Image evaluation method and apparatus, storage medium, and electronic device

Abstract

This disclosure discloses an image evaluation method, apparatus, storage medium, and electronic device, relating to image processing technology. The image evaluation method includes: determining a fundus image to be evaluated corresponding to an object to be evaluated; segmenting and extracting data from the fundus image to be evaluated to obtain a lesion segmentation result image, wherein the lesion segmentation result image includes lesion regions; and determining the grade evaluation result corresponding to the fundus image to be evaluated based on the lesion segmentation result image. Based on the lesion segmentation result image, characteristic data of the lesions, such as quantity, area, and location data, can be obtained to determine the grade evaluation result corresponding to the fundus image to be evaluated. Since the characteristic data of the lesions are closely related to the degree of disease—for example, a larger area and a greater number of lesions result in a higher grade evaluation—the image evaluation method provided in this disclosure achieves the determination of the grade evaluation result corresponding to the fundus image to be evaluated in a non-invasive manner, thereby helping doctors understand the fundus condition of the object to be evaluated.

Description

Technical Field

[0001] This disclosure relates to the field of image processing technology, specifically to an image evaluation method and apparatus, a storage medium, and an electronic device. Background Technology

[0002] Because many fundus diseases cannot be detected by the naked eye, fundus examination has become an important tool for doctors to diagnose them. During a fundus examination, doctors can understand many fundus diseases based on fundus images. Fundus images are a crucial component of fundus examinations and a primary means of assisting doctors in diagnosing fundus diseases; their importance is self-evident.

[0003] However, traditional techniques for processing fundus images are insufficient for doctors to understand the condition of related fundus diseases. This usually requires combining these techniques with other examinations, such as blood tests. For some patients with weakened immune systems, blood tests can easily lead to infection, causing their condition to worsen. Therefore, there is an urgent need for a method that allows doctors to understand fundus diseases non-invasively. Summary of the Invention

[0004] In view of this, the present disclosure provides an image evaluation method and apparatus, storage medium and electronic device, to help doctors understand the condition of related fundus diseases under non-invasive conditions.

[0005] In a first aspect, an embodiment of this disclosure provides an image evaluation method, comprising: determining a fundus image to be evaluated corresponding to an object to be evaluated; segmenting and extracting the fundus image to be evaluated to obtain a lesion segmentation result image, wherein the lesion segmentation result image includes a lesion region; and determining a grade evaluation result corresponding to the fundus image to be evaluated based on the lesion segmentation result image.

[0006] In conjunction with the first aspect, in some implementations of the first aspect, before segmenting and extracting the fundus image to be evaluated to obtain the lesion segmentation result image, the method further includes: determining the location data of the basic features of the fundus image to be evaluated, the basic features of the fundus image to be evaluated including blood vessels and the macula; determining the grade evaluation result corresponding to the fundus image to be evaluated based on the lesion segmentation result image, including: determining the location data, area data, and color data of the lesion region; and determining the grade evaluation result corresponding to the fundus image to be evaluated based on the location data of the basic features, combined with at least one of the location data of the lesion region, the area data of the lesion region, and the color data of the lesion region.

[0007] In conjunction with the first aspect, in certain implementations of the first aspect, the grade assessment result corresponding to the fundus image to be evaluated includes severe, moderate, and mild. The grade assessment conditions include a distance threshold of the lesion region relative to the macular center, a lesion region area threshold, a preset color value range of the lesion region, and a distance threshold of the lesion region relative to the blood vessel. Based on the location data of the basic features, and combined with at least one of the location data of the lesion region, the area data of the lesion region, and the color data of the lesion region, the grade assessment result corresponding to the fundus image to be evaluated is determined, including: determining the location data of the macular center and the location data of the blood vessels in the basic features; determining the first distance between the lesion region and the macular center based on the location data of the lesion region and the location data of the macular center; and determining the lesion based on the location data of the lesion region and the location data of the blood vessels. The second distance between the region and the blood vessel; or, if the first distance is less than or equal to the first distance threshold, the grade assessment result corresponding to the fundus image to be evaluated is determined to be severe; if the first distance is greater than the first distance threshold and less than the second distance threshold, and if the area data of the lesion region is greater than the lesion region area threshold and / or the color data of the lesion region meets the range of preset color values ​​of the lesion region, the grade assessment result corresponding to the fundus image to be evaluated is determined to be severe; if the first distance is greater than the first distance threshold and less than the second distance threshold, and if the second distance data is greater than the distance threshold between the lesion region and the blood vessel, the grade assessment result corresponding to the fundus image to be evaluated is determined to be moderate; if the first distance is greater than the second distance threshold, the grade assessment result corresponding to the fundus image to be evaluated is determined to be mild.

[0008] In conjunction with the first aspect, in certain implementations of the first aspect, determining the fundus image to be evaluated corresponding to the object to be evaluated includes: determining the effective region fundus image corresponding to the object to be evaluated; classifying the effective region fundus image using a retinitis classification model, and determining the classification result of the effective region fundus image, the classification result including whether the effective region fundus image is a fundus image of cytomegalovirus retinitis or not a fundus image of cytomegalovirus retinitis; if the effective region fundus image is a fundus image of cytomegalovirus retinitis, determining the effective region fundus image as the fundus image to be evaluated.

[0009] In conjunction with the first aspect, in some implementations of the first aspect, before classifying the original fundus image to be evaluated using the classification model of retinitis, the method further includes: determining the classification results corresponding to the retinal fundus image samples; establishing an initial network model, and training the initial network model using the classification results corresponding to the retinal fundus image samples to generate a classification model of retinitis.

[0010] In conjunction with the first aspect, in certain implementations of the first aspect, determining the effective region fundus image corresponding to the object to be evaluated includes: determining the original fundus image corresponding to the object to be evaluated; extracting color channels from the original fundus image to determine the single-channel image corresponding to the original fundus image; performing threshold segmentation processing on the single-channel image corresponding to the original fundus image to obtain the binarized image corresponding to the original fundus image; extracting fundus features based on the binarized image corresponding to the original fundus image to obtain the fundus feature region; and performing roundness fitting on the fundus feature region to determine the effective region fundus image corresponding to the object to be evaluated.

[0011] In conjunction with the first aspect, in some implementations of the first aspect, segmenting and extracting the fundus image to be evaluated to obtain a lesion segmentation result image includes: extracting color channels from the fundus image to be evaluated to determine the single-channel image corresponding to the fundus image to be evaluated; performing binarization processing on the single-channel image to obtain the binarized image corresponding to the single-channel image; and segmenting the fundus image to be evaluated based on the binarized image to obtain a lesion segmentation result image.

[0012] Secondly, one embodiment of this disclosure provides an image evaluation device, comprising: a determining module for determining a fundus image to be evaluated corresponding to an object to be evaluated; a segmentation module for segmenting and extracting the fundus image to be evaluated to obtain a lesion segmentation result image, wherein the lesion segmentation result image includes a lesion region; and an evaluation module for determining a grade evaluation result corresponding to the fundus image to be evaluated based on the lesion segmentation result image.

[0013] Thirdly, an embodiment of this disclosure provides an electronic device comprising: a processor; and a memory for storing processor-executable instructions, wherein the processor is configured to perform the method mentioned in the first aspect above.

[0014] Fourthly, one embodiment of this disclosure provides a computer-readable storage medium storing a computer program for performing the methods mentioned in the first aspect above.

[0015] This disclosure embodiment obtains a lesion segmentation result image by segmenting and extracting the image of the eye to be evaluated. The lesion segmentation result image includes the lesion region. Based on the lesion segmentation result image, characteristic data of the lesions can be obtained, such as quantity, area, and location data, to determine the grade assessment result corresponding to the fundus image to be evaluated. Since the characteristic data of the lesions are closely related to the severity of the disease—for example, a larger area and more lesions indicate a higher degree of disease and a higher grade assessment—or, the location of the lesions is related to the type of lesions, the type of lesions can be determined based on the lesion location data, and different types of lesions have different effects on the degree of disease, resulting in different grade assessment results. Therefore, the image evaluation method provided by this disclosure embodiment can determine the grade assessment result corresponding to the fundus image to be evaluated in a non-invasive manner, thereby helping doctors understand the fundus condition of the subject being evaluated. Attached Figure Description

[0016] The above and other objects, features, and advantages of this disclosure will become more apparent from the more detailed description of the embodiments thereof in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this disclosure and form part of the specification. They are used together with the embodiments of this disclosure to explain the disclosure and do not constitute a limitation thereof.

[0017] Figure 1 The diagram shown is an application scenario illustration provided by an embodiment of this disclosure.

[0018] Figure 2 The diagram shown is a flowchart of an image evaluation method provided in an embodiment of this disclosure.

[0019] Figure 3 The diagram shown is a flowchart illustrating another image evaluation method provided in another embodiment of this disclosure.

[0020] Figure 4 The diagram shown is a flowchart illustrating the process of determining the fundus image corresponding to the object to be evaluated, according to an embodiment of this disclosure.

[0021] Figure 5 The diagram shown is a flowchart of an image evaluation method provided in another embodiment of this disclosure.

[0022] Figure 6 The diagram shown is a flowchart illustrating the process of determining the effective region of the fundus image corresponding to the object to be evaluated, according to an embodiment of this disclosure.

[0023] Figure 7 The diagram shown is a flowchart illustrating the process of segmenting and extracting a fundus image to be evaluated to obtain a lesion segmentation result image, according to an embodiment of this disclosure.

[0024] Figure 8 The diagram shown is a structural schematic of an image evaluation apparatus provided in an embodiment of this disclosure.

[0025] Figure 9 The diagram shown is a structural schematic of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0026] The technical solutions of the present disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present disclosure, and not all embodiments.

[0027] The fundus is composed of the retina, optic nerve head, optic nerve fibers, and the choroid behind the retina. Because the retina, optic nerve head, optic nerve fibers, and the choroid behind the retina are all located at the bottom of the posterior segment of the eyeball, many fundus diseases cannot be detected by naked eye. Therefore, fundus examination is usually required to assist doctors in diagnosis. Fundus examination is an important auxiliary means for doctors to diagnose fundus diseases.

[0028] During a fundus examination, doctors can understand many fundus diseases based on fundus images, such as retinal detachment, optic neuritis, or macular degeneration. Fundus images are an important component of fundus examinations and a primary means of assisting doctors in diagnosing fundus diseases; their importance is self-evident.

[0029] However, traditional techniques for analyzing fundus images are insufficient for doctors to understand the condition of related fundus diseases. To assess the progression and severity of these diseases, doctors typically need to combine these methods with other examinations, such as blood tests. However, for some immunocompromised patients, especially those with cytomegalovirus retinitis (CMV), blood tests can easily lead to infection and worsen their condition. This is because CMV is a common opportunistic ocular infection in patients with Human Immunodeficiency Virus (HIV) infection / Acquired Immunodeficiency Syndrome (AIDS). These patients already have weakened immune systems, and blood tests can cause skin damage, which can hinder wound healing and lead to infection, potentially exacerbating the condition. Therefore, there is an urgent need for a non-invasive method to help doctors understand the condition of the fundus.

[0030] The following is combined Figure 1 A brief introduction will be given to an application scenario of one embodiment of this disclosure.

[0031] Figure 1 The diagram illustrates an application scenario of one embodiment of this disclosure. Figure 1As shown, this scenario is a fundus image evaluation scenario for patient A. Specifically, the fundus image evaluation scenario for patient A includes a server 110, a user terminal 120 and a data storage device 130, which are respectively communicatively connected to the server 110. The server 110 is used to execute the methods mentioned in the embodiments of this disclosure.

[0032] For example, in practical application, a user sends a command to the server via user terminal 120 to evaluate the fundus image of patient A (the subject to be evaluated). Upon receiving the command, server 110 retrieves the fundus image corresponding to patient A from data storage device 130, segments and extracts the fundus image to obtain a lesion segmentation result image, which includes the lesion region and the macular center. Based on the lesion segmentation result image, the grade evaluation result corresponding to the fundus image to be evaluated is determined. Server 110 then sends the grade evaluation result corresponding to the fundus image to be evaluated to user terminal 120 so that user terminal 120 can display the evaluation result to the user.

[0033] For example, the data stored in the aforementioned data storage device 130 includes, but is not limited to, all image data saved by the medical institution during the treatment of patient A, relevant fundus image data of patient A input by the user, and fundus image data taken by patient A as needed during the consultation.

[0034] For example, the user terminal 120 mentioned above includes, but is not limited to, computer terminals such as desktop computers and laptops, and mobile terminals such as tablet computers and mobile phones.

[0035] Figure 2 The diagram shown is a flowchart illustrating an image evaluation method provided in an embodiment of this disclosure. Figure 2 As shown, the image evaluation method provided in this embodiment includes the following steps.

[0036] Step S210: Determine the fundus image corresponding to the object to be evaluated.

[0037] For example, the subject of evaluation can be a patient with cytomegalovirus retinitis or other types of retinitis. The fundus image to be evaluated for the subject is a fundus image of a patient with cytomegalovirus retinitis or a fundus image of a patient with other types of retinitis.

[0038] Step S220: Segment and extract the fundus image to be evaluated to obtain the lesion segmentation result image.

[0039] The resulting image of the lesion segmentation includes the lesion region.

[0040] For example, a trained deep learning network segmentation model can be used to segment the lesion, resulting in a segmented image. Alternatively, traditional image segmentation methods, such as thresholding or edge detection, can be used to segment the lesion, resulting in a segmented image.

[0041] Step S230: Based on the lesion segmentation result image, determine the grade assessment result corresponding to the fundus image to be evaluated.

[0042] For example, based on the lesion region in the lesion segmentation result image, characteristic data of the lesion are obtained, such as number, area, and location data. Based on the characteristic data of the lesion, the grade assessment result corresponding to the fundus image to be evaluated is determined; or based on other features of the lesion image, such as the distance of the lesion from the center of the macula, the grade assessment result corresponding to the fundus image to be evaluated is determined. For example, the larger the area and / or the more numerous the lesions, the greater the degree of disease may be, and the higher the grade assessment. Alternatively, the type of lesion is determined based on the lesion location data. Since the lesion location is related to the lesion type, different types of lesions correspond to different degrees of disease, and the grade assessment results are also different.

[0043] The evaluation method provided in this disclosure involves segmenting and extracting the fundus image to be evaluated to obtain a lesion segmentation result image, wherein the lesion segmentation image includes the lesion region. Then, based on the lesion segmentation result image, the grade evaluation result corresponding to the fundus image to be evaluated is determined. Based on the lesion segmentation result image, characteristic data of the lesions can be obtained, such as quantity, area, and location data. Based on the characteristic data of the lesions, the grade evaluation result corresponding to the fundus image to be evaluated is determined. Since the characteristic data of the lesions are closely related to the degree of disease, for example, the larger the area and the more numerous the lesions, the greater the degree of disease and the higher the grade evaluation. Alternatively, the location of the lesions is related to the type of lesions. The type of lesions is determined based on the location data of the lesions. Different types of lesions have different effects on the degree of disease, and the grade evaluation results are also different. Therefore, this disclosure embodiment achieves the determination of the grade evaluation result corresponding to the fundus image to be evaluated in a non-invasive manner, thereby helping doctors understand the fundus condition of the subject being evaluated.

[0044] Figure 3 The diagram shown is a flowchart illustrating another image evaluation method provided in another embodiment of this disclosure. Figure 2 Extending from the illustrated embodiment Figure 3 The illustrated embodiment will be described in detail below. Figure 3 The illustrated embodiments and Figure 2 The differences between the embodiments shown are not repeated here, and the similarities are not repeated here.

[0045] like Figure 3As shown, another embodiment of the image evaluation method provided in this disclosure further includes the following steps before segmenting and extracting the fundus image to be evaluated to obtain the lesion segmentation result image.

[0046] Step S310: Determine the location data of the basic features of the fundus image to be evaluated.

[0047] The basic features of the fundus images to be evaluated include blood vessels and the macula.

[0048] For example, the location data of blood vessels and the location data of the macula in the fundus image to be evaluated are determined according to a custom coordinate system. The location data of the macula can be determined by selecting the coordinates of the center of the macula. The custom coordinate system can be determined according to the requirements, with the center of the optic disc or the center of the macula as the origin, or with the line connecting the center of the optic disc and the center of the macula as the X-axis and the perpendicular line as the Y-axis, or with the midpoint of the line connecting the center of the optic disc and the center of the macula as the origin, or other coordinate systems determined according to the requirements.

[0049] Step S320: Calculate the area of ​​the lesion region to determine the area data of the lesion region.

[0050] For example, the area data of the lesion region is determined by calculating the number of pixels occupied by the area of ​​the lesion region from the image of the lesion region.

[0051] In some embodiments, the specific implementation of step S230 above includes: determining the location data, area data, and color data of the lesion region; and determining the grade assessment result corresponding to the fundus image to be evaluated based on the location data of the basic features, combined with at least one of the location data, area data, and color data of the lesion region. For example, according to the characteristics of different fundus diseases, grade assessment conditions are selected to determine the assessment result corresponding to the fundus image to be evaluated. For instance, according to the lesion characteristics of giant cell retinitis, the location of the lesion relative to the macula center is selected as the grade assessment condition, that is, the grade assessment result is determined based on the location data of the lesion region, or the grade assessment result is determined based on the location data of the lesion region combined with the area data and / or color data of the lesion region. Alternatively, according to the lesion characteristics of giant cell retinitis, the size of the lesion area can be selected as the grade assessment condition, that is, the grade assessment result is determined based on the area data of the lesion, or the grade assessment result is determined based on the area data of the lesion region combined with the location data and / or color data of the lesion region. Alternatively, the type of lesion can be selected as the grading criterion based on the characteristics of the lesions in giant cell retinitis. Lesions in giant cell retinitis include: yellowish-white necrotic retinal lesions, retinal hemorrhage, retinal vascular sheath, and granular retinal opacities. The type of lesion can be determined based on the color data of the lesion, and the grading result can be determined based on the type of lesion. Alternatively, the grading result can be determined based on the color data of the lesion combined with the location data and / or the area data of the lesion region.

[0052] In some embodiments, the grade assessment result corresponding to the fundus image to be evaluated includes severe, moderate, and mild. The grade assessment conditions include a distance threshold of the lesion area relative to the macular center, a lesion area threshold, a preset color value range of the lesion area, and a distance threshold of the lesion area relative to the blood vessel. A specific implementation method for determining the grade assessment result corresponding to the fundus image to be evaluated based on the location data of the basic features, combined with at least one of the location data of the lesion area, the area data of the lesion area, and the color data of the lesion area, includes: determining the location data of the macular center and the location data of the blood vessels in the basic features; determining a first distance between the lesion area and the macular center based on the location data of the lesion area and the location data of the macular center; and determining the distance between the lesion area and the blood vessel based on the location data of the lesion area and the preset color value range of the blood vessel. The second distance to the blood vessel is specified; or, if the first distance is less than or equal to the first distance threshold, the grade assessment result corresponding to the fundus image to be evaluated is determined to be severe; if the first distance is greater than the first distance threshold and less than the second distance threshold, and if the area data of the lesion region is greater than the lesion region area threshold and / or the color data of the lesion region meets the range of preset color values ​​of the lesion region, the grade assessment result corresponding to the fundus image to be evaluated is determined to be severe; if the first distance is greater than the first distance threshold and less than the second distance threshold, and if the second distance data is greater than the lesion region relative to the blood vessel distance threshold, the grade assessment result corresponding to the fundus image to be evaluated is determined to be moderate; if the first distance is greater than the second distance threshold, the grade assessment result corresponding to the fundus image to be evaluated is determined to be mild.

[0053] For example, the distance threshold between the lesion area and the center of the macula can be the optic disc's long axis diameter, or other distance values ​​selected according to actual needs. The area threshold of the lesion area can be set according to actual needs, and this embodiment does not specifically limit the value of the area threshold. For example, the preset color value range of the lesion area can be a color value range set according to needs.

[0054] For example, when the fundus image to be evaluated is a fundus image of cytomegalovirus retinitis, a first distance between the lesion area and the macula is determined based on the location data of the macular center and the lesion area in the basic features; a second distance between the lesion area and the location data of the blood vessels is determined based on the location data of the blood vessels and the lesion area. When the first distance is less than or equal to a first distance threshold, that is, when the distance between the lesion area and the macula is less than or equal to a preset threshold, the lesion has a greater impact on vision, and the patient's disease is more severe. Therefore, the grade assessment result corresponding to the fundus image to be evaluated can be determined as severe. It should be understood that the first distance threshold can be selected according to needs, for example, the distance from the macula center is one optic disc long axis diameter. This embodiment of the disclosure does not specifically limit the value of the first threshold.

[0055] For example, when the first distance is greater than a first distance threshold and less than a second distance threshold—that is, when the distance from the lesion region to the macula center is greater than the first threshold and less than the second threshold—the factors affecting vision mainly include the type and area of ​​the lesion. Therefore, when the area of ​​the lesion region is greater than the total area of ​​the lesion region, the lesion region is larger, which has a greater impact on the patient's vision and indicates a more severe condition. Thus, the grade assessment result corresponding to the fundus image being evaluated can be determined as severe.

[0056] Alternatively, when the color data of the lesion area meets a preset color value range, such as RGB(255, 255, 255) for white and RGB(255, 255, 0) for yellow, the color data indicates that the lesion area appears whitish or yellowish. This confirms that the lesion is a yellowish-white necrotic retinal lesion. In cytomegalovirus retinitis, yellowish-white necrotic retinal lesions have a significant impact on the patient's vision, indicating a severe condition. Therefore, the grade assessment result corresponding to the fundus image being evaluated can be determined as severe.

[0057] Alternatively, if the area of ​​the lesion exceeds a threshold and the color of the lesion meets a preset color value, indicating a yellowish-white necrotic retina, a large lesion area, significant impact on the patient's vision, and a severe condition, then the assessment result of the fundus image being evaluated can be determined to be severe.

[0058] For example, when the fundus image to be evaluated is that of a patient with cytomegalovirus retinitis, if the first distance is greater than a first distance threshold and less than a second distance threshold, and if the second distance is greater than the distance threshold between the lesion area and the blood vessel (i.e., the lesion area is far from the blood vessel), it can be determined that the lesion type is a retinal hemorrhage lesion. Retinal hemorrhage lesions have a smaller impact on the patient's vision compared to yellowish-white necrotic retinal lesions, but a larger impact compared to other types of lesions such as retinal vascular sheath lesions and granular retinal opacities. Therefore, the grade assessment result corresponding to the fundus image to be evaluated can be determined as moderate.

[0059] For example, when the fundus image to be evaluated is that of a patient with cytomegalovirus retinitis, if the first distance is greater than a second distance threshold, the assessment result of the fundus image is determined to be mild. Specifically, when the first distance is greater than the second threshold, the impact on the patient's vision is reduced because the lesion is farther from the macula's center, thus the assessment result of the fundus image can be determined to be mild. In the mild case, different lesion types correspond to different degrees of disease, with the severity from worst being: yellowish-white necrotic retinal lesions, retinal hemorrhage lesions, retinal vascular sheath lesions, and granular retinal opacities. The type of lesion can be determined based on the first distance and the second threshold, as well as the distance thresholds corresponding to the first and second thresholds respectively. Therefore, when the assessment result is mild, the severity within the mild condition can be determined in conjunction with the lesion type, indicating the extent of the patient's disease, to assist the physician in determining the patient's treatment plan.

[0060] In some embodiments, if the area of ​​the lesion region is greater than an area threshold, the area threshold can be selected as needed, within a range that has a relatively small impact on the severity of the disease. The average grayscale value of the fundus image to be evaluated and the grayscale value of the lesion region are calculated. When the grayscale value of the lesion region is greater than the average grayscale value of the fundus image to be evaluated, and the distance of the lesion region relative to the blood vessel is greater than a preset distance between the lesion region and the blood vessel, it can be determined that the lesion region has a large area, a large grayscale value, and is far from the blood vessel. Based on the lesion characteristics of cytomegalovirus retinitis, it can be identified as a yellowish-white necrotic retinal lesion, and the grade assessment result corresponding to the fundus image to be evaluated can be determined as severe. Alternatively, when the area of ​​the lesion region is greater than the area threshold, the grayscale value of the lesion region is less than the average grayscale value of the fundus image to be evaluated, and the color data meets a preset color value range, such as red, it can be determined that the lesion region has a large area, a small grayscale value, and a reddish color. Based on the lesion characteristics of cytomegalovirus retinitis, it can be identified as a retinal hemorrhage lesion, and the grade assessment result corresponding to the fundus image to be evaluated can be determined as moderate.

[0061] In some embodiments, based on the morphological data of the lesion area, combined with other data of the lesion, such as area data, location data, and color data, the grade assessment result corresponding to the fundus image to be evaluated is determined. The morphological data can reflect the morphology and / or shape of the lesion, and based on the morphological data, the type of lesion can be determined. For example, the characteristics of retinal vascular sheath lesions include: distribution along the blood vessel boundary, linear distribution on both sides of the blood vessel, elongated shape, and yellowish color. Based on the morphological data of the lesion area combined with location data and / or color data, it can be determined whether the lesion is a retinal vascular sheath lesion. If it is a retinal vascular sheath lesion, since the severity of retinal vascular sheath lesions is milder than that of yellowish-white necrotic retinal lesions and retinal hemorrhage lesions under the same conditions, the grade assessment result corresponding to the retinal vascular sheath lesion can be mild within a preset distance range. The preset distance range is selected according to needs, for example, a range greater than a second threshold. It should be understood that within a preset distance range, when the assessment result corresponds to mild, severity can also be distinguished within mild. For example, retinal sheath lesions are considered a more severe case within the mild category, while granular retinal opacities are milder than retinal sheath lesions and can be classified as a milder case within the mild category. The characteristics of granular retinal opacities include: scattered distribution in the fundus image being assessed, small individual opacities, clear granular borders, and a whitish-yellow color. Assuming that location data confirms a scattered distribution of lesions, area data confirms a small individual lesion area, morphological data confirms a granular lesion shape, and color data confirms a whitish-yellow color, the lesion type can be identified as granular retinal opacities. Therefore, the assessment result is determined to be mild, falling within the milder category of mild cases.

[0062] This embodiment of the disclosure determines the grade assessment result corresponding to the fundus image to be evaluated by using the location data of the lesion area relative to the macula center, the location data of the lesion area relative to the macula center, the preset color value range of the lesion area, and the distance threshold of the lesion area relative to the blood vessel. The grade assessment result is determined according to the degree of impact on the patient's vision caused by the lesion, which can obtain a more accurate grade assessment result. In addition, this embodiment of the disclosure classifies the assessment grade by using the location data of the basic features, combined with at least one of the location data of the lesion area, the area data of the lesion area, and the color data of the lesion area, so as to determine the grade assessment result corresponding to the fundus image to be evaluated. Since the assessment result is determined by combining the location data, area data, and color data of the lesion area, the accuracy of the grade assessment result is further improved. This embodiment of the disclosure can also combine the gray value data and morphological data of the lesion area to determine the grade assessment result, which can obtain a more accurate grade assessment result, so as to assist doctors in determining relevant treatment plans based on the accurate grade assessment result.

[0063] Figure 4 The diagram shown is a flowchart illustrating the process of determining the fundus image corresponding to the object to be evaluated, according to an embodiment of this disclosure. Figure 4 As shown in the embodiments of this disclosure, the steps for determining the fundus image corresponding to the object to be evaluated include the following.

[0064] Step S410: Determine the effective region fundus image corresponding to the object to be evaluated.

[0065] For example, the region of interest (ROI) in the fundus image corresponding to the object to be evaluated is determined, and the ROI is determined as the effective region for the object to be evaluated.

[0066] Step S420: Using a classification model for retinitis, classify the fundus images of the effective region and determine the classification results of the fundus images of the effective region.

[0067] The classification results include fundus images of the effective area that are cytomegalovirus retinitis or fundus images of the effective area that are not cytomegalovirus retinitis.

[0068] For example, the classification model for retinitis can classify fundus images of cytomegalovirus retinitis and non-cytomegalovirus retinitis. That is, it can determine whether the fundus image of the effective region is a fundus image of cytomegalovirus retinitis or not. It can also determine the type of fundus image of cytomegalovirus retinitis. That is, when the fundus image of the effective region is a fundus image of cytomegalovirus retinitis, it can determine whether the fundus image of the effective region is a fundus image of fulminant (edematous) cytomegalovirus retinitis or a fundus image of indolent (granular) cytomegalovirus retinitis.

[0069] Step S430: Determine whether the fundus image of the effective area is a fundus image of cytomegalovirus retinitis.

[0070] For example, if the fundus image of the effective region is a fundus image of cytomegalovirus retinitis, then step S440 is performed; if the fundus image of the effective region is not a fundus image of cytomegalovirus retinitis, then step S450 is performed.

[0071] Step S440: The fundus image of the effective region is identified as the fundus image to be evaluated.

[0072] For example, through the above classification, the fundus image to be evaluated is either a fundus image of fulminant (edematous) cytomegalovirus retinitis or a fundus image of indolent (granular) cytomegalovirus retinitis. After determining that it is a fundus image to be evaluated, it can output whether the fundus image of cytomegalovirus retinitis is a fundus image of fulminant (edematous) cytomegalovirus retinitis or a fundus image of indolent (granular) cytomegalovirus retinitis.

[0073] Step S450: End the image evaluation task.

[0074] This embodiment utilizes a classification model for retinitis to determine whether the fundus image of the effective region is a fundus image of cytomegalovirus retinitis. If so, the fundus image of cytomegalovirus retinitis is used as the subsequent image to be evaluated. This allows for the evaluation of fundus images of cytomegalovirus retinitis, reducing computational load. Furthermore, by using a classification method, the subsequent grade evaluation results can be made more accurate.

[0075] Figure 5 The diagram shown is a flowchart illustrating an image evaluation method provided in another embodiment of this disclosure. Figure 4 Extending from the illustrated embodiment Figure 5 The illustrated embodiment will be described in detail below. Figure 5 The illustrated embodiments and Figure 4 The differences between the embodiments shown are not repeated here, and the similarities are not repeated here.

[0076] Figure 5 As shown, another embodiment of this disclosure provides another image evaluation method, which further includes the following steps before classifying the original fundus image to be evaluated using a classification model of retinitis.

[0077] Step S510: Determine the classification results corresponding to the retinal fundus image samples and the retinal fundus image samples.

[0078] For example, the retinal fundus image samples can be fundus images of fulminant (edematous) cytomegalovirus retinitis, fundus images of lazy (granular) cytomegalovirus retinitis, and fundus images of non-cytomegalovirus retinitis.

[0079] Step S520: Establish an initial network model and train the initial network model using retinal fundus image samples and the classification results corresponding to the retinal fundus image samples to generate a classification model for retinitis.

[0080] For example, the initial network model can be a deep learning classification model, such as the ResNet 50 model or other deep learning network classification models.

[0081] This embodiment of the disclosure trains an initial network model using retinal fundus image samples and the corresponding classification results to generate a classification model for retinitis, making the classification results more accurate. Furthermore, the retinal fundus image samples include fundus images of fulminant (edematous) cytomegalovirus retinitis and fundus images of indolent (granular) cytomegalovirus retinitis, achieving the goal of classifying these two types of images using the classification model. This helps doctors determine the type of cytomegalovirus retinitis based on the classification results, further enhancing their understanding of the fundus condition.

[0082] Figure 6 The diagram shown is a flowchart illustrating a process for determining the effective region of the fundus image corresponding to the object to be evaluated, according to an embodiment of this disclosure. Figure 6 As shown in the embodiments of this disclosure, the determination of the effective region fundus image corresponding to the object to be evaluated includes the following steps.

[0083] Step S610: Determine the original fundus image corresponding to the object to be evaluated.

[0084] Step S620: Extract color channels from the original fundus image to determine the single-channel image corresponding to the original fundus image.

[0085] For example, the red channel is extracted from the original fundus image to determine the single-channel image corresponding to the original fundus image. Alternatively, the three channels of the original fundus image are extracted by averaging to obtain a grayscale image.

[0086] Step S630: Based on the single-channel image corresponding to the original fundus image, threshold segmentation is performed to obtain the binarized image corresponding to the original fundus image.

[0087] For example, based on the single-channel image corresponding to the original fundus image, according to actual needs, such as the average value of the red channel grayscale or by calculation to determine the optimal threshold, values ​​greater than this threshold are set to 255, and values ​​less than this threshold are set to 0, thus obtaining the binarized image corresponding to the original fundus image.

[0088] Step S640: Based on the binarized image corresponding to the original fundus image, perform feature extraction of fundus features to obtain the fundus feature region.

[0089] Step S650: Perform roundness fitting on the fundus feature region to determine the effective fundus image corresponding to the object to be evaluated.

[0090] For example, the effective fundus image corresponding to the object to be evaluated is determined by fitting the circularity of the fundus feature region to the circumscribed circle of the fundus feature region.

[0091] This embodiment processes fundus images by performing operations such as color channel separation and binarization, making the contours of fundus image features clearer, which improves the subsequent lesion area segmentation and extraction, and ultimately results in a more accurate grade assessment.

[0092] Figure 7 The diagram shown is a flowchart illustrating the process of segmenting and extracting a fundus image to be evaluated, and obtaining a lesion segmentation result image, according to an embodiment of this disclosure. Figure 7 The following steps are shown in the embodiment of this disclosure for segmenting and extracting the fundus image to be evaluated to obtain the lesion segmentation result image.

[0093] Step S710: Extract color channels from the fundus image to be evaluated to determine the single-channel image corresponding to the fundus image to be evaluated.

[0094] For example, based on actual needs, a color channel is selected to determine the single-channel image corresponding to the fundus image to be evaluated.

[0095] Step S720: Binarize the single-channel image to obtain the binarized image corresponding to the single-channel image.

[0096] Step S730: Based on the binarized image, segment the fundus image to be evaluated to obtain the lesion segmentation result image.

[0097] For example, the lesion segmentation result image can be combined with the original fundus image, and the color data of the final lesion area image is obtained from the original fundus image so that the lesion segmentation result image can be processed subsequently.

[0098] This embodiment processes the lesion area by performing operations such as color channel separation and binarization, making the outline of the lesion area clear. The segmented image of the fundus to be evaluated is then segmented, resulting in a more accurate lesion segmentation image, which further improves the accuracy of the grade assessment results.

[0099] The above text combined Figures 2 to 7 The present disclosure describes in detail the method embodiments, which are then combined with the following. Figure 8 and Figure 9 The present disclosure provides a detailed description of the apparatus embodiments. Furthermore, it should be understood that the descriptions of the method embodiments correspond to the descriptions of the apparatus embodiments; therefore, any parts not described in detail can be found in the foregoing method embodiments.

[0100] Figure 8 The diagram shown is a structural schematic of an image evaluation apparatus provided in an embodiment of this disclosure.

[0101] like Figure 8As shown, the image evaluation device provided in this embodiment includes: a determination module 801, a segmentation module 802, and an evaluation module 803. Specifically, the determination module 801 is used to determine the fundus image to be evaluated corresponding to the object to be evaluated; the segmentation module 802 is used to segment and extract the fundus image to be evaluated to obtain a lesion segmentation result image, wherein the lesion segmentation result image includes a lesion region; the evaluation module 803 determines the grade evaluation result corresponding to the fundus image to be evaluated based on the lesion segmentation result image.

[0102] In some embodiments, the evaluation module 803 is further configured to: determine the location data of basic features of the fundus image to be evaluated, the basic features including blood vessels and the macula; calculate the area of ​​the lesion region to determine the area data of the lesion region; and determine the grade evaluation result corresponding to the fundus image to be evaluated based on the lesion segmentation result image, including: determining the location data, area data, and color data of the lesion region; and determine the grade evaluation result corresponding to the fundus image to be evaluated based on the location data of the basic features, combined with at least one of the location data of the lesion region, the area data of the lesion region, and the color data of the lesion region.

[0103] In some embodiments, the grade assessment result corresponding to the fundus image to be evaluated includes severe, moderate, and mild, and the grade assessment conditions include a distance threshold of the lesion area relative to the macular center, a lesion area threshold, a preset color value range of the lesion area, and a distance threshold of the lesion area relative to the blood vessel. The assessment module 803 is further configured to determine the grade assessment result corresponding to the fundus image to be evaluated based on the location data of the basic features, combined with at least one of the location data of the lesion area, the area data of the lesion area, and the color data of the lesion area, including: determining the location data of the macular center and the location data of the blood vessels in the basic features; determining a first distance between the lesion area and the macular center based on the location data of the lesion area and the location data of the macular center; and determining the lesion area based on the location data of the lesion area and the location data of the blood vessels. The second distance between the lesion and the blood vessel; or, if the first distance is less than or equal to the first distance threshold, the grade assessment result corresponding to the fundus image to be evaluated is determined to be severe; if the first distance is greater than the first distance threshold and less than the second distance threshold, and if the area data of the lesion region is greater than the lesion region area threshold and / or the color data of the lesion region meets the range of preset color values ​​of the lesion region, the grade assessment result corresponding to the fundus image to be evaluated is determined to be severe; if the first distance is greater than the first distance threshold and less than the second distance threshold, and if the second distance data is greater than the distance threshold between the lesion region and the blood vessel, the grade assessment result corresponding to the fundus image to be evaluated is determined to be moderate; if the first distance is greater than the second distance threshold, the grade assessment result corresponding to the fundus image to be evaluated is determined to be mild.

[0104] In some embodiments, the determining module 801 is further configured to: determine the effective region fundus image corresponding to the object to be evaluated; classify the effective region fundus image using a classification model for retinitis, and determine the classification result of the effective region fundus image, wherein the classification result includes whether the effective region fundus image is a fundus image of cytomegalovirus retinitis or whether the effective region fundus image is not a fundus image of cytomegalovirus retinitis; if the effective region fundus image is a fundus image of cytomegalovirus retinitis, determine the effective region fundus image as the fundus image to be evaluated.

[0105] In some embodiments, the determining module 801 is further configured to determine the classification results corresponding to the retinal fundus image samples and the retinal fundus image samples; establish an initial network model and train the initial network model using the classification results corresponding to the retinal fundus image samples and the retinal fundus image samples to generate a classification model for retinitis.

[0106] In some embodiments, the determining module 801 is further configured to: determine the original fundus image corresponding to the object to be evaluated; extract color channels from the original fundus image to determine the single-channel image corresponding to the original fundus image; perform threshold segmentation processing based on the single-channel image corresponding to the original fundus image to obtain a binarized image corresponding to the original fundus image; extract fundus features based on the binarized image corresponding to the original fundus image to obtain a fundus feature region; and perform roundness fitting on the fundus feature region to determine the effective region fundus image corresponding to the object to be evaluated.

[0107] In some embodiments, the determining module 801 is further configured to: extract color channels from the fundus image to be evaluated to determine the single-channel image corresponding to the fundus image to be evaluated; perform binarization processing on the single-channel image to obtain the binarized image corresponding to the single-channel image; and segment the fundus image to be evaluated based on the binarized image to obtain the lesion segmentation result image.

[0108] Figure 9 The diagram shown is a schematic representation of the structure of an electronic device provided in an exemplary embodiment of this disclosure. The electronic device 900 (specifically, it may be a computer device) includes a memory 901, a processor 902, a communication interface 903, and a bus 904. The memory 901, processor 902, and communication interface 903 are interconnected via the bus 904.

[0109] The memory 901 may be a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 15301 may store a program, and when the program stored in the memory 901 is executed by the processor 902, the processor 902 and the communication interface 903 are used to perform the various steps in the image evaluation method of the embodiments of this disclosure.

[0110] The processor 902 may be a general-purpose central processing unit (CPU), microprocessor, application-specific integrated circuit (ASIC), graphics processing unit (GPU), or one or more integrated circuits, used to execute relevant programs to achieve the functions required by each unit in the image evaluation method of this disclosure embodiment.

[0111] The processor 902 can also be an integrated circuit chip with signal processing capabilities. In implementation, each step of the image evaluation method of this disclosure can be completed by the integrated logic circuits in the hardware of the processor 902 or by instructions in software form. The processor 902 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of this disclosure can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in the memory 901. The processor 902 reads the information in the memory 901 and, in conjunction with its hardware, performs the functions required by the units included in the image evaluation apparatus of this disclosure, or executes the image evaluation method of this disclosure.

[0112] The communication interface 903 uses a transceiver device, such as, but not limited to, a transceiver, to enable communication between the electronic device 900 and other devices or communication networks. For example, the communication interface 903 can be used to acquire fundus images corresponding to the object to be evaluated.

[0113] Bus 904 may include a pathway for transmitting information between various components of electronic device 900 (e.g., memory 901, processor 902, communication interface 903).

[0114] It should be noted that, although Figure 9 The illustrated electronic device 900 only shows the memory, processor, and communication interface. However, those skilled in the art should understand that in specific implementations, the electronic device 900 may also include other devices necessary for normal operation. Furthermore, depending on specific needs, those skilled in the art should understand that the electronic device 900 may also include hardware devices for implementing other additional functions. Moreover, those skilled in the art should understand that the electronic device 900 may only include the devices necessary for implementing the embodiments of this disclosure, and may not necessarily include... Figure 9 All the devices shown.

[0115] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this disclosure.

[0116] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0117] In the embodiments provided in this disclosure, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0118] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0119] In addition, the functional units in the various embodiments of this disclosure can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0120] Embodiments of this disclosure can also be computer-readable storage media storing computer program instructions that, when executed by a processor, cause the processor to perform the steps of the methods described above according to various embodiments of this disclosure. If the functionality is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks. The computer-readable storage medium can be any combination of one or more readable media. A readable medium can be a readable signal medium or a readable storage medium. A readable storage medium can be, for example, including but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof.

[0121] The above description is merely a specific embodiment of this disclosure, but the scope of protection of this disclosure is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this disclosure should be included within the scope of protection of this disclosure. Therefore, the scope of protection of this disclosure should be determined by the scope of the claims.

Claims

1. An image evaluation method, characterized in that, include: Identify the fundus image corresponding to the object to be evaluated; The fundus image to be evaluated is segmented and extracted to obtain a lesion segmentation result image, wherein the lesion segmentation result image includes the lesion region; Determine the location data of the basic features of the fundus image to be evaluated, the basic features of the fundus image to be evaluated including blood vessels and the macula; Determine the location, area, and color data of the lesion region; Based on the location data of the basic features, and combined with at least one of the location data of the lesion area, the area data of the lesion area, and the color data of the lesion area, the grade assessment result corresponding to the fundus image to be evaluated is determined; The grade assessment results corresponding to the fundus images to be evaluated include severe, moderate and mild. The grade assessment conditions include the distance threshold of the lesion area relative to the center of the macula, the area threshold of the lesion area, the preset color value range of the lesion area and the distance threshold of the lesion area relative to the blood vessel. The step of determining the grade assessment result corresponding to the fundus image to be evaluated based on the location data of the basic features, combined with at least one of the location data of the lesion region, the area data of the lesion region, and the color data of the lesion region, includes: Determine the location data of the macular center and the location data of blood vessels in the aforementioned basic features; Based on the location data of the lesion area and the location data of the macular center, a first distance between the lesion area and the macular center is determined; Based on the location data of the lesion area and the location data of the blood vessel, a second distance between the lesion area and the blood vessel is determined; If the first distance is less than or equal to the first distance threshold, the grade assessment result corresponding to the fundus image to be evaluated is determined to be severe; or, if the first distance is greater than the first distance threshold and less than the second distance threshold, and if the area data of the lesion region is greater than the lesion region area threshold and / or the color data of the lesion region meets the range of the preset color value of the lesion region, the grade assessment result corresponding to the fundus image to be evaluated is determined to be severe. If the first distance is greater than a first distance threshold and less than a second distance threshold, and if the second distance is greater than the distance threshold between the lesion area and the blood vessel, the grade assessment result corresponding to the fundus image to be evaluated is determined to be moderate. If the first distance is greater than the second distance threshold, the grade assessment result corresponding to the fundus image to be evaluated is determined to be mild.

2. The method of claim 1, wherein, The process of determining the fundus image corresponding to the object to be evaluated includes: Determine the effective region fundus image corresponding to the object to be evaluated; Using a classification model for retinitis, the fundus images of the effective region are classified to determine the classification result of the fundus images of the effective region. The classification result includes whether the fundus images of the effective region are fundus images of cytomegalovirus retinitis or fundus images of the effective region are not fundus images of cytomegalovirus retinitis. If the fundus image of the effective region is a fundus image of cytomegalovirus retinitis, the fundus image of the effective region is identified as the fundus image to be evaluated.

3. The method of claim 2, wherein, Before classifying the original fundus images to be evaluated using the retinitis classification model, the method further includes: Determine the classification results of the retinal fundus image samples and the corresponding retinal fundus image samples; An initial network model is established, and the initial network model is trained using retinal fundus image samples and the classification results corresponding to the retinal fundus image samples to generate a classification model for retinitis.

4. The method according to claim 2, characterized in that, Determining the effective region fundus image corresponding to the object to be evaluated includes: Determine the original fundus image corresponding to the object to be evaluated; Color channels are extracted from the original fundus image to determine the single-channel image corresponding to the original fundus image; Based on the single-channel image corresponding to the original fundus image, threshold segmentation is performed to obtain the binarized image corresponding to the original fundus image. Based on the binarized image corresponding to the original fundus image, feature extraction of fundus features is performed to obtain the fundus feature region; The fundus feature region is fitted with a circularity to determine the effective fundus image corresponding to the object to be evaluated.

5. The method of claim 1, wherein, The step of segmenting and extracting the fundus image to be evaluated to obtain the lesion segmentation result image includes: Color channels are extracted from the fundus image to be evaluated to determine the single-channel image corresponding to the fundus image to be evaluated. The single-channel image is binarized to obtain the corresponding binarized image. Based on the binarized image, the fundus image to be evaluated is segmented to obtain the lesion segmentation result image.

6. An image evaluation apparatus, characterized by include: The determination module is used to determine the fundus image corresponding to the object to be evaluated; The segmentation module is used to segment and extract the fundus image to be evaluated to obtain a lesion segmentation result image, wherein the lesion segmentation result image includes the lesion region; The evaluation module is used to determine the location data of the basic features of the fundus image to be evaluated, which include blood vessels and the macula; determine the location data, area data, and color data of the lesion area; and determine the grade evaluation result corresponding to the fundus image to be evaluated based on the location data of the basic features, combined with at least one of the location data of the lesion area, the area data of the lesion area, and the color data of the lesion area. The grade assessment results corresponding to the fundus images to be evaluated include severe, moderate and mild. The grade assessment conditions include the distance threshold of the lesion area relative to the center of the macula, the area threshold of the lesion area, the preset color value range of the lesion area and the distance threshold of the lesion area relative to the blood vessel. The step of determining the grade assessment result corresponding to the fundus image to be evaluated based on the location data of the basic features, combined with at least one of the location data of the lesion region, the area data of the lesion region, and the color data of the lesion region, includes: Determine the location data of the macular center and the location data of blood vessels in the aforementioned basic features; Based on the location data of the lesion area and the location data of the macular center, a first distance between the lesion area and the macular center is determined; Based on the location data of the lesion area and the location data of the blood vessel, a second distance between the lesion area and the blood vessel is determined; If the first distance is less than or equal to the first distance threshold, the grade assessment result corresponding to the fundus image to be evaluated is determined to be severe; or, if the first distance is greater than the first distance threshold and less than the second distance threshold, and if the area data of the lesion region is greater than the lesion region area threshold and / or the color data of the lesion region meets the range of the preset color value of the lesion region, the grade assessment result corresponding to the fundus image to be evaluated is determined to be severe. If the first distance is greater than a first distance threshold and less than a second distance threshold, and if the second distance is greater than the distance threshold between the lesion area and the blood vessel, the grade assessment result corresponding to the fundus image to be evaluated is determined to be moderate. If the first distance is greater than the second distance threshold, the grade assessment result corresponding to the fundus image to be evaluated is determined to be mild.

7. An electronic device, comprising: include: processor; Memory used to store the processor's executable instructions. The processor is used to execute the method described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The storage medium stores a computer program for performing the method described in any one of claims 1 to 5.

Images