then, a loop is executed. In the first loop, a second radius is generated based on the first radius and a preset radius increment parameter. The preset radius increment parameter represents the magnitude of each radius increase. For example, it can be a uniform increase of one pixel width each time, or a uniform increase of one pixel width each time before exceeding a preset number of times, and a uniform increase of two pixels width each time after exceeding the preset number of times, etc. Then, a second reference circle is generated based on the second radius, with point O as the center. The second reference circle intersects the first segmentation boundary of the cornea at point A2 (not shown in the figure) and the second segmentation boundary of the first iris at point B2 (not shown in the figure). With point O as the vertex, the angle between OA2 and OB2 forms the second candidate anterior chamber angle. The second candidate angle θ2 of the second candidate anterior chamber angle is automatically measured and obtained. Finally, it is determined whether the variance std of the first two candidate angles θ1 and θ2 is greater than a third threshold.

[0058] If the variance std is greater than the third threshold, it means that the fluctuation of the current two candidate angles is large. Continuing to select more candidate angles will increase the error. At this time, we will no longer continue to calculate, but directly determine the average value of the two candidate angles as the first initial angle of the first room angle.

[0059] If the variance std is not greater than the third threshold, it means that the volatility of the current two candidate angles is still within expectations, and more candidate angles can be selected to participate in the calculation to improve accuracy. At this time, i is incremented by 1, and the next loop is started. The third radius generation operation, the third reference circle generation operation, the third candidate angle acquisition operation, and the judgment operation on whether the variance std of the first three candidate angles is greater than the third threshold are executed in sequence until the judgment result of a certain loop is yes. The loop stops, and the average value of all candidate angles obtained when the loop stops is taken as the first initial angle of the first room angle.

[0060] For the second anterior chamber angle, point P is used as the target point, and the above process is repeated to obtain the second initial angle of the second anterior chamber angle. The first and second initial angles obtained through the above process are the average values ​​of multiple candidate angles, and the fluctuation of multiple candidate angles is small, so the calculation results are relatively accurate.

[0061] In one embodiment, S4 specifically includes: determining one of the first UBM image and the second UBM image as the original image, and mirroring the other to obtain a mirror image; converting the original image and the mirror image to grayscale and adjusting them to a preset width and a preset height; determining a preset sampling precision based on the preset width and preset height; determining N original sampling pixels in the original image and N mirror sampling pixels in the mirror image based on the preset sampling precision; forming a pixel pair with each original pixel sampling pixel and its corresponding mirror sampling pixel; obtaining the grayscale value of the original sampling pixel and the grayscale value of the mirror sampling pixel in each pixel pair to obtain N sets of grayscale differences; obtaining the global similarity between the original image and the mirror image based on the N sets of grayscale differences; and obtaining the mirror similarity between the first UBM image and the second UBM image based on the global similarity, the preset width, and the preset height.

[0062] One of the first and second UBM images is designated as the original image, denoted as img10. The other is mirrored to obtain a mirror image, denoted as img01. Both img01 and img10 are converted to grayscale, ensuring that the RGB channel values ​​of each pixel in the image are equal. Then, both are adjusted to a preset width W and a preset height H, such as W = 512 and H = 512, or W = 224 and H = 224, or other values, all in units of one pixel width. These preprocessing steps remove image details, retaining only basic information such as structure and brightness, and discarding image differences caused by different sizes or proportions, thus avoiding unnecessary interference with subsequent similarity calculations.

[0063] Then, the preset sampling precision is determined based on W and H. For example, if W is 512 and H is 512, the preset sampling precision is 512*512. For img01, there are 512*512 mirrored sampling pixels, and for img10, there are 512*512 original sampling pixels. Each original sampling pixel and its corresponding mirrored sampling pixel form a pixel pair, resulting in a total of 512*512 pixel pairs. For each pixel pair, the grayscale value of the original sampling pixel and the grayscale value of the mirrored sampling pixel are obtained, and the two are subtracted to obtain the grayscale difference. This process is repeated for all pixel pairs, resulting in a total of 512*512 sets of grayscale differences. By combining the 512*512 groups of grayscale differences, the global similarity λ between img01 and img10 can be obtained. Let img01(i,j) represent the grayscale value of each mirrored pixel sampling point, and img10(i,j) represent the grayscale value of each original sampling pixel point. Then, the global similarity λ satisfies the following formula:

[0064]

[0065] λ is used to characterize the superposition of grayscale differences between 512*512 pixel pairs. A larger λ value indicates more dissimilar pixels, and vice versa. The mirror similarity between img01 and img10 is expressed by the following formula:

[0066]

[0067] The value ranges from 0 to 1. A larger value indicates less similarity between mg01 and img10, and vice versa. The mirror similarity obtained by the above method is more suitable for the application scenarios of UBM images in the medical field compared with existing hash algorithms, histograms, and other methods.

[0068] S5: Based on the second initial angle and mirror similarity, the first initial angle is corrected to obtain the target angle of the first anterior chamber angle. When the target angle is less than the second threshold, the shape of the first anterior chamber angle is determined to be a narrow angle. When the target angle is not less than the second threshold, the shape of the first anterior chamber angle is determined to be an open angle.

[0069] Since the first initial angle is obtained solely from the relevant segmentation boundaries of the first UBM image, and is influenced by various factors such as the initial shooting method, model segmentation effect, and subsequent measurement method, there is still a certain error between the room angle obtained from a single UBM image and the true angle. However, because the second room angle in the second UBM image is a mirror image of the first room angle, the second initial angle can be used to correct the first initial angle. Furthermore, considering that the second UBM image and the first UBM image are not completely mirror images but have a certain degree of mirror similarity, both the second initial angle and the mirror similarity can be combined to correct the first initial angle, thus obtaining the target angle. Because the target angle is obtained by combining relevant data from both the first and second UBM images, the error between it and the true angle is effectively reduced, and the accuracy is significantly improved.

[0070] After obtaining the target angle, it is compared with a second threshold T2, which can be 49 degrees or other values. If the target angle is less than 49 degrees, the first anterior chamber angle is determined to be a narrow angle; otherwise, it is determined to be an open angle.

[0071] In one embodiment, S5 specifically includes: obtaining the absolute difference between a first initial angle and a second initial angle; obtaining a correction factor based on the absolute difference and mirror similarity; comparing the magnitudes of the first initial angle and the second initial angle; determining the first angle and the second angle among the first initial angle and the second initial angle based on the comparison result; the first angle being greater than the second angle; when the first initial angle is the first angle, determining the difference between the first angle and the correction factor as the target angle of the first ventricular angle; when the first initial angle is the second angle, determining the sum of the second angle and the correction factor as the target angle of the first ventricular angle.

[0072] Let the first anterior chamber angle be located to the left of the left eye, and let the first initial angle be θ. Z0 This indicates that the second anterior chamber angle is located to the right of the right eye, and the second initial angle is represented by θ. Y1 Let the correction factor be K, then it satisfies the following condition:

[0073]

[0074] Let the target angle be θ Z0N Then it satisfies one of the following formulas:

[0075] θ Z0N =min(θ) Z0 ,θ Y1 )+K (Formula 4)

[0076] θ Z0N =max(θ) Z0 ,θ Y1 -K (Formula 5)

[0077] Specifically, if the second anterior chamber angle is closed, since the first anterior chamber angle is not closed, the first initial angle must be the first angle, and the target angle is calculated using Formula 5. If both the first and second anterior chamber angles are not closed, first compare θ. Z0 and θ Y1 The size of θ Z0 Greater than θ Y1 At this point, the first initial room angle is the first angle, and the target angle is calculated using formula 5. When θ Z0 Less than θ Y1 When the first initial angle is the second angle, the target angle is calculated using Formula 4. Verification in real-world scenarios shows that the correction method in this embodiment can meet the correction requirements in most cases, and the accuracy of the target angle's shape determination is high after correction.

[0078] In one embodiment, S5 specifically includes: obtaining the absolute difference between the first initial angle and the second initial angle; obtaining a correction factor based on the absolute difference and mirror similarity; when the first anterior chamber angle is located to the left of the left eye or to the right of the right eye, determining the sum of the first initial angle and the correction factor as the target angle of the first anterior chamber angle; when the first anterior chamber angle is located to the right of the left eye or to the left of the right eye, determining the difference between the first initial angle and the correction factor as the target angle of the first anterior chamber angle.

[0079] Due to factors such as the eye's structure, imaging methods, and measurement techniques, in certain scenarios, the measured value of the anterior chamber angle on the lateral side of the eye (including the left side of the left eye and the right side of the right eye) may be smaller, while the measured value of the anterior chamber angle on the medial side of the eye (including the right side of the left eye and the left side of the right eye) may be larger. In such cases, different correction methods can be used depending on the position of the first anterior chamber angle to obtain the target angle. Let the target angle be θ. Z0N Then it satisfies one of the following formulas:

[0080] θ Z0N =θ Z0 +K (Formula 6)

[0081] θ Z0N =θ Z0 -K (Formula 7)

[0082] Specifically, when the first anterior chamber angle is located to the left of the left eye or to the right of the right eye, Formula 6 is used for calculation; when the first anterior chamber angle is located to the right of the left eye or to the left of the right eye, Formula 7 is used for calculation.

[0083] Those skilled in the art can flexibly choose one of the two methods mentioned above to modify the target angle according to the needs of the actual scenario, and then determine the narrow angle and the open angle.

[0084] It should be noted that the above embodiments are all illustrated with the example of the first anterior chamber angle being located to the left of the left eye. When the first anterior chamber angle is located to the right of the left eye, to the left of the right eye, or to the right of the right eye, the target angle can also be obtained in the same way. That is, those skilled in the art can select at least one of the four anterior chamber angles of both eyes as the first anterior chamber angle as needed, and obtain the final determination result through the above process.

[0085] like Figure 7 The diagram shown is a second flowchart illustrating the ocular angle morphology determination method provided in this application embodiment. This flowchart serves as a summary description of the above embodiments and specifically includes:

[0086] S101: Begin.

[0087] S102: Obtain UBM image pairs, which include left-eye UBM images and right-eye UBM images.

[0088] S103: Segment each UBM image in the UBM image pair to obtain the corneal boundary, iris boundary, and lens boundary.

[0089] S104: In the UBM image pair, the four anterior chamber angles of the left eye (left side), left eye (right side), right eye (left side), and right eye (right side) are taken as the first anterior chamber angles in sequence, and the d1 and d2 values ​​of each first anterior chamber angle are calculated respectively.

[0090] S105: Determine whether d1 / d2 of each first anterior chamber corner is greater than the first threshold T1.

[0091] If so, determine that the first anterior chamber angle is a closed angle, execute S110: end, and output the determination result of the current first anterior chamber angle.

[0092] If not, determine that the first anterior chamber angle is not closed, and continue to execute S106: calculate the first initial angle of the first anterior chamber angle and the second initial angle of the second anterior chamber angle.

[0093] S107: Calculate the mirror similarity between the left-eye UBM image and the right-eye UBM image.

[0094] S108: Based on the second initial angle and mirror similarity, the first initial angle is corrected to obtain the target angle θ of the first room angle.

[0095] S109: Determine whether θ is less than the second threshold T2.

[0096] If so, determine that the first anterior chamber angle is a narrow angle, execute S110: end, and output the determination result of the current first anterior chamber angle.

[0097] If not, determine that the first room angle is an open angle, execute S110: end, and output the determination result of the current first room angle.

[0098] As can be seen from the above embodiments, the ocular angle morphology determination method of this application first acquires the left and right UBM images of the same case eye, and then calls the trained image segmentation model for processing. In each UBM image, the cornea, iris, and lens are segmented to obtain the segmentation boundaries of each object. Then, a corneal reference center point is determined with the corneal segmentation boundary as a reference, and the closest and furthest distances between the first iris and the corneal reference center point in the target direction are determined with the first iris segmentation boundary as a reference. The ratio of the closest to the furthest distance and the first threshold are used to determine whether an ocular angle is closed or not. For closed angles, this method transforms the angle problem into a distance problem for measurement, resulting in more accurate calculations and a simpler, more efficient determination method. When the first anterior chamber angle is not closed, the first anterior chamber angle is corrected based on the second initial angle of the mirrored second anterior chamber angle and the mirror similarity between the left and right UBM images. The relationship between the corrected target angle and a second threshold is then used to determine whether it is a narrow or open angle. Since the final target angle incorporates the effects of mirror and similarity factors, the error between it and the true angle is effectively reduced, and the accuracy of determining whether it is an open or narrow angle is significantly improved. In other words, after obtaining the UBM image pairs, this application does not require manual measurement but automatically determines the shape of each anterior chamber angle based on relevant algorithms, thus improving the accuracy of the determination.

[0099] Based on the methods described in the above embodiments, this embodiment will further describe the method from the perspective of an ocular angle morphology determination device. Please refer to [link to relevant documentation]. Figure 8 The device for determining the shape of the anterior chamber angle may include:

[0100] The acquisition module 10 is used to acquire a pair of UBM images of the same eye in the same case. The pair of UBM images includes a first UBM image and a second UBM image. The first UBM image is one of the left eye UBM image and the right eye UBM image, and the second UBM image is the other one.

[0101] Processing module 20 is used to call the trained image segmentation model to process the UBM image pairs and obtain the first segmentation boundary of the cornea, the second segmentation boundary of the iris and the third segmentation boundary of the lens in each UBM image. In each UBM image, the iris includes a first iris and a second iris. The first iris is located on one side of the left and right sides of the eye, and the second iris is located on the other side.

[0102] The first determination module 30 is used to determine a corneal reference center point based on the first segmentation boundary of the cornea in the first UBM image, and to determine the closest distance and the farthest distance between the first iris and the corneal reference center point in the target direction based on the second segmentation boundary of the first iris in the first UBM image and the corneal reference center point. When the ratio of the closest distance to the farthest distance is greater than a first threshold, the first anterior chamber angle morphology on the side where the first iris is located is determined to be a closed angle. When the ratio is not greater than the first threshold, the first anterior chamber angle morphology is determined to be a non-closed angle.

[0103] The determination module 40 is used to determine the first initial angle of the first anterior chamber angle based on the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image when the first anterior chamber angle is not closed; and to determine the second initial angle of the second anterior chamber angle on the side where the second iris is located in the second UBM image based on the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image, and to obtain the mirror similarity between the first UBM image and the second UBM image.

[0104] The second determination module 50 is used to correct the first initial angle based on the second initial angle and the mirror similarity to obtain the target angle of the first ovarian angle. When the target angle is less than the second threshold, the first ovarian angle shape is determined to be a narrow angle. When the target angle is not less than the second threshold, the first ovarian angle shape is determined to be an open angle.

[0105] In one embodiment, the first determination module 30 includes:

[0106] The first determining submodule is used to determine the first minimum bounding rectangle of the cornea based on the first segmentation boundary of the cornea in the first UBM image, and to determine the center of the first minimum bounding rectangle as the corneal reference center point;

[0107] The second determining submodule is used to determine the second minimum bounding rectangle of the first iris based on the second segmentation boundary of the first iris in the first UBM image, and to determine the closest distance and the farthest distance between the first iris and the corneal reference center point in the target direction based on the left and right bounding rectangles of the second minimum bounding rectangle and the corneal reference center point.

[0108] In one embodiment, the determining module 40 includes:

[0109] The third determining submodule is used to determine the first intersection point of the cornea and the first iris in the first UBM image based on the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image, and to determine the second intersection point of the cornea and the second iris in the second UBM image based on the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image, and to determine the first intersection point and the second intersection point as target points in sequence.

[0110] The first acquisition submodule is used to set i to 1, acquire the i-th radius, generate the i-th reference circle with the target point as the center based on the i-th radius, take the target point as the vertex of the room corner, obtain the i-th candidate room corner according to the first intersection point formed by the i-th reference circle and the first dividing boundary, the second intersection point formed by the i-th reference circle and the second dividing boundary, and the target point, and acquire the i-th candidate angle of the i-th candidate room corner.

[0111] The loop submodule is used to repeatedly execute the (i+1)th radius generation operation, the (i+1)th reference circle generation operation, the (i+1)th candidate angle acquisition operation, and the judgment operation on whether the variance of the first (i+1)th candidate angles is greater than a third threshold. The (i+1)th radius generation operation includes generating the (i+1)th radius based on the ith radius and a preset radius incrementing parameter. The (i+1)th reference circle generation operation includes generating the (i+1)th reference circle with the target point as the center based on the (i+1)th radius. The (i+1)th candidate angle acquisition operation includes obtaining the (i+1)th candidate room corner based on the target point as the corner vertex, the first intersection point formed by the (i+1)th reference circle and the first dividing boundary, the second intersection point formed by the (i+1)th reference circle and the second dividing boundary, and the target point, and obtaining the (i+1)th candidate angle of the (i+1)th candidate room corner. The judgment operation includes judging whether the variance of the first (i+1)th candidate angles is greater than a third threshold.

[0112] The fourth determination submodule is used in each loop to determine the average value of the first i+1 candidate angles as the first initial angle of the first ovarian angle when the target point is the first boundary point, and to determine the average value of the first i+1 candidate angles as the second initial angle of the second ovarian angle when the target point is the second boundary point; if the determination result is not, i is incremented by 1 and the next loop is performed.

[0113] In one embodiment, the ocular anterior chamber angle morphology determination device further includes a first processing module that operates before the first determination module 30, and a second processing module that operates after the first determination module 30.

[0114] The first processing module is used to perform a first binarization on the segmented UBM image pairs so that the area inside each segmentation boundary is a white area and the area outside each segmentation boundary is a black area.

[0115] The second processing module is used to perform a second binarization on the segmented UBM image pairs so that the area inside each segmentation boundary is a black area and the area outside each segmentation boundary is a white area.

[0116] In one embodiment, the determining module 40 includes:

[0117] The processing submodule is used to determine one of the first UBM image and the second UBM image as the original image, and to perform mirroring processing on the other to obtain a mirror image. The original image and the mirror image are then grayscaled and adjusted to a preset width and a preset height.

[0118] The fifth determining submodule is used to determine the preset sampling precision based on the preset width and the preset height, and to determine N original sampling pixels in the original image and N mirror sampling pixels in the mirror image based on the preset sampling precision. Each original pixel sampling point and the corresponding mirror sampling pixel point form a pixel pair.

[0119] The second acquisition submodule is used to acquire the gray values ​​of the original sampled pixels and the mirror sampled pixels in each pixel pair respectively, to obtain N sets of gray value differences, and to obtain the global similarity between the original image and the mirror image based on the N sets of gray value differences.

[0120] The first submodule is used to obtain the mirror similarity between the first UBM image and the second UBM image based on the global similarity, the preset width, and the preset height.

[0121] In one embodiment, the second determination module 50 includes:

[0122] The second submodule is used to obtain the absolute difference between the first initial angle and the second initial angle, and to obtain a correction factor based on the absolute difference and the mirror similarity.

[0123] The sixth determining submodule is used to compare the magnitudes of the first initial angle and the second initial angle, and based on the comparison result, determine a first angle and a second angle among the first initial angle and the second initial angle; the first angle is greater than the second angle.

[0124] The seventh determining submodule is used to determine the difference between the first angle and the correction factor as the target angle of the first room angle when the first initial angle is the first angle.

[0125] The eighth determining submodule is used to determine the sum of the second angle and the correction factor as the target angle of the first room angle when the first initial angle is the second angle.

[0126] In one embodiment, the second determination module 50 includes:

[0127] The third submodule is used to obtain the absolute difference between the first initial angle and the second initial angle, and to obtain a correction factor based on the absolute difference and the mirror similarity.

[0128] The ninth determining submodule is used to determine the sum of the first initial angle and the correction factor as the target angle of the first anterior chamber angle when the first anterior chamber angle is located to the left of the left eye or to the right of the right eye.

[0129] The tenth determining submodule is used to determine the difference between the first initial angle and the correction factor as the target angle of the first anterior chamber angle when the first anterior chamber angle is located to the right of the left eye or to the left of the right eye.

[0130] Unlike existing technologies, the ocular angle morphology determination device provided in this application first acquires UBM images of the left and right eyes of the same case, and then processes them using a trained image segmentation model. In each UBM image, the cornea, iris, and lens are segmented to obtain the segmentation boundaries of each object. Then, a corneal reference center point is determined using the corneal segmentation boundary as a reference, and the closest and furthest distances between the first iris and the corneal reference center point in the target direction are determined using the first iris segmentation boundary as a reference. Finally, the device determines whether an ocular angle is closed or not based on the relationship between the ratio of the closest to the furthest distance and a first threshold. For closed angles, this method transforms the angle problem into a distance problem for measurement, resulting in more accurate calculations and a simpler, more efficient determination method. When the first anterior chamber angle is not closed, the first anterior chamber angle is corrected based on the second initial angle of the mirrored second anterior chamber angle and the mirror similarity between the left and right UBM images. The relationship between the corrected target angle and a second threshold is then used to determine whether it is a narrow or open angle. Since the final target angle incorporates the effects of mirror and similarity factors, the error between it and the true angle is effectively reduced, and the accuracy of determining whether it is an open or narrow angle is significantly improved. In other words, after obtaining the UBM image pairs, this application does not require manual measurement but automatically determines the shape of each anterior chamber angle based on relevant algorithms, thus improving the accuracy of the determination.

[0131] Accordingly, embodiments of this application also provide an electronic device, such as... Figure 9As shown, the electronic device may include a radio frequency (RF) circuit 1001, a memory 1002 including one or more computer-readable storage media, an input unit 1003, a display unit 1004, a sensor 1005, an audio circuit 1006, a WiFi module 1007, a processor 1008 including one or more processing cores, and a power supply 1009, among other components. Those skilled in the art will understand that... Figure 9 The electronic device structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein:

[0132] The radio frequency circuit 1001 can be used for receiving and transmitting signals during information transmission or calls. Specifically, it receives downlink information from the base station and hands it over to one or more processors 1008 for processing; additionally, it transmits uplink data to the base station. The memory 1002 can be used to store software programs and modules. The processor 1008 executes various functional applications and eye angle morphology determination by running the software programs and modules stored in the memory 1002. The input unit 1003 can be used to receive input digital or character information and generate keyboard, mouse, joystick, optical, or trackball signal inputs related to customer settings and function control.

[0133] The display unit 1004 can be used to display information input by the customer or information provided to the customer, as well as various graphical client interfaces of the server. These graphical client interfaces can be composed of graphics, text, icons, videos, and any combination thereof.

[0134] The electronic device may also include at least one sensor 1005, such as a light sensor, a motion sensor, and other sensors. The audio circuitry 1006 includes a speaker that provides an audio interface between the client and the electronic device.

[0135] WiFi is a short-range wireless transmission technology. Electronic devices using the WiFi module 1007 can help customers send and receive emails, browse web pages, and access streaming media, providing customers with wireless broadband internet access. Although Figure 9 WiFi module 1007 is shown, but it is understood that it is not a necessary component of electronic devices and can be omitted as needed without changing the nature of the application.

[0136] The processor 1008 is the control center of the electronic device. It connects various parts of the phone through various interfaces and lines. By running or executing software programs and / or modules stored in the memory 1002, and calling data stored in the memory 1002, it performs various functions of the electronic device and processes data, thereby monitoring the phone as a whole.

[0137] The electronic device also includes a power supply 1009 (such as a battery) that supplies power to various components. Preferably, the power supply can be logically connected to the processor 1008 through a power management system, thereby enabling functions such as managing charging, discharging, and power consumption through the power management system.

[0138] Although not shown, electronic devices may also include cameras, Bluetooth modules, etc., which will not be described in detail here. Specifically, in this embodiment, the processor 1008 in the server loads the executable files corresponding to the processes of one or more applications into the memory 1002 according to the following instructions, and the processor 1008 runs the applications stored in the memory 1002 to achieve the following functions:

[0139] Obtain a pair of UBM images of the same eye, the pair of UBM images including a first UBM image and a second UBM image, the first UBM image being one of the left eye UBM image and the right eye UBM image, and the second UBM image being the other;

[0140] The trained image segmentation model is called to process the UBM image pairs to obtain the first segmentation boundary of the cornea, the second segmentation boundary of the iris, and the third segmentation boundary of the lens in each UBM image. In each UBM image, the iris includes a first iris and a second iris. The first iris is located on one side of the left and right sides of the eye, and the second iris is located on the other side.

[0141] A corneal reference center point is determined based on the first segmentation boundary of the cornea in the first UBM image. The closest distance and the farthest distance between the first iris and the corneal reference center point in the target direction are determined based on the second segmentation boundary of the first iris in the first UBM image and the corneal reference center point. When the ratio of the closest distance to the farthest distance is greater than a first threshold, the first anterior chamber angle on the side where the first iris is located is determined to be a closed angle. When the ratio is not greater than the first threshold, the first anterior chamber angle is determined to be a non-closed angle.

[0142] When the first anterior chamber angle is not closed, the first initial angle of the first anterior chamber angle is determined according to the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image. The second initial angle of the second anterior chamber angle on the side where the second iris is located in the second UBM image is determined according to the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image. The mirror similarity between the first UBM image and the second UBM image is obtained.

[0143] The first initial angle is corrected based on the second initial angle and the mirror similarity to obtain the target angle of the first ovarian angle. When the target angle is less than the second threshold, the first ovarian angle is determined to be a narrow angle. When the target angle is not less than the second threshold, the first ovarian angle is determined to be an open angle.

[0144] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the detailed description above, and they will not be repeated here.

[0145] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0146] Therefore, embodiments of this application provide a computer-readable storage medium storing a plurality of instructions that can be loaded by a processor to perform the following functions:

[0147] Obtain a pair of UBM images of the same eye, the pair of UBM images including a first UBM image and a second UBM image, the first UBM image being one of the left eye UBM image and the right eye UBM image, and the second UBM image being the other;

[0148] The trained image segmentation model is called to process the UBM image pairs to obtain the first segmentation boundary of the cornea, the second segmentation boundary of the iris, and the third segmentation boundary of the lens in each UBM image. In each UBM image, the iris includes a first iris and a second iris. The first iris is located on one side of the left and right sides of the eye, and the second iris is located on the other side.

[0149] A corneal reference center point is determined based on the first segmentation boundary of the cornea in the first UBM image. The closest distance and the farthest distance between the first iris and the corneal reference center point in the target direction are determined based on the second segmentation boundary of the first iris in the first UBM image and the corneal reference center point. When the ratio of the closest distance to the farthest distance is greater than a first threshold, the first anterior chamber angle on the side where the first iris is located is determined to be a closed angle. When the ratio is not greater than the first threshold, the first anterior chamber angle is determined to be a non-closed angle.

[0150] When the first anterior chamber angle is not closed, the first initial angle of the first anterior chamber angle is determined according to the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image. The second initial angle of the second anterior chamber angle on the side where the second iris is located in the second UBM image is determined according to the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image. The mirror similarity between the first UBM image and the second UBM image is obtained.

[0151] The first initial angle is corrected based on the second initial angle and the mirror similarity to obtain the target angle of the first ovarian angle. When the target angle is less than the second threshold, the first ovarian angle is determined to be a narrow angle. When the target angle is not less than the second threshold, the first ovarian angle is determined to be an open angle.

[0152] The foregoing has provided a detailed description of the ocular angle morphology determination method, apparatus, electronic device, and computer-readable storage medium provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the technical solutions and core ideas of this application. Those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for determining the morphology of the anterior chamber angle, characterized in that, include: Obtain a pair of UBM images of the same eye, the pair of UBM images including a first UBM image and a second UBM image, the first UBM image being one of the left eye UBM image and the right eye UBM image, and the second UBM image being the other; The trained image segmentation model is called to process the UBM image pairs to obtain the first segmentation boundary of the cornea, the second segmentation boundary of the iris, and the third segmentation boundary of the lens in each UBM image. In each UBM image, the iris includes a first iris and a second iris. The first iris is located on one side of the left and right sides of the eye, and the second iris is located on the other side. A corneal reference center point is determined based on the first segmentation boundary of the cornea in the first UBM image. The closest distance and the farthest distance between the first iris and the corneal reference center point in the target direction are determined based on the second segmentation boundary of the first iris in the first UBM image and the corneal reference center point. When the ratio of the closest distance to the farthest distance is greater than a first threshold, the first anterior chamber angle on the side where the first iris is located is determined to be a closed angle. When the ratio is not greater than the first threshold, the first anterior chamber angle is determined to be a non-closed angle. When the first anterior chamber angle is not closed, the first initial angle of the first anterior chamber angle is determined according to the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image. The second initial angle of the second anterior chamber angle on the side where the second iris is located in the second UBM image is determined according to the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image. The mirror similarity between the first UBM image and the second UBM image is obtained. The first initial angle is corrected based on the second initial angle and the mirror similarity to obtain the target angle of the first ovarian angle. When the target angle is less than the second threshold, the first ovarian angle is determined to be a narrow angle. When the target angle is not less than the second threshold, the first ovarian angle is determined to be an open angle.

2. The method for determining the morphology of the anterior chamber angle according to claim 1, characterized in that, The steps of determining a corneal reference center point based on a first segmentation boundary of the cornea in the first UBM image, and determining the closest and furthest distances between the first iris and the corneal reference center point in the target direction based on a second segmentation boundary of the first iris in the first UBM image and the corneal reference center point, include: Based on the first segmentation boundary of the cornea in the first UBM image, the first minimum bounding rectangle of the cornea is determined, and the center of the first minimum bounding rectangle is determined as the corneal reference center point; Based on the second segmentation boundary of the first iris in the first UBM image, the second minimum bounding rectangle of the first iris is determined. Based on the left and right borders of the second minimum bounding rectangle and the corneal reference center point, the closest and furthest distances between the first iris and the corneal reference center point in the target direction are determined.

3. The method for determining the morphology of the anterior chamber angle according to claim 1, characterized in that, The steps of determining the first initial angle of the first anterior chamber angle based on the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image, and determining the second initial angle of the second anterior chamber angle on the side where the second iris is located in the second UBM image based on the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image, include: Based on the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image, the first intersection point of the cornea and the first iris in the first UBM image is determined. Based on the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image, the second intersection point of the cornea and the second iris in the second UBM image is determined. The first intersection point and the second intersection point are then determined as target points in sequence. Set i to 1, obtain the i-th radius, and generate the i-th reference circle with the target point as the center based on the i-th radius. Take the target point as the vertex of the room corner, obtain the i-th candidate room corner based on the first intersection point formed by the i-th reference circle and the first dividing boundary, the second intersection point formed by the i-th reference circle and the second dividing boundary, and the target point, and obtain the i-th candidate angle of the i-th candidate room corner. The process iteratively executes the (i+1)th radius generation operation, the (i+1)th reference circle generation operation, the (i+1)th candidate angle acquisition operation, and the judgment operation on whether the variance of the first (i+1)th candidate angles is greater than a third threshold. The (i+1)th radius generation operation includes generating the (i+1)th radius based on the ith radius and a preset radius incrementing parameter. The (i+1)th reference circle generation operation includes generating the (i+1)th reference circle with the target point as the center based on the (i+1)th radius. The (i+1)th candidate angle acquisition operation includes obtaining the (i+1)th candidate room corner based on the target point as the corner vertex, the first intersection point formed by the (i+1)th reference circle and the first dividing boundary, the second intersection point formed by the (i+1)th reference circle and the second dividing boundary, and the target point, and obtaining the (i+1)th candidate angle of the (i+1)th candidate room corner. The judgment operation includes judging whether the variance of the first (i+1)th candidate angles is greater than a third threshold. Within each loop, if the judgment result is yes, when the target point is the first boundary point, the average value of the first i+1 candidate angles is determined as the first initial angle of the first ventricular angle; when the target point is the second boundary point, the average value of the first i+1 candidate angles is determined as the second initial angle of the second ventricular angle. If the judgment result is no, i is incremented by 1 and the next loop is performed.

4. The method for determining the morphology of the anterior chamber angle according to claim 1, characterized in that, Before the step of determining the corneal reference center point based on the first segmentation boundary of the cornea in the first UBM image, the method further includes: The segmented UBM image pairs are binarized for the first time so that the area inside each segmentation boundary is a white area and the area outside each segmentation boundary is a black area. After determining that the first anterior chamber angle is not closed, the following steps are also included: The segmented UBM image pairs are binarized a second time so that the area inside each segmentation boundary is a black area and the area outside each segmentation boundary is a white area.

5. The method for determining the morphology of the anterior chamber angle according to claim 1, characterized in that, The step of obtaining the mirror similarity between the first UBM image and the second UBM image includes: One of the first UBM image and the second UBM image is determined as the original image, and the other is mirrored to obtain a mirror image. The original image and the mirror image are then grayscaled and adjusted to a preset width and a preset height. A preset sampling precision is determined based on the preset width and the preset height. N original sampling pixels in the original image and N mirror sampling pixels in the mirror image are determined based on the preset sampling precision. Each original sampling pixel and the corresponding mirror sampling pixel form a pixel pair. The gray values ​​of the original sampled pixels and the mirror sampled pixels in each pixel pair are obtained respectively, resulting in N sets of gray value differences. The global similarity between the original image and the mirror image is obtained based on the N sets of gray value differences. The mirror similarity between the first UBM image and the second UBM image is obtained based on the global similarity, the preset width, and the preset height.

6. The method for determining the morphology of the anterior chamber angle according to claim 1, characterized in that, The step of correcting the first initial angle based on the second initial angle and the mirror similarity to obtain the target angle of the first room angle includes: Obtain the absolute difference between the first initial angle and the second initial angle, and obtain a correction factor based on the absolute difference and the mirror similarity. Compare the magnitudes of the first initial angle and the second initial angle, and based on the comparison result, determine the first angle and the second angle between the first initial angle and the second initial angle; the first angle is greater than the second angle; When the first initial angle is the first angle, the difference between the first angle and the correction factor is determined as the target angle of the first room angle; When the first initial angle is the second angle, the sum of the second angle and the correction factor is determined as the target angle of the first room angle.

7. The method for determining the morphology of the anterior chamber angle according to claim 1, characterized in that, The step of correcting the first initial angle based on the second initial angle and the mirror similarity to obtain the target angle of the first room angle includes: Obtain the absolute difference between the first initial angle and the second initial angle, and obtain a correction factor based on the absolute difference and the mirror similarity. When the first anterior chamber angle is located to the left of the left eye or to the right of the right eye, the sum of the first initial angle and the correction factor is determined as the target angle of the first anterior chamber angle. When the first anterior chamber angle is located to the right of the left eye or to the left of the right eye, the difference between the first initial angle and the correction factor is determined as the target angle of the first anterior chamber angle.

8. A device for determining the morphology of the anterior chamber angle, characterized in that, include: The acquisition module is used to acquire a pair of UBM images of the same eye in the same case. The pair of UBM images includes a first UBM image and a second UBM image. The first UBM image is one of the left eye UBM image and the right eye UBM image, and the second UBM image is the other one. The processing module is used to call the trained image segmentation model to process the UBM image pairs and obtain the first segmentation boundary of the cornea, the second segmentation boundary of the iris and the third segmentation boundary of the lens in each UBM image. In each UBM image, the iris includes a first iris and a second iris. The first iris is located on one side of the left and right sides of the eye, and the second iris is located on the other side. The first determination module is used to determine a corneal reference center point based on the first segmentation boundary of the cornea in the first UBM image, and to determine the closest distance and the farthest distance between the first iris and the corneal reference center point in the target direction based on the second segmentation boundary of the first iris in the first UBM image and the corneal reference center point. When the ratio of the closest distance to the farthest distance is greater than a first threshold, the first anterior chamber angle morphology on the side where the first iris is located is determined to be a closed angle. When the ratio is not greater than the first threshold, the first anterior chamber angle morphology is determined to be a non-closed angle. The determination module is used to determine the first initial angle of the first anterior chamber angle based on the first segmentation boundary of the cornea and the second segmentation boundary of the first iris in the first UBM image when the first anterior chamber angle is not closed; and to determine the second initial angle of the second anterior chamber angle on the side where the second iris is located in the second UBM image based on the first segmentation boundary of the cornea and the second segmentation boundary of the second iris in the second UBM image; and to obtain the mirror similarity between the first UBM image and the second UBM image. The second determination module is used to correct the first initial angle based on the second initial angle and the mirror similarity to obtain the target angle of the first ovarian angle. When the target angle is less than the second threshold, the first ovarian angle is determined to be a narrow angle. When the target angle is not less than the second threshold, the first ovarian angle is determined to be an open angle.

9. An electronic device, characterized in that, It includes a memory and a processor; the memory stores an application program, and the processor runs the application program in the memory to perform the steps in the ocular angle morphology determination method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which is executed by a processor to implement the steps in the method for determining the anterior chamber angle morphology according to any one of claims 1 to 7.

Images