Diabetic retinopathy lesion visualization method and device and readable medium

Abstract

The invention discloses a diabetic retinopathy lesion visualization method and device and a readable medium, and the method comprises the steps: extracting a region of interest from an original fundus image, and generating a region of interest image; determining a feature stroke code of a feature segmentation region in the region-of-interest image in the original fundus image, and determining a focus stroke code of a diabetic retinopathy focus segmentation region in the region-of-interest image in the original fundus image; based on the feature run-length codes, a reference area used for assisting in judging the diabetic retinopathy focus is obtained, and corresponding reference run-length codes are generated; and displaying the feature stroke code, the focus stroke code and the reference stroke code in a blank image differently by using different colors to generate a fundus focus visual image. Therefore, the original fundus image is simplified, and the simplified image is visually displayed, so that a doctor and a patient can check the image conveniently, the doctor-patient communication time is saved, and the utilization rate of medical resources is improved.

Description

Technical Field

[0001] This invention belongs to the field of image processing technology, and in particular relates to a method, device and computer-readable medium for visualizing diabetic retinopathy lesions. Background Technology

[0002] The fundus is the only image in the human body that allows for simple, non-invasive, and clear observation of changes in capillaries. According to incomplete statistics, more than 60 kinds of fundus diseases or systemic chronic diseases will cause corresponding lesions in the fundus. Therefore, fundus images are usually a clinical diagnostic reference for these diseases.

[0003] In current technologies, fundus reports for patients with diabetic retinopathy cannot visually present the characteristics of the lesions to doctors and patients. Especially for inexperienced doctors, some tiny lesions are invisible to the naked eye and may be missed during diagnosis, thus delaying treatment. Even in fundus images of patients with severe diabetic retinopathy, although the lesions are visible to the naked eye, patients without medical knowledge cannot distinguish between normal and abnormal fundus features. Summary of the Invention

[0004] To address the aforementioned problems in existing technologies, embodiments of the present invention provide a method, apparatus, and computer-readable medium for visualizing diabetic retinopathy lesions. This method can visualize simplified raw fundus images, making them understandable to both doctors and patients, thus saving time in doctor-patient communication and improving the utilization rate of medical resources.

[0005] According to a first aspect of the present invention, a method for visualizing diabetic retinopathy lesions is provided. The method includes: extracting a region of interest from an original fundus image to generate a region of interest image; determining the feature run-length encoding of a segmented region in the region of interest image in the original fundus image, and determining the lesion run-length encoding of a segmented region of diabetic retinopathy lesions in the region of interest image in the original fundus image; obtaining a reference region for assisting in the judgment of diabetic retinopathy lesions based on the feature run-length encoding, and generating a corresponding reference run-length encoding; and displaying the feature run-length encoding, the lesion run-length encoding, and the reference run-length encoding in a blank image using different colors to generate a fundus lesion visualization image.

[0006] Optionally, the feature segmentation region includes at least an optic disc segmentation region, a macular segmentation region, and an arteriovenous vessel segmentation region; determining the feature run-length encoding of the feature segmentation region in the region of interest image in the original fundus image includes: extracting the optic disc, macula, and arteriovenous vessels from the region of interest image to generate corresponding optic disc segmentation regions, macular segmentation regions, and arteriovenous vessel segmentation regions; determining a first run-length encoding of the optic disc in the original fundus image based on the position information of the optic disc segmentation region in the region of interest image; determining a second run-length encoding of the macula in the original fundus image based on the position information of the macular segmentation region in the region of interest image; determining a third run-length encoding of the arteriovenous vessels in the original fundus image based on the position information of the arteriovenous vessel segmentation region in the region of interest image; and determining the first run-length encoding, the second run-length encoding, and the third run-length encoding as the feature run-length encoding of the feature segmentation region in the original fundus image.

[0007] Optionally, the segmented region of the diabetic retinopathy lesion includes a hemorrhage segmentation region, and / or a microaneurysm segmentation region, and / or a soft exudate segmentation region, and / or a hard exudate segmentation region; determining the lesion path encoding of the segmented region of the diabetic retinopathy lesion in the original fundus image in the region of interest image includes: extracting vascular hemorrhage, and / or microaneurysm, and / or soft exudate, and / or hard exudate from the region of interest image respectively, generating corresponding hemorrhage segmentation regions, and / or microaneurysm segmentation regions, and / or soft exudate segmentation regions, and / or hard exudate segmentation regions; based on the location information of the hemorrhage segmentation region in the region of interest image, determining the fourth segmentation region of the vascular hemorrhage in the original fundus image. Run-length encoding; and / or, based on the location information of the microaneurysm segmentation region in the region of interest image, determine the fifth run-length encoding of the microaneurysm in the original fundus image; and / or, based on the location information of the soft exudate segmentation region in the region of interest image, determine the sixth run-length encoding of the soft exudate in the original fundus image; and / or, based on the location information of the hard exudate segmentation region in the region of interest image, determine the seventh run-length encoding of the hard exudate in the original fundus image; and determine the fourth run-length encoding, and / or the fifth run-length encoding, and / or the sixth run-length encoding, and / or the seventh run-length encoding as the lesion run-length encoding of the diabetic retinopathy lesion segmentation region in the original fundus image.

[0008] Optionally, obtaining a reference region for assisting in the determination of diabetic retinopathy lesions based on the feature travel coding, and generating a corresponding reference travel coding, includes: determining a crosshair centered on the optic disc based on a first travel coding of the optic disc in the original fundus image and a first preset distance dataset, and generating a corresponding crosshair travel coding; and determining the crosshair travel coding as the reference travel coding.

[0009] Optionally, obtaining a reference region for assisting in the determination of diabetic retinopathy lesions based on the feature travel coding, and generating a corresponding reference travel code, includes: determining a first circle centered on the macula based on a second travel code of the macula in the original fundus image and a first preset radius, and generating a corresponding first circle travel code; determining a second circle centered on the macula based on a second travel code of the macula in the original fundus image and a second preset radius, and generating a corresponding second circle travel code; determining a third circle centered on the macula based on a second travel code of the macula in the original fundus image and a third preset radius, and generating a corresponding third circle travel code; wherein the first preset radius is smaller than the second preset radius, and the second preset radius is smaller than the third preset radius; and determining the first circle travel code, the second circle travel code, and the third circle travel code as reference travel codes.

[0010] Optionally, the method further includes: statistically analyzing the types of diabetic retinopathy lesions in the visualized fundus lesion image to obtain statistical results; if the statistical results indicate that only microaneurysm segmentation areas exist in the visualized fundus lesion image, the analysis result of the fundus lesion is output as mild non-proliferative diabetic retinopathy; if the statistical results indicate that hemorrhage segmentation areas exist in the visualized fundus lesion image and the number of hemorrhage segmentation areas is not greater than a preset threshold, the analysis result of the fundus lesion is output as moderate non-proliferative diabetic retinopathy; if the statistical results indicate that hard exudate segmentation areas exist in the visualized fundus lesion image and the number of hemorrhage segmentation areas in each of the four quadrants divided by the crosshairs is greater than a preset threshold, the analysis result of the fundus lesion is output as severe non-proliferative diabetic retinopathy.

[0011] Optionally, if the statistical results indicate that there is a hard exudate segmentation region in the visualized image of the fundus lesion and the hard exudate segmentation region is located within a first circle, then the output analysis result of the fundus lesion is diabetic retinopathy with severe macular edema; if the statistical results indicate that there is a hard exudate segmentation region in the visualized image of the fundus lesion and the hard exudate segmentation region is located within a second circle, then the output analysis result of the fundus lesion is diabetic retinopathy with moderate macular edema; if the statistical results indicate that there is a hard exudate segmentation region in the visualized image of the fundus lesion and the hard exudate segmentation region is located within a third circle, then the output analysis result of the fundus lesion is diabetic retinopathy with mild macular edema.

[0012] Optionally, determining the feature run-length encoding of the feature segmentation region in the region of interest image in the original fundus image includes: extracting the optic disc, macula, and blood vessels from the region of interest image to generate corresponding optic disc segmentation regions, macula segmentation regions, and blood vessel segmentation regions; extracting the optic cup from the optic disc segmentation region to generate an optic cup segmentation region; extracting arteries and veins from the blood vessel segmentation region to generate arteriovenous vessel segmentation regions; determining the first run-length encoding of the optic disc in the original fundus image based on the position information of the optic disc segmentation region in the region of interest image; and determining the first run-length encoding of the macula in the original fundus image based on the position information of the macula segmentation region in the region of interest image. The second stroke code in the base image; the third stroke code of the blood vessel in the original fundus image based on the position information of the blood vessel segmentation region in the region of interest image; the eighth stroke code of the optic cup in the original fundus image based on the position information of the optic cup segmentation region in the optic disc segmentation region; the ninth stroke code of the arteriovenous blood vessel in the original fundus image based on the position information of the arteriovenous blood vessel segmentation region in the blood vessel segmentation region; the first stroke code, the second stroke code, the third stroke code, the eighth stroke code, and the ninth stroke code are determined as the feature stroke codes of the feature segmentation region in the original fundus image.

[0013] According to a second aspect of the present invention, a visualization device for diabetic retinopathy lesions is also provided. The device includes: an extraction module for extracting a region of interest from an original fundus image and generating a region of interest image; a determination module for determining the feature run-length encoding of a segmented region in the region of interest image in the original fundus image, and determining the lesion run-length encoding of a segmented region of diabetic retinopathy lesions in the region of interest image in the original fundus image; a generation module for obtaining a reference region for assisting in the judgment of diabetic retinopathy lesions based on the feature run-length encoding, and generating a corresponding reference run-length encoding; and a visualization module for displaying the feature run-length encoding, the lesion run-length encoding, and the reference run-length encoding in a blank image using different colors to generate a visualization image of fundus lesions.

[0014] According to a third aspect of the present invention, an electronic device is also provided, the electronic device comprising: a processor; a memory for storing executable instructions of the processor; the processor being configured to read the executable instructions from the memory and execute the instructions to implement the method as described in the first aspect.

[0015] According to a fourth aspect of the present invention, a computer-readable medium is also provided, on which a computer program is stored, wherein the program, when executed by a processor, implements the method described in the first aspect.

[0016] This invention provides a method for generating a fundus follow-up report. The method includes: first, extracting a region of interest (ROI) from an original fundus image to generate a ROI image; second, determining the feature run-length encoding of a segmented region in the ROI image within the original fundus image, and determining the lesion run-length encoding of a segmented region of diabetic retinopathy lesions in the ROI image within the original fundus image; then, based on the feature run-length encoding, obtaining a reference region for assisting in the identification of diabetic retinopathy lesions, and generating a corresponding reference run-length encoding; finally, displaying the feature run-length encoding, lesion run-length encoding, and reference run-length encoding in a blank image using different colors to generate a fundus lesion visualization image. In this embodiment, a segmented region and a segmented region of diabetic retinopathy lesions are extracted from the ROI of the original fundus image, and the feature run-length encoding, the lesion run-length encoding, and the reference run-length encoding are determined. Then, the feature run-length encoding, lesion run-length encoding, and reference run-length encoding are visualized using different colors to generate a fundus lesion visualization image. Therefore, the original fundus images were simplified and then visualized, making it easier for doctors and patients to view them, thus saving time in doctor-patient communication and improving the utilization rate of medical resources. Attached Figure Description

[0017] The following sections will describe some specific embodiments of the invention in detail by way of example and not limitation, with reference to the accompanying drawings. The same reference numerals in the drawings denote the same or similar parts or portions. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings: Figure 1 This is a flowchart illustrating a method for visualizing diabetic retinopathy lesions according to an embodiment of the present invention. Figure 2 This is a schematic diagram of fundus images according to another embodiment of the present invention; wherein, Figure a is a mild non-proliferative diabetic retinopathy; Figure b is a visualized image of the fundus lesion corresponding to Figure a; Figure 3 This is a schematic diagram of severe nonproliferative diabetic retinopathy in another embodiment of the present invention; Figure 4 for Figure 3 A schematic diagram of the corresponding visualized fundus lesions; Figure 5 for Figure 4 Analysis results of visual images of lesions in the middle fundus.

[0018] Figure 6 This is a schematic diagram of the structure of a device for visualizing diabetic retinopathy lesions provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention 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 invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0020] like Figure 1 The diagram shown is a flowchart illustrating a method for visualizing diabetic retinopathy lesions according to an embodiment of the present invention.

[0021] A method for visualizing diabetic retinopathy lesions, the method comprising at least the following steps: S101, Extract the region of interest from the original fundus image and generate a region of interest image; S102, determine the feature run-length encoding of the feature segmentation region in the region of interest image in the original fundus image, and determine the lesion run-length encoding of the segmentation region of diabetic retinopathy in the region of interest image in the original fundus image; S103, Based on feature travel coding, obtain the reference region for assisting in the identification of diabetic retinopathy lesions and generate the corresponding reference travel coding; S104. The feature travel code, lesion travel code, and reference travel code are displayed in the blank image using different colors to generate a visual image of fundus lesions.

[0022] In S101, a region of interest is extracted from the original fundus image based on a neural network algorithm, and the extracted region of interest is cropped from the original fundus image to obtain a region of interest image.

[0023] In S102, the feature segmentation regions include, but are not limited to, optic disc segmentation regions, macular segmentation regions, and arteriovenous vessel segmentation regions. The optic disc is extracted from the region of interest (ROI) image using a neural network algorithm to generate the optic disc segmentation region; the macula is extracted from the ROI image using a neural network algorithm to generate the macular segmentation region; and arteries and veins are extracted from the ROI image using a neural network algorithm to generate the arteriovenous vessel segmentation region. The position information of the optic disc segmentation region in the ROI image is obtained, and a first stroke code for the optic disc in the original fundus image is determined based on this position information. The position information of the macular segmentation region in the ROI image is obtained, and a second stroke code for the macula in the original fundus image is determined based on this position information. The position information of the arteriovenous vessel segmentation region in the ROI image is obtained, and a third stroke code for the arteriovenous vessels in the original fundus image is determined based on this position information. The first stroke code, the second stroke code, and the third stroke code are determined as the feature stroke codes of the feature segmentation regions in the original fundus image.

[0024] The segmentation regions for diabetic retinopathy lesions include hemorrhage segmentation regions, and / or microaneurysm segmentation regions, and / or soft exudate segmentation regions, and / or hard exudate segmentation regions. Hemorrhage is extracted from the region of interest image using a neural network algorithm to generate a hemorrhage segmentation region; microaneurysms are extracted from the region of interest image using a neural network algorithm to generate a microaneurysm segmentation region; hard exudates are extracted from the region of interest image using a neural network algorithm to generate a hard exudate segmentation region; and soft exudates are extracted from the region of interest image using a neural network algorithm to generate a soft exudate segmentation region. The location information of the hemorrhage segmentation region in the region of interest image is obtained, and based on the location information of the hemorrhage segmentation region, a fourth-stroke code for vascular hemorrhage in the original fundus image is determined; and / or, the location information of the microaneurysm segmentation region in the region of interest image is obtained, and based on the location information of the microaneurysm segmentation region, a fifth-stroke code for the microaneurysm in the original fundus image is determined; and / or, the location information of the soft exudate segmentation region in the region of interest image is obtained, and based on the location information of the soft exudate segmentation region, a sixth-stroke code for the soft exudate in the original fundus image is determined; and / or, the location information of the hard exudate segmentation region in the region of interest image is obtained, and based on the location information of the hard exudate segmentation region, a seventh-stroke code for the hard exudate in the original fundus image is determined. The fourth-stroke code, and / or the fifth-stroke code, and / or the sixth-stroke code, and / or the seventh-stroke code are determined as the lesion stroke codes of the diabetic retinopathy lesion segmentation region in the original fundus image.

[0025] The generation process for the first through seventh line-of-sight codes is similar and will not be listed individually. A detailed explanation will be provided using the generation of the seventh line-of-sight code as an example, as follows: The positional information of the hard exudate segmentation region in the region of interest image is obtained; and the positional information of the region of interest image in the original fundus image is obtained; based on the positional information of the hard exudate segmentation region and the positional information of the region of interest image, the relative positional information of the hard exudate in the original fundus image is determined, and a seventh run-length code is generated.

[0026] It should be noted that run-length encoding is used to indicate the horizontal and vertical coordinate values ​​corresponding to image pixels. For example, seventh run-length encoding is used to indicate the horizontal and vertical coordinate values ​​corresponding to all pixels in the hard exudation region.

[0027] This embodiment extracts feature segmentation regions and diabetic retinopathy lesion segmentation regions from the original fundus image. Then, based on the relative position information of the extracted feature segmentation regions and diabetic retinopathy lesion segmentation regions in the original fundus image, the corresponding run-length encoding is determined. Thus, by denoising the original fundus image, the original fundus image can be simplified, making it easier for patients or doctors to review. This reduces diagnostic omissions and delays in treatment caused by doctors not being able to capture lesion information with the naked eye, thereby improving the user experience.

[0028] In S103, a reference area is used to assist in the identification of diabetic retinopathy lesions, such as a cross line drawn with the optic disc as the center. This cross line can divide the visual image of fundus lesions into four quadrants to assist in the identification of diabetic retinopathy lesions.

[0029] For example, based on the feature travel coding, a reference region for assisting in the determination of diabetic retinopathy lesions is obtained, and a corresponding reference travel coding is generated; including: based on the first travel coding of the optic disc in the original fundus image and the first preset distance dataset, determining a crosshair centered on the optic disc, and generating a corresponding crosshair travel coding; and determining the crosshair travel coding as the reference travel coding.

[0030] In S104, the first stroke code, the second stroke code, the third stroke code, the fourth stroke code, the fifth stroke code, the sixth stroke code, the seventh stroke code, and the reference stroke code are displayed visually in different colors to generate a visual image of fundus lesions.

[0031] This embodiment extracts and processes the original fundus image to generate feature travel codes, diabetic retinopathy lesion travel codes, and reference travel codes. Then, the various codes are displayed in different colors in a blank image and a detection report is output. This makes it easier for doctors and patients to quickly understand fundus images, saves time in doctor-patient communication, and improves the utilization rate of medical resources.

[0032] It should be noted that the segmented areas of diabetic retinopathy lesions may be of only one type or may include multiple categories. For example, the extracted segmented areas of diabetic retinopathy lesions may only include microaneurysm segmented areas; or, the extracted segmented areas of diabetic retinopathy lesions may include hemorrhage segmented areas, microaneurysm segmented areas, soft exudate segmented areas, and hard exudate segmented areas, etc.

[0033] In a preferred embodiment of this invention, the method further includes: determining a first circle centered on the macula based on a second stroke code of the macula in the original fundus image and a first preset radius, and generating a corresponding first circle stroke code; determining a second circle centered on the macula based on a second stroke code of the macula in the original fundus image and a second preset radius, and generating a corresponding second circle stroke code; determining a third circle centered on the macula based on a second stroke code of the macula in the original fundus image and a third preset radius, and generating a corresponding third circle stroke code; wherein the first preset radius is smaller than the second preset radius, and the second preset radius is smaller than the third preset radius; and determining the first circle stroke code, the second circle stroke code, and the third circle stroke code as reference stroke codes.

[0034] Specifically, if hard exudates are present in the extracted diabetic retinopathy lesions, the target subject typically has both diabetic retinopathy and macular edema. Macular edema is a complication of diabetic retinopathy. A target subject may have diabetic retinopathy but not necessarily macular edema; however, the presence of macular edema in a target subject definitely indicates diabetic retinopathy. The first, second, and third circles are used to determine the distance between the hard exudates in the diabetic retinopathy lesions and the macula; the closer the hard exudates are to the macula, the more severe the macular edema, a complication of diabetic retinopathy.

[0035] This embodiment uses the drawing of a first circle, a second circle, and a third circle in a visualized image of fundus lesions to identify complications of diabetic retinopathy in the target subject. This verifies the accuracy of the diagnosis of diabetic retinopathy lesions, helps doctors understand the progression of the patient's condition, facilitates patient treatment, avoids delaying the best time for treatment, and improves the patient's experience.

[0036] In another preferred embodiment of this example, the method further includes: statistically analyzing the types of diabetic retinopathy lesions in the visualized fundus lesion image to obtain statistical results; if the statistical results indicate that only microaneurysm segmentation areas exist in the visualized fundus lesion image, the analysis result of the fundus lesion is output as mild non-proliferative diabetic retinopathy; if the statistical results indicate that hemorrhage segmentation areas exist in the visualized fundus lesion image and the number of hemorrhage segmentation areas is not greater than a preset threshold, the analysis result of the fundus lesion is output as moderate non-proliferative diabetic retinopathy; if the statistical results indicate that hard exudate segmentation areas exist in the visualized fundus lesion image and the number of hemorrhage segmentation areas in each of the four quadrants divided by the crosshairs is greater than a preset threshold, the analysis result of the fundus lesion is output as severe non-proliferative diabetic retinopathy.

[0037] Here, typically, if there is a hemorrhage segmentation area in the visualized image of a fundus lesion, it may be accompanied by a microaneurysm segmentation area; if there is a hard exudate segmentation area in the visualized image of a fundus lesion, it may be accompanied by both a hemorrhage segmentation area and a microaneurysm segmentation area.

[0038] This embodiment combines the quadrants divided by the crosshairs with diabetic retinopathy lesions to determine the severity of diabetic retinopathy, thereby outputting the analysis results of diabetic retinopathy. This not only makes it convenient for doctors and patients to review, but also avoids misdiagnosis or omissions due to doctors' lack of experience or the inability to detect minute lesion features with the naked eye. It saves time in doctor-patient communication and improves the utilization rate of medical resources.

[0039] In a preferred embodiment of this example, the method further includes: statistically analyzing the types of diabetic retinopathy lesions in the visualized fundus lesion image to obtain statistical results; if the statistical results indicate that there is a hard exudate segmentation region in the visualized fundus lesion image and the hard exudate segmentation region is located within a first circle, then the analysis result of the fundus lesion is output as diabetic retinopathy with severe macular edema; if the statistical results indicate that there is a hard exudate segmentation region in the visualized fundus lesion image and the hard exudate segmentation region is located within a second circle, then the analysis result of the fundus lesion is output as diabetic retinopathy with moderate macular edema; if the statistical results indicate that there is a hard exudate segmentation region in the visualized fundus lesion image and the hard exudate segmentation region is located within a third circle, then the analysis result of the fundus lesion is output as diabetic retinopathy with mild macular edema.

[0040] This embodiment combines circles formed by different radii with diabetic retinopathy lesions to determine the severity of diabetic retinopathy complications. This not only makes it convenient for doctors to directly review the data, but also avoids misdiagnosis due to insufficient doctor experience, which could delay treatment, save time in doctor-patient communication, and improve the utilization rate of medical resources.

[0041] The visualization method for diabetic retinopathy lesions provided in this embodiment will be described in detail below, taking into account specific application scenarios.

[0042] A method for visualizing diabetic retinopathy lesions includes at least the following steps: S1, Extract the region of interest from the original fundus image and generate a region of interest image; S2, extract the optic disc, macula, and arteries and veins from the region of interest image to generate corresponding optic disc segmentation regions, macula segmentation regions, and arteriovenous segmentation regions; based on the position information of the optic disc segmentation region in the region of interest image, determine the first stroke code of the optic disc in the original fundus image; based on the position information of the macula segmentation region in the region of interest image, determine the second stroke code of the macula in the original fundus image; based on the position information of the arteriovenous segmentation region in the region of interest image, determine the third stroke code of the arteries and veins in the original fundus image; and determine the first stroke code, the second stroke code, and the third stroke code as the feature stroke code of the feature segmentation region in the original fundus image.

[0043] S3, extract vascular hemorrhage, microaneurysms, soft exudates, and hard exudates from the region of interest image, respectively, and generate corresponding hemorrhage segmentation regions, microaneurysm segmentation regions, soft exudate segmentation regions, and hard exudate segmentation regions; based on the position information of the hemorrhage segmentation region in the region of interest image, determine the fourth stroke code of vascular hemorrhage in the original fundus image; based on the position information of the microaneurysm segmentation region in the region of interest image, determine the fifth stroke code of microaneurysm in the original fundus image; based on the position information of the soft exudate segmentation region in the region of interest image, determine the sixth stroke code of soft exudate in the original fundus image; based on the position information of the hard exudate segmentation region in the region of interest image, determine the seventh stroke code of hard exudate in the original fundus image; determine the fourth stroke code, and / or the fifth stroke code, and / or the sixth stroke code, and / or the seventh stroke code as the lesion stroke code of the diabetic retinopathy lesion segmentation region in the original fundus image.

[0044] S4. Based on the first stroke code of the optic disc in the original fundus image and the first preset distance dataset, a crosshair centered on the optic disc is determined, and a corresponding crosshair stroke code is generated. Based on the second stroke code of the macula in the original fundus image and the first preset radius, a first circle centered on the macula is determined, and a corresponding first circle stroke code is generated. Based on the second stroke code of the macula in the original fundus image and the second preset radius, a second circle centered on the macula is determined, and a corresponding second circle stroke code is generated. Based on the second stroke code of the macula in the original fundus image and the third preset radius, a third circle centered on the macula is determined, and a corresponding third circle stroke code is generated. Wherein, the first preset radius is smaller than the second preset radius, and the second preset radius is smaller than the third preset radius. The crosshair stroke code, the first circle stroke code, the second circle stroke code, and the third circle stroke code are determined as reference stroke codes.

[0045] S5, the first stroke code, second stroke code, third stroke code, fourth stroke code, fifth stroke code, sixth stroke code, seventh stroke code, crosshair stroke code, first circle stroke code, second circle stroke code, and third circle stroke code are displayed visually in different colors to generate a visual image of fundus lesions.

[0046] S6, statistically analyze the types of diabetic retinopathy lesions in the visualized fundus lesion image to obtain statistical results; if the statistical results indicate that only microaneurysm segmentation areas exist in the visualized fundus lesion image, the analysis result of the fundus lesion is output as mild non-proliferative diabetic retinopathy; if the statistical results indicate that hemorrhage segmentation areas exist in the visualized fundus lesion image and the number of hemorrhage segmentation areas is not greater than a preset threshold, the analysis result of the fundus lesion is output as moderate non-proliferative diabetic retinopathy; if the statistical results indicate that hard exudate segmentation areas exist in the visualized fundus lesion image and the number of hemorrhage segmentation areas in each of the four quadrants divided by the crosshairs is greater than a preset threshold, the analysis result of the fundus lesion is output as severe non-proliferative diabetic retinopathy.

[0047] S7, if the statistical results indicate that there is a hard exudate segmentation region in the visualized image of the fundus lesion and the hard exudate segmentation region is located within the first circle, then the analysis result of the fundus lesion is output as diabetic retinopathy with severe macular edema; if the statistical results indicate that there is a hard exudate segmentation region in the visualized image of the fundus lesion and the hard exudate segmentation region is located within the second circle, then the analysis result of the fundus lesion is output as diabetic retinopathy with moderate macular edema; if the statistical results indicate that there is a hard exudate segmentation region in the visualized image of the fundus lesion and the hard exudate segmentation region is located within the third circle, then the analysis result of the fundus lesion is output as diabetic retinopathy with mild macular edema.

[0048] For example: First, the Region of Interest (ROI) is extracted from the original fundus image to generate a region of interest (ROI) image. Second, the optic disc is extracted from the ROI image to generate an optic disc segmentation region; the macula is extracted from the ROI image to generate a macular segmentation region; arteries and veins are extracted from the ROI image to generate arteriovenous segmentation regions; fundus hemorrhages and microaneurysms are extracted from the ROI image to generate hemorrhage segmentation regions and microaneurysm segmentation regions; and hard and soft exudates are extracted from the ROI image to generate hard and soft exudate segmentation regions. Then, the first pass coding of the optic disc in the original fundus image is determined, the second pass coding of the macula in the original fundus image is determined, the third pass coding of the arteries and veins in the original fundus image is determined, the fourth pass coding of the hemorrhage in the original fundus image is determined, the fifth pass coding of the microaneurysm in the original fundus image is determined, the sixth pass coding of the soft exudate in the original fundus image is determined, and the seventh pass coding of the hard exudate in the original fundus image is determined. Finally, based on the first stroke encoding, a crosshair centered on the optic disc is obtained, generating the corresponding crosshair stroke encoding. Based on the second stroke encoding, a first circle with a radius of 500 μm, a second circle with a radius of 1500 μm, and a third circle with a radius of 3000 μm, centered on the macula, are determined, resulting in the first circle stroke encoding, the second circle stroke encoding, and the third circle stroke encoding. All stroke encodings are displayed in different colors on a blank image to obtain a visualized image of fundus lesions.

[0049] Different colors represent different types of diabetic retinopathy lesions. Fixed fundus features are rendered with cool tones, while abnormal fundus features are highlighted with warm tones. This results in strong contrast, making the images intuitive, clear, and easy to understand, especially suitable for medical imaging examination reports.

[0050] like Figure 2As shown in Figure a, which depicts mild non-proliferative diabetic retinopathy, the lesions are difficult to see with the naked eye. However, Figure b, a visualized image of the lesions, clearly shows the number and location of the lesions. This aids in the detection of minute lesions and occult fundus changes, avoiding missed diagnoses and misdiagnoses.

[0051] like Figure 3 As shown, Figure 3 It is severe nonproliferative diabetic retinopathy. Figure 4 yes Figure 3 The image shows a visualized image of fundus lesions. As shown in the figure, the method of this application simplifies the processing of fundus images and reflects them in the examination report, making it understandable for both doctors and patients, saving time in doctor-patient communication and improving the utilization rate of medical resources. The method of this application enables the display and analysis of diabetic retinopathy lesions, and has unique value, especially in medical image analysis reports. The analysis results of the visualized fundus lesion image are as follows. Figure 5 As shown.

[0052] like Figure 6 The diagram shown is a structural schematic of a device for visualizing diabetic retinopathy lesions provided in an embodiment of the present invention.

[0053] A device for visualizing diabetic retinopathy lesions, the device 600 comprising at least: an extraction module 601, configured to extract a region of interest from an original fundus image and generate a region of interest image; a determination module 602, configured to determine the feature run-length encoding of a segmented region in the region of interest image in the original fundus image, and to determine the lesion run-length encoding of a segmented region of diabetic retinopathy lesion in the region of interest image in the original fundus image; a generation module 603, configured to obtain a reference region for assisting in the identification of diabetic retinopathy lesions based on the feature run-length encoding, and generate a corresponding reference run-length encoding; and a visualization module 604, configured to display the feature run-length encoding, the lesion run-length encoding, and the reference run-length encoding in a blank image using different colors, respectively, to generate a fundus lesion visualization image.

[0054] In a preferred embodiment, the feature segmentation region includes at least an optic disc segmentation region, a macular segmentation region, and an arteriovenous vessel segmentation region; the determination module includes: a first extraction unit, used to extract the optic disc, macula, and arteriovenous vessels from the region of interest image, respectively, to generate corresponding optic disc segmentation regions, macular segmentation regions, and arteriovenous vessel segmentation regions; a first determination unit, used to determine a first stroke code of the optic disc in the original fundus image based on the position information of the optic disc segmentation region in the region of interest image; a second determination unit, used to determine a second stroke code of the macula in the original fundus image based on the position information of the macular segmentation region in the region of interest image; a third determination unit, used to determine a third stroke code of the arteriovenous vessels in the original fundus image based on the position information of the arteriovenous vessel segmentation region in the region of interest image; and a fourth determination unit, used to determine the first stroke code, the second stroke code, and the third stroke code as the feature stroke code of the feature segmentation region in the original fundus image.

[0055] In a preferred embodiment, the segmented region of the diabetic retinopathy lesion includes a hemorrhage segmentation region, and / or a microaneurysm segmentation region, and / or a soft exudate segmentation region, and / or a hard exudate segmentation region; the determination module includes: a second extraction unit, used to extract vascular hemorrhage, and / or microaneurysm, and / or soft exudate, and / or hard exudate from the region of interest image, respectively, to generate corresponding hemorrhage segmentation regions, and / or microaneurysm segmentation regions, and / or soft exudate segmentation regions, and / or hard exudate segmentation regions; a fifth determination unit, used to determine the fourth stroke code of vascular hemorrhage in the original fundus image based on the position information of the hemorrhage segmentation region in the region of interest image; a sixth determination unit, used to and / or based on the microaneurysm segmentation... The location information of the region in the region of interest image is used to determine the fifth stroke code of the microaneurysm in the original fundus image; the seventh determining unit is used to determine the sixth stroke code of the soft exudate segmentation region in the original fundus image based on the location information of the soft exudate segmentation region in the region of interest image; the eighth determining unit is used to determine the seventh stroke code of the hard exudate segmentation region in the original fundus image based on the location information of the hard exudate segmentation region in the region of interest image; the ninth determining unit is used to determine the fourth stroke code, and / or the fifth stroke code, and / or the sixth stroke code, and / or the seventh stroke code as the lesion stroke code of the diabetic retinopathy lesion segmentation region in the original fundus image.

[0056] In a preferred embodiment, the generation module includes: a first generation unit, configured to determine a crosshair centered on the optic disc based on a first travel code of the optic disc in the original fundus image and a first preset distance dataset, and generate a corresponding crosshair travel code; and a first determination unit, configured to determine the crosshair travel code as a reference travel code.

[0057] In a preferred embodiment, the generation module includes: a second generation unit, configured to determine a first circle centered on the macula based on a second stroke code of the macula in the original fundus image and a first preset radius, and generate a corresponding first circle stroke code; a third generation unit, configured to determine a second circle centered on the macula based on the second stroke code of the macula in the original fundus image and a second preset radius, and generate a corresponding second circle stroke code; a fourth generation unit, configured to determine a third circle centered on the macula based on the second stroke code of the macula in the original fundus image and a third preset radius, and generate a corresponding third circle stroke code; wherein the first preset radius is smaller than the second preset radius, and the second preset radius is smaller than the third preset radius; and a second determination unit, configured to determine the first circle stroke code, the second circle stroke code, and the third circle stroke code as reference stroke codes.

[0058] In a preferred embodiment, the device further includes: a statistics module for statistically analyzing the types of diabetic retinopathy lesions in the visualized fundus lesion image to obtain statistical results; a first output module for outputting the analysis result of fundus lesions as mild non-proliferative diabetic retinopathy if the statistical results indicate that only microaneurysm segmentation areas exist in the visualized fundus lesion image; a second output module for outputting the analysis result of fundus lesions as moderate non-proliferative diabetic retinopathy if the statistical results indicate that hemorrhage segmentation areas exist in the visualized fundus lesion image and the number of hemorrhage segmentation areas is not greater than a preset threshold; and a third output module for outputting the analysis result of fundus lesions as severe non-proliferative diabetic retinopathy if the statistical results indicate that hard exudate segmentation areas exist in the visualized fundus lesion image and the number of hemorrhage segmentation areas in each of the four quadrants divided by the crosshairs is greater than a preset threshold.

[0059] In a preferred embodiment, the device further includes: a statistics module, further configured to statistically analyze the types of diabetic retinopathy lesions in the visualized fundus lesion image to obtain statistical results; a first output module, further configured to output the analysis result of the fundus lesion as diabetic retinopathy with severe macular edema if the statistical results indicate that there is a hard exudate segmentation region in the visualized fundus lesion image and the hard exudate segmentation region is located within a first circle; a second output module, further configured to output the analysis result of the fundus lesion as diabetic retinopathy with moderate macular edema if the statistical results indicate that there is a hard exudate segmentation region in the visualized fundus lesion image and the hard exudate segmentation region is located within a second circle; and a third output module, further configured to output the analysis result of the fundus lesion as diabetic retinopathy with mild macular edema if the statistical results indicate that there is a hard exudate segmentation region in the visualized fundus lesion image and the hard exudate segmentation region is located within a third circle.

[0060] In a preferred embodiment, the determining module includes: a first extraction unit, further configured to extract the optic disc, macula, and blood vessels from the region of interest image, respectively, generating corresponding optic disc segmentation regions, macula segmentation regions, and blood vessel segmentation regions; a second extraction unit, configured to extract the optic cup from the optic disc segmentation regions, generating an optic cup segmentation region; a third extraction unit, configured to extract arteries and veins from the blood vessel segmentation regions, generating arteriovenous vessel segmentation regions; a first determining unit, further configured to determine a first stroke code of the optic disc in the original fundus image based on the position information of the optic disc segmentation region in the region of interest image; and a second determining unit, further configured to determine a second stroke of the macula in the original fundus image based on the position information of the macula segmentation region in the region of interest image. The third determining unit is further configured to determine the third stroke code of the blood vessel in the original fundus image based on the position information of the blood vessel segmentation region in the region of interest image; the fifth determining unit is configured to determine the eighth stroke code of the optic cup in the original fundus image based on the position information of the optic cup segmentation region in the optic disc segmentation region; the sixth determining unit is configured to determine the ninth stroke code of the arteriovenous vessels in the original fundus image based on the position information of the arteriovenous vessel segmentation region in the blood vessel segmentation region; the fourth determining unit is further configured to determine the first stroke code, the second stroke code, the third stroke code, the eighth stroke code, and the ninth stroke code as the feature stroke codes of the feature segmentation region in the original fundus image.

[0061] The above-described device can execute the method for visualizing diabetic retinopathy lesions provided in an embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the method for visualizing diabetic retinopathy lesions. Technical details not described in detail in this embodiment can be found in the method for visualizing diabetic retinopathy lesions provided in an embodiment of the present invention.

[0062] The present invention also provides an electronic device, comprising: a processor; a memory for storing executable instructions of the processor; the processor being configured to read the executable instructions from the memory and execute the instructions to implement the method for visualizing diabetic retinopathy lesions according to the present invention.

[0063] In addition to the methods and apparatus described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps in the methods according to various embodiments of this application described in the "Exemplary Methods" section above.

[0064] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0065] Furthermore, embodiments of this application may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps in the methods according to the following embodiments of this application described in the "Exemplary Methods" section above.

[0066] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0067] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the application to the necessity of employing the aforementioned specific details for implementation.

[0068] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0069] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.

[0070] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0071] The above description has been given for illustrative and descriptive purposes. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

[0072] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0073] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0074] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for visualizing diabetic retinopathy lesions, characterized in that, include: Extract the region of interest from the original fundus image and generate a region of interest image; The feature run-length encoding of the feature segmentation region in the region of interest image is determined in the original fundus image, and the lesion run-length encoding of the diabetic retinopathy lesion segmentation region in the region of interest image is determined in the original fundus image; Based on the aforementioned feature travel coding, a reference region for assisting in the identification of diabetic retinopathy lesions is obtained, and a corresponding reference travel coding is generated. The feature travel code, lesion travel code, and reference travel code are displayed in a blank image using different colors to generate a visual image of fundus lesions.

2. The method according to claim 1, characterized in that, The feature segmentation region includes at least an optic disc segmentation region, a macular segmentation region, and a vascular segmentation region; determining the feature run-length encoding of the feature segmentation region in the original fundus image within the region of interest image includes: The optic disc, macula, and blood vessels are extracted from the region of interest image to generate corresponding optic disc segmentation regions, macula segmentation regions, and blood vessel segmentation regions. Based on the position information of the optic disc segmentation region in the region of interest image, the first stroke code of the optic disc in the original fundus image is determined; Based on the location information of the macular segmentation region in the region of interest image, the second run-length encoding of the macula in the original fundus image is determined; Based on the location information of the segmented blood vessel region in the region of interest image, the third-stroke encoding of the blood vessel in the original fundus image is determined; The first travel code, the second travel code, and the third travel code are determined as the feature travel codes of the feature segmentation region in the original fundus image.

3. The method according to claim 1, characterized in that, The segmentation areas of the diabetic retinopathy lesions include hemorrhage segmentation areas, and / or microaneurysm segmentation areas, and / or soft exudate segmentation areas, and / or hard exudate segmentation areas. The step of determining the segmentation region of diabetic retinopathy lesions in the region of interest image within the original fundus image includes: Hemorrhage, and / or microaneurysm, and / or soft exudate, and / or hard exudate are extracted from the region of interest image, respectively, to generate corresponding hemorrhage segmentation region, and / or microaneurysm segmentation region, and / or soft exudate segmentation region, and / or hard exudate segmentation region. Based on the location information of the hemorrhage segmentation region in the region of interest image, the fourth stroke code of vascular hemorrhage in the original fundus image is determined; And / or based on the location information of the microaneurysm segmentation region in the region of interest image, determine the fifth-stroke code of the microaneurysm in the original fundus image; And / or based on the location information of the soft exudate segmentation region in the region of interest image, determine the sixth-stroke code of the soft exudate in the original fundus image; And / or, based on the location information of the hard exudate segmentation region in the region of interest image, determine the seventh stroke code of the hard exudate in the original fundus image; The fourth stroke code, and / or the fifth stroke code, and / or the sixth stroke code, and / or the seventh stroke code are determined as the stroke codes of the diabetic retinopathy lesion segmentation region in the original fundus image.

4. The method according to claim 2, characterized in that, The step of obtaining a reference region for assisting in the identification of diabetic retinopathy lesions based on the feature travel coding, and generating a corresponding reference travel coding, includes: Based on the first stroke code of the optic disc in the original fundus image and the first preset distance dataset, the cross line centered on the optic disc is determined, and the corresponding cross line stroke code is generated. The crosshair travel code is determined as the reference travel code.

5. The method according to claim 3, characterized in that, The step of obtaining a reference region for assisting in the identification of diabetic retinopathy lesions based on the feature travel coding, and generating a corresponding reference travel coding, includes: Based on the second stroke code of the macula in the original fundus image and the first preset radius, a first circle centered on the macula is determined, and a corresponding first circle stroke code is generated. Based on the second stroke code of the macula in the original fundus image and the second preset radius, a second circle centered on the macula is determined, and a corresponding second circle stroke code is generated. Based on the second stroke code of the macula in the original fundus image and the third preset radius, a third circle centered on the macula is determined, and a corresponding third circle stroke code is generated; wherein, the first preset radius is smaller than the second preset radius, and the second preset radius is smaller than the third preset radius; The first circular travel code, the second circular travel code, and the third circular travel code are determined as reference travel codes.

6. The method according to claim 4, characterized in that, Also includes: The types of diabetic retinopathy lesions in the visualized fundus lesion images were statistically analyzed, and the statistical results were obtained. If the statistical results indicate that only microaneurysm segmentation areas exist in the visualized image of the fundus lesion, then the output analysis result of the fundus lesion is mild non-proliferative diabetic retinopathy. If the statistical results indicate that there are hemorrhage segmentation regions in the visualized image of the fundus lesion and the number of hemorrhage segmentation regions is not greater than a preset threshold, then the analysis result of the fundus lesion is output as moderate non-proliferative diabetic retinopathy. If the statistical results indicate that there are hard exudate segmentation regions in the visualized image of the fundus lesion and the number of hemorrhage segmentation regions in each of the four quadrants divided by the crosshairs is greater than a preset threshold, then the analysis result of the fundus lesion is output as severe nonproliferative diabetic retinopathy.

7. The method according to claim 5, characterized in that, Also includes: The types of diabetic retinopathy lesions in the visualized fundus lesion images were statistically analyzed, and the statistical results were obtained. If the statistical results indicate that there is a hard exudate segmentation region in the visualized image of the fundus lesion and the hard exudate segmentation region is located within the first circle, then the output analysis result of the fundus lesion is diabetic retinopathy with severe macular edema. If the statistical results indicate that there is a hard exudate segmentation region in the visualized image of the fundus lesion and the hard exudate segmentation region is located within the second circle, then the output analysis result of the fundus lesion is diabetic retinopathy with moderate macular edema. If the statistical results indicate that there is a hard exudate segmentation region in the visualized image of the fundus lesion and the hard exudate segmentation region is located within the third circle, then the output analysis result of the fundus lesion is diabetic retinopathy with mild macular edema.

8. The method according to claim 1, characterized in that, The step of determining the feature run-length encoding of the feature segmentation region in the region of interest image in the original fundus image includes: The optic disc, macula, and blood vessels are extracted from the region of interest image to generate corresponding optic disc segmentation regions, macula segmentation regions, and blood vessel segmentation regions. Extract the optic cup from the optic disc segmentation region to generate the optic cup segmentation region; Arteries and veins are extracted from the segmented vascular region to generate an arteriovenous segmentation region; Based on the position information of the optic disc segmentation region in the region of interest image, the first stroke code of the optic disc in the original fundus image is determined; Based on the location information of the macular segmentation region in the region of interest image, the second run-length encoding of the macula in the original fundus image is determined; Based on the location information of the segmented blood vessel region in the region of interest image, the third-stroke encoding of the blood vessel in the original fundus image is determined; Based on the position information of the optic cup segmentation region in the optic disc segmentation region, the eighth stroke code of the optic cup in the original fundus image is determined; Based on the location information of the arteriovenous vessel segmentation region in the vessel segmentation region, the ninth stroke code of the arteriovenous vessel in the original fundus image is determined; the first stroke code, the second stroke code, the third stroke code, the eighth stroke code, and the ninth stroke code are determined as the feature stroke code of the feature segmentation region in the original fundus image.

9. A device for visualizing diabetic retinopathy lesions, characterized in that, include: The extraction module is used to extract the region of interest from the original fundus image and generate a region of interest image; The determination module is used to determine the feature run-length encoding of the feature segmentation region in the region of interest image in the original fundus image, and to determine the lesion run-length encoding of the diabetic retinopathy lesion segmentation region in the region of interest image in the original fundus image; The generation module is used to obtain a reference region for assisting in the identification of diabetic retinopathy lesions based on the feature travel coding, and generate a corresponding reference travel coding; The visualization module is used to display the feature travel code, lesion travel code, and reference travel code in a blank image using different colors to generate a visualization image of fundus lesions.

10. An electronic device, comprising: processor; Memory used to store the processor's executable instructions; The processor is configured to read the executable instructions from the memory and execute the instructions to implement the method as described in any one of claims 1-8.

11. A computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the method as claimed in any one of claims 1-8.

Images