Method, device, medium and equipment for evaluating fundus vascular morphological characteristics

Abstract

Embodiments of the present invention provide a method, apparatus, medium, and device for assessing fundus vascular morphological characteristics. The method comprises: S1: preprocessing a fundus image; S2: performing vessel segmentation on the preprocessed fundus image to obtain a segmented fundus vascular image; and S3: determining, based on the segmented fundus vascular image, multiple parameters describing fundus vascular morphological characteristics. The advantages of embodiments of the present invention are that they use digital language to comprehensively describe vascular morphological characteristics from various angles, thereby providing a detailed understanding of fundus vascular morphology, facilitating the detailed assessment of abnormalities and changes in fundus vascular structure.

Description

Technical Field

[0001] The present invention relates to the technical field of fundus image processing, and in particular to a method, device, medium and equipment for evaluating fundus blood vessel morphological characteristics. Background Art

[0002] Retinal microvessels are the only deep-seated microvessels in the human body that can be directly observed non-invasively. Changes in their state or structure can indicate the occurrence of a variety of diseases, such as hypertension and diabetes. Therefore, accurate segmentation of retinal vascular structures in the fundus can assist clinicians in diagnosing and treating diseases. Furthermore, retinal vascular analysis can enable early diagnosis and prevention of certain diseases, which is of great significance both in theoretical research and clinical practice. However, a suitable technical approach is currently lacking for comprehensive and detailed assessment of fundus vascular morphology. Summary of the Invention

[0003] In view of this, an object of the embodiments of the present invention is to provide a method, apparatus, medium and device for evaluating fundus vascular morphological characteristics, so as to comprehensively reflect the vascular morphological characteristics.

[0004] To achieve the above objectives, in a first aspect, the present invention provides a method for evaluating fundus vascular morphology, the method comprising:

[0005] S1: preprocessing the fundus image to obtain a preprocessed fundus image;

[0006] S2: performing blood vessel segmentation processing on the pre-processed fundus image to obtain a fundus blood vessel segmentation image;

[0007] S3: Determine a plurality of parameters describing the morphological characteristics of the fundus blood vessels according to the fundus blood vessel segmentation image obtained by the segmentation process.

[0008] In some possible implementations, determining a region of interest (ROI) based on the fundus image, step S3 includes:

[0009] S31: determining a first set of morphological parameters of the fundus blood vessels or a first set of morphological parameters corresponding to any target sub-region within the ROI according to the fundus blood vessel segmentation image obtained by the segmentation process;

[0010] The first group of morphological parameters includes any one of fundus vessel fractal dimension, fundus vessel area, fundus vessel density, fundus vessel area ratio and fundus vessel spacing.

[0011] In some possible implementations, step S31 specifically includes:

[0012] Determining a fundus artery image and a fundus vein image from the fundus blood vessel segmentation image obtained by segmentation processing; or directly identifying and determining the fundus artery image and the fundus vein image based on the preprocessed fundus image;

[0013] Determining a first set of morphological parameters associated with the fundus arteries according to the fundus artery image, or a first set of morphological parameters associated with the fundus arteries corresponding to any target sub-region within the ROI;

[0014] A first set of morphological parameters associated with the fundus venous vessels is determined according to the fundus venous vessel image, or a first set of morphological parameters associated with the fundus venous vessels corresponding to any target sub-region within the ROI.

[0015] In some possible implementations, determining a first set of morphological parameters corresponding to any target sub-region within the ROI based on the fundus vascular segmentation image obtained by segmentation processing in step S31 specifically includes:

[0016] S311: Taking the optic disc or macula as a reference, extracting a segmented image of the fundus vessels in a preset area around the optic disc or macula from the segmented image of the fundus vessels, and determining a first set of morphological parameters corresponding to the preset area based on the segmented image of the fundus vessels in the preset area; or

[0017] S312: Determine, based on the multiple blood vessel branch points selected by the user, a first set of morphological parameters corresponding to the selected sub-region including the multiple blood vessel branch points.

[0018] In some possible implementations, step S311 specifically includes:

[0019] Taking the optic disc or macula as a reference, extracting the fundus arterial vessel image and the fundus venous vessel image in a preset area around the optic disc or macula from the fundus vessel segmentation image;

[0020] Determining a first set of morphological parameters associated with the fundus arteries corresponding to the preset area based on the fundus artery image within the preset area around the optic disc or macula;

[0021] Determining a first set of morphological parameters associated with the fundus venous vessels corresponding to the preset area based on the fundus venous vessel image within the preset area around the optic disc or macula;

[0022] Step S312 specifically includes:

[0023] Determining, based on the plurality of vascular branch points selected by the user, a selected subregion including the plurality of vascular branch points;

[0024] extracting a fundus arterial vessel image and a fundus venous vessel image from the selected subregion;

[0025] Determining a first set of morphological parameters associated with the fundus arteries corresponding to the selected subregion based on the fundus artery image extracted from the selected subregion;

[0026] According to the fundus venous vessel image extracted from the selected sub-region, a first group of morphological parameters associated with the fundus venous vessels corresponding to the selected sub-region is determined.

[0027] The method for determining the proportion of fundus blood vessel area is as follows:

[0028] Obtaining the sum of the areas of all fundus blood vessels in the fundus blood vessel segmentation image as a first fundus blood vessel total area;

[0029] Obtaining the sum of the areas of the fundus blood vessels in the fundus blood vessel segmentation image in a preset area around the optic disc or macula as a second total fundus blood vessel area;

[0030] The ratio of the total area of ​​the second fundus vessels to the total area of ​​the first fundus vessels is determined as the fundus vessel area ratio in the preset area around the optic disc or macula. The method for determining the fundus artery area ratio is as follows:

[0031] Obtaining the sum of the areas of all fundus arteries in the fundus vessel segmentation image as the first fundus artery total area;

[0032] Obtaining the sum of the areas of the fundus arteries in the fundus artery image within a preset area around the optic disc or macula as a second total fundus artery area;

[0033] The ratio of the total area of ​​the second fundus arteries to the total area of ​​the first fundus arteries is determined as the proportion of the fundus artery area in the preset area around the optic disc or macula.

[0034] The method for determining the proportion of fundus venous area is as follows:

[0035] Obtaining the sum of the areas of all fundus venous vessels in the fundus venous segmentation image as the first fundus venous vessel total area;

[0036] Obtaining the sum of the areas of the fundus venous vessels in the fundus venous vessel image within a preset area around the optic disc or macula as a second total fundus venous vessel area;

[0037] The ratio of the total area of ​​the second fundus venous vessels to the total area of ​​the first fundus venous vessels is determined as the proportion of the fundus venous vessel area in the preset area around the optic disc or macula.

[0038] In some possible implementations, step S3 further includes:

[0039] S32: determining a fundus vessel centerline based on the fundus vessel segmentation image obtained by the segmentation process, or obtaining the fundus vessel centerline based on the preprocessed fundus image; and determining a second set of morphological parameters or a second set of morphological parameters corresponding to any target subregion within the ROI based on the fundus vessel centerline;

[0040] The second group of morphological parameters includes any multiple of fundus vessel fractal dimension, fundus vessel length, fundus vessel density and fundus vessel spacing.

[0041] In some possible implementations, step S32 specifically includes:

[0042] Determining a fundus artery image and a fundus venous image from the fundus blood vessel segmentation image obtained by segmentation processing; or directly identifying and determining the fundus artery image and the fundus venous image based on the preprocessed fundus image;

[0043] Determine a fundus artery centerline based on the fundus artery image, or obtain the fundus artery centerline based on the preprocessed fundus image, and determine a second set of morphological parameters associated with the fundus artery based on the fundus artery centerline, or determine the second set of morphological parameters associated with the fundus artery corresponding to any target sub-region within the ROI;

[0044] The fundus venous centerline is determined based on the fundus venous image, or the fundus venous centerline is obtained based on the preprocessed fundus image, and a second set of morphological parameters associated with the fundus venous vessels is determined based on the fundus venous centerline, or the second set of morphological parameters associated with the fundus venous vessels corresponding to any target sub-region within the ROI.

[0045] In some possible implementations, the second set of morphological parameters determined in step S32 based on the fundus blood vessel centerline or the second set of morphological parameters corresponding to any target sub-region within the ROI specifically include:

[0046] S321: extracting a fundus vessel segmentation image within a preset area around the optic disc or macula from the fundus vessel segmentation image, taking the optic disc or macula as a reference; determining a fundus vessel centerline within the preset area based on the fundus vessel segmentation image within the preset area around the optic disc or macula, or obtaining the fundus vessel centerline within the preset area based on a preprocessed fundus image within the preset area; determining a second set of morphological parameters corresponding to the preset area based on the fundus vessel centerline within the preset area; or,

[0047] S322: Determine a selected sub-region including multiple vascular branch points based on the multiple vascular branch points selected by the user; determine the fundus vascular centerline within the selected sub-region based on the fundus vascular segmentation image within the selected sub-region, and determine a second set of morphological parameters corresponding to the selected sub-region based on the fundus vascular centerline within the selected sub-region.

[0048] In some possible implementations, step S321 specifically includes:

[0049] Taking the optic disc or macula as a reference, extracting the fundus arterial vessel image and the fundus venous vessel image in a preset area around the optic disc or macula from the fundus vessel segmentation image;

[0050] Determining a centerline of the fundus artery according to an image of the fundus artery in a preset area around the optic disc or macula, or obtaining the centerline of the fundus artery in the preset area based on a preprocessed fundus image; determining a second set of parameters associated with the fundus artery corresponding to the preset area according to the centerline of the fundus artery in the preset area;

[0051] Determining a centerline of the fundus veins based on an image of the fundus veins in a preset area around the optic disc or macula, or obtaining the centerline of the fundus veins in the preset area based on a preprocessed fundus image; determining a second set of parameters associated with the fundus veins corresponding to the preset area based on the centerline of the fundus veins in the preset area;

[0052] Step S322 specifically includes:

[0053] Determining, based on the plurality of vascular branch points selected by the user, a selected subregion including the plurality of vascular branch points;

[0054] extracting a fundus arterial vessel image and a fundus venous vessel image from the selected subregion;

[0055] Determining a centerline of the fundus artery within the selected subregion based on the fundus artery image extracted from the selected subregion, and determining a second set of morphological parameters associated with the fundus artery corresponding to the selected subregion based on the centerline of the fundus artery within the selected subregion;

[0056] Based on the fundus venous vessel image extracted from the selected sub-region, the centerline of the fundus venous vessel in the selected sub-region is determined, and based on the centerline of the fundus venous vessel in the selected sub-region, a second set of morphological parameters associated with the fundus venous vessel corresponding to the selected sub-region is determined.

[0057] In some possible implementations, the fundus vascular density includes fundus vascular linear density and fundus vascular surface density, and the fundus vascular surface density can be determined based on the following method:

[0058] Extract the region of interest (ROI) of the fundus image;

[0059] Performing threshold segmentation on the pre-processed fundus image to obtain fundus vascular areas;

[0060] Calculate the area of ​​fundus vascular region;

[0061] Calculate the area of ​​the region of interest (ROI);

[0062] The blood vessel surface density is obtained according to the following formula:

[0063] Blood vessel surface density = area of ​​fundus blood vessels / area of ​​region of interest (ROI).

[0064] Preset area blood vessel density = preset area fundus blood vessel area / preset area fundus area;

[0065] Arterial blood vessel surface density in the preset area = fundus arterial blood vessel area in the preset area / fundus area in the preset area;

[0066] The venous blood vessel surface density of the preset area=the fundus venous blood vessel area of ​​the preset area / the fundus area of ​​the preset area.

[0067] In some possible implementations, the fundus vessel density may be determined based on the following method:

[0068] Extract the region of interest (ROI) of the fundus image;

[0069] Performing threshold segmentation on the pre-processed fundus image to obtain the fundus vascular area;

[0070] Separate the fundus vascular area to obtain each independent blood vessel;

[0071] Perform morphological operations on each blood vessel to obtain the blood vessel skeleton line or blood vessel center line, which is recorded as the blood vessel line;

[0072] Calculate the length of each blood vessel line and then obtain the total length of all blood vessel lines in the region of interest (ROI);

[0073] Calculate the area of ​​the region of interest (ROI);

[0074] The vascular line density was obtained according to the following formula: vascular line density = total length of all vascular lines in the region of interest (ROI) / area of ​​the region of interest (ROI).

[0075] The blood vessel line density of the preset area = the total length of the fundus blood vessels in the preset area / the fundus area of ​​the preset area;

[0076] Arterial blood vessel linear density in the preset area = total length of fundus arterial blood vessels in the preset area / fundus area in the preset area;

[0077] The venous blood vessel linear density in the preset area = the total length of the fundus venous blood vessels in the preset area / the fundus area in the preset area.

[0078] In some possible implementations, step S1 specifically includes:

[0079] S11: extracting the region of interest (ROI) of the fundus image;

[0080] S12: performing denoising processing on the ROI;

[0081] S13: performing normalization processing on the denoised image;

[0082] S14: performing enhancement processing on the normalized image to obtain a pre-processed fundus image.

[0083] In some possible implementations, step S2 specifically includes:

[0084] S21: performing threshold segmentation on the preprocessed fundus image using the color and morphological features of the fundus blood vessels to obtain an initial fundus blood vessel segmentation image, and correcting the initial fundus blood vessel segmentation image to obtain a final sample image;

[0085] S22: Based on the final sample image, a semantic segmentation network is used to perform model training, and forward processing is performed through the trained model to output a confidence probability map of the same size as the trained final sample image;

[0086] S23: Convert the confidence probability map into a binary image according to a set threshold.

[0087] S24: Obtaining a fundus blood vessel segmentation image including a fundus blood vessel region according to the binary image.

[0088] In a second aspect, a device for evaluating fundus vascular morphology is provided, the device comprising:

[0089] A fundus image preprocessing module is used to preprocess the fundus image to obtain a preprocessed fundus image;

[0090] a fundus blood vessel segmentation processing module, configured to perform blood vessel segmentation processing on the fundus image to obtain a fundus blood vessel segmentation image;

[0091] The fundus vascular morphology characteristic index determination module is used to determine a plurality of parameters describing the fundus vascular morphology characteristics based on the fundus vascular segmentation image.

[0092] Optionally, the multiple parameters include: fundus vessel fractal dimension, fundus vessel length and fundus vessel density.

[0093] In some possible implementations, the fundus image preprocessing module is specifically configured to:

[0094] Extract the region of interest (ROI) of the fundus image;

[0095] performing denoising processing on the ROI;

[0096] Perform normalization on the denoised image;

[0097] The normalized image is enhanced to obtain a preprocessed fundus image.

[0098] In some possible implementations, the fundus blood vessel segmentation processing module is specifically configured to:

[0099] Performing threshold segmentation on the preprocessed fundus image using the color and morphological features of the fundus blood vessels to obtain an initial fundus blood vessel segmentation image, and correcting the initial fundus blood vessel segmentation image to obtain a final sample image;

[0100] Based on the final sample image, a semantic segmentation network is used to train a model, and forward processing is performed through the trained model to output a confidence probability map of the same size as the trained final sample image;

[0101] According to a set threshold, the confidence probability map is converted into a binary image;

[0102] According to the binary image, a fundus vascular area and a fundus vascular segmentation image are obtained.

[0103] In a third aspect, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, any one of the methods for evaluating fundus vascular morphological characteristics as described above is implemented.

[0104] According to a fourth aspect, a computer device is provided, comprising:

[0105] processor;

[0106] a memory for storing instructions executable by the processor;

[0107] The processor is configured to execute the instructions to implement any one of the methods for evaluating fundus vascular morphological characteristics as described above.

[0108] The above technical solution has the following beneficial effects:

[0109] The advantages of the embodiments of the present invention are that they use digital language to comprehensively describe the morphological characteristics of blood vessels from various angles, thereby accurately grasping the morphological conditions of fundus blood vessels, which is conducive to the accurate assessment of abnormalities and changes in fundus blood vessel structures. BRIEF DESCRIPTION OF THE DRAWINGS

[0110] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0111] Figure 1A This is a flow chart of a method for evaluating fundus vascular morphological characteristics according to an embodiment of the present invention;

[0112] Figure 1B is a schematic diagram of a classification logic structure of multiple parameters according to an embodiment of the present invention;

[0113] Figure 2 is a flowchart of the preprocessing of an embodiment of the present invention;

[0114] Figure 3 1 is an example of a processing process image of the preprocessing of an embodiment of the present invention;

[0115] Figure 4 is a flowchart of fundus blood vessel segmentation processing according to an embodiment of the present invention;

[0116] Figure 5 FIG. 1 is an image showing a processing process of fundus blood vessel segmentation as an example in an embodiment of the present invention;

[0117] Figure 6 is a flow chart of determining multiple indicators indicating fundus vascular morphological characteristics according to an embodiment of the present invention;

[0118] Figure 7 is a diagram of a calculation process of box counting dimension as an example in an embodiment of the present invention;

[0119] Figure 8 This is a functional block diagram of a device for evaluating fundus vascular morphological characteristics according to an embodiment of the present invention;

[0120] Figure 9 is a functional block diagram of a computer-readable storage medium according to an embodiment of the present invention;

[0121] Figure 10 This is a functional block diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0122] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0123] The purpose of the embodiment of the present invention is to automatically identify and segment the human fundus blood vessels based on the bionic mechanism of human vision, and distinguish between arteries and veins, and then calculate indicators such as fractal dimension, blood vessel density, and blood vessel length to comprehensively reflect the vascular morphological characteristics.

[0124] The advantages of the embodiments of the present invention are that they use digital language to comprehensively describe the morphological characteristics of blood vessels from various angles, thereby accurately grasping the morphological conditions of fundus blood vessels, which is conducive to the accurate assessment of abnormalities and changes in fundus blood vessel structures.

[0125] Example 1

[0126] The embodiment of the present invention uses artificial intelligence image processing technology to extract fundus blood vessels and calculate indicators such as fundus blood vessel fractal dimension, fundus blood vessel density, and fundus blood vessel centerline length to represent the morphological characteristics of fundus blood vessels. It specifically includes three processes: image preprocessing, fundus blood vessel segmentation, and morphological feature indicator calculation.

[0127] Figure 1A FIG. 1 is a flow chart of a method for evaluating fundus vascular morphology characteristics provided by an embodiment of the present invention. Figure 1A As shown, the method includes the following steps:

[0128] S1: preprocessing the fundus image to obtain a preprocessed fundus image;

[0129] S2: performing blood vessel segmentation processing on the pre-processed fundus image to obtain a fundus blood vessel segmentation image;

[0130] S3: Determine multiple parameters describing the morphological characteristics of the fundus blood vessels based on the fundus blood vessel segmentation image.

[0131] In some embodiments, the pre-processing step in S1 may be omitted.

[0132] Figure 1B FIG. 1 is a schematic diagram of a classification logic structure of multiple parameters in an embodiment of the present invention. Figure 1BAs shown, the multiple parameters may include a first set of morphological parameters and a second set of morphological parameters. The first set of morphological parameters and the second set of morphological parameters respectively include morphological parameters corresponding to the ROI and morphological parameters corresponding to the target sub-region. The morphological parameters corresponding to the ROI and the morphological parameters corresponding to the target sub-region respectively include morphological parameters associated with fundus arteries and morphological parameters associated with fundus venous vessels. The target sub-region includes two cases: one is a preset area around the optic disc or macula, and the other is a selected sub-region selected by the user and including multiple vascular branch points. In this embodiment, the vascular morphological parameters of any target area or selected area of ​​the optic disc or macula period can be calculated with the optic disc or macula as the reference.

[0133] In some embodiments, in step S1, a region of interest (ROI) is further determined based on the fundus image, and step S3 specifically includes:

[0134] S31: determining a first set of morphological parameters corresponding to the ROI or a first set of morphological parameters corresponding to any target sub-region within the ROI based on the fundus blood vessel segmentation image obtained by the segmentation process;

[0135] The first group of morphological parameters includes any one of fundus vessel fractal dimension, fundus vessel area, fundus vessel density, fundus vessel area ratio or fundus vessel spacing.

[0136] Figure 2 4 is a flowchart of preprocessing according to an embodiment of the present invention. Figure 3 FIG. 1 is an image of an example of a preprocessing process of an embodiment of the present invention. Figure 3 (a) is the original image, Figure 3 (b) is the ROI image, Figure 3 (c) is the image before normalization. Figure 3 (d) is the normalized image. Figure 3 (e) is the image before enhancement processing, Figure 3 (f) is the enhanced image. Figure 2 and Figure 3 As shown, step S1 may specifically include:

[0137] S11: extracting the region of interest (ROI) of the fundus image;

[0138] Specifically, ROI extraction refers to extracting the valid area of ​​the fundus image, removing invalid areas such as the background, and reducing the interference of non-fundus content on subsequent blood vessel segmentation. First, the image is channel-separated, and the red channel image is segmented using a threshold segmentation method to obtain ROI candidate regions. The candidate regions are then screened using at least one or more features such as position, area, and circularity. Morphological erosion and opening operations (opening operations are basic image morphology algorithms) are used to determine the boundaries, thereby determining the final ROI region.

[0139] S12: performing denoising on the ROI; the main purpose of the denoising is to reduce the noise formed on the image during the shooting and camera imaging process, which is achieved by a low-pass filtering method.

[0140] S13: performing normalization processing on the denoised image;

[0141] Specifically, normalization processing is mainly used to reduce the differences between images, including brightness normalization, color normalization and size normalization. Among them, color normalization refers to: converting the fundus image from RGB color space to LAB color space, then performing mean calibration in LAB color space, and then converting it back to RGB (Red Green Blue) color space to obtain it. Brightness normalization refers to: converting the image from RGB color space to HSI (Hue Saturation Intensity) color space, then performing mean calibration on each channel, and converting it back to RGB color space after calibration. Size normalization refers to resampling the fundus image so that the size of the image is within a uniform range.

[0142] S14: performing enhancement processing on the normalized image to obtain a pre-processed fundus image.

[0143] Image enhancement processing refers to using the contrast-limited adaptive histogram equalization (CLAHE) algorithm to enhance the image in the ROI area.

[0144] Before blood vessel segmentation, the fundus image is first subjected to a series of fundus image preprocessing, including ROI (Region Of Interest) extraction, denoising, normalization, and enhancement, so as to improve the stability of subsequent blood vessel extraction.

[0145] The above is only an example of a preprocessing method, but is not limited thereto. Other preprocessing methods can also be applied to this embodiment.

[0146] Figure 4is a flowchart of fundus blood vessel segmentation processing according to an embodiment of the present invention; Figure 5 FIG. 1 is an image showing an example of the fundus blood vessel segmentation process according to an embodiment of the present invention. Figure 5 (a) is the original image, Figure 5 (b) is the fundus vascular segmentation map. The vascular segmentation based on the image obtained by fundus image preprocessing mainly includes three steps: sample labeling, model training and vascular feature segmentation. Figure 4 and Figure 5 As shown, step S2 specifically includes:

[0147] S21: performing threshold segmentation on the preprocessed fundus image using the color and morphological features of the fundus blood vessels to obtain an initial fundus blood vessel segmentation image, and correcting the initial fundus blood vessel segmentation image to obtain a final sample image;

[0148] Specifically, this step first performs threshold segmentation using the color and morphological features of the blood vessels to obtain an initial fundus blood vessel segmentation image, and then corrects the initial fundus blood vessel segmentation image to obtain a final sample image.

[0149] S22: Based on the final sample image, a semantic segmentation network is used to train the model and output a confidence probability map of the same size as the final sample image.

[0150] Model training: Based on the labeled sample images, a semantic segmentation network such as resnet101-unet is used for model training. The final output is a confidence probability map of the same size as the training image, and then a vascular segmentation image is obtained.

[0151] S23: Convert the confidence probability map into a binary image according to the set threshold.

[0152] S24: Obtaining a fundus blood vessel segmentation image including the fundus blood vessel region based on the binary image.

[0153] Specifically, in the blood vessel segmentation step, the trained model is used to segment and extract the blood vessels.

[0154] The above is only an example of a blood vessel segmentation method, and the present invention does not specifically limit the blood vessel segmentation and the method for determining the blood vessel centerline.

[0155] Figure 6 FIG. 1 is a flow chart of determining multiple parameters describing the morphological characteristics of fundus blood vessels according to an embodiment of the present invention. Figure 6 As shown, step S31 specifically includes:

[0156] Determine the fundus artery image and the fundus vein image from the fundus blood vessel segmentation image obtained by the segmentation process; or directly identify and determine the fundus artery image and the fundus vein image based on the preprocessed fundus image;

[0157] Determining a first set of morphological parameters associated with the fundus arteries according to the fundus artery image, or a first set of morphological parameters associated with the fundus arteries corresponding to any target sub-region within the ROI;

[0158] A first set of morphological parameters associated with the fundus venous vessels is determined according to the fundus venous vessel image, or a first set of morphological parameters associated with the fundus venous vessels corresponding to any target sub-region within the ROI.

[0159] The fundus vessel fractal dimension includes any one or more of the following: Hausdorff dimension, similarity dimension, box counting dimension, capacity dimension, correlation dimension, and information dimension. All morphological parameters can be distinguished between arterial morphological parameters and venous morphological parameters.

[0160] Specifically, fractal dimension calculations may include: overall fractal dimension, arterial fractal dimension, and venous fractal dimension. Fractal dimension is the most important parameter for describing fractals, reflecting the effectiveness of the space occupied by complex shapes and measuring the irregularity of complex shapes. Fractal dimension reflects the morphological characteristics of a graph. In the embodiments of the present invention, the fractal characteristics of retinal vessels are represented by calculating the box-counting dimension of the retinal vessels. This dimension can, to a certain extent, reflect the extension and coverage of the vessels, objectively reflecting the complexity and richness of the vessels as a whole.

[0161] Figure 7 FIG. 1 is a diagram showing the calculation process of the box-counting dimension in an embodiment of the present invention as an example. Figure 7 As shown, the fundus image is first gridded, and multiple grids with the same side length (εi) are formed on the fundus image each time, where i is a positive integer ranging from 1 to n; the number of grids (Ni) that intersect with the fundus blood vessels in the multiple grids with the same side length (εi) is detected; the above process is repeated until the number of grids (Nn) corresponding to the grid with the side length (εn) is obtained; then the number of grids (Ni) that intersect with the fundus blood vessels obtained in each operation and the inverse of the side length (εi) of the grid corresponding to this operation are fitted with a straight line under logarithmic conditions, and the slope of the straight line is the box counting dimension.

[0162] For example, Figure 7In the figure, five grids with increasing side lengths are plotted as examples. These correspond to grids with side lengths of ε1, ε2, ε3, ε4, and ε5. Each time, multiple grids with the same side length (εi) are formed on the fundus image, and the number of grids (N) that intersect with fundus vessels in each grid with the corresponding side length (ε1, ε2, ε3, ε4, ε5) is calculated, namely N1, N2, N3, N4, and N5, respectively. Each change in the side length (ε) of a grid and the calculation of the number of grids (N) that intersect with fundus vessels is counted as one operation.

[0163] like Figure 7 As shown in the figure, rectangles of different colors or side lengths represent grids of different side lengths. It can be seen that as the grid side length changes, the intersection of the grid and the fundus vessels also changes. The number of grids (N) that intersect with the fundus vessels obtained in each operation and the inverse of the side length of the grid corresponding to this operation (ε) are fitted to a straight line under logarithmic conditions. The slope of the straight line is the box counting dimension, which is expressed as:

[0164]

[0165] Vascular spacing:

[0166]

[0167] Λ represents the mean intervascular space, where σ is the standard deviation of the pixel values ​​of each grid pixel under a given size ε, μ is the mean of the pixel values ​​of all pixels of each grid under a given size ε, and n is the number of grid sizes (grid side lengths).

[0168] Fractal dimension: Fractal theory, known as nature's geometry, reflects the effectiveness of complex shapes occupying space. It is a measure of the irregularity of complex shapes and is also called fractal dimension. The fractal dimension of a straight line is 1, the fractal dimension of a rectangular plane is 2, and the fractal dimension of a three-dimensional figure is 3. Therefore, the present invention introduces the concept of fractal dimension to retinal blood vessels to represent a characteristic of these vessels. Several fractal dimensions can be calculated in the present invention: Hausdorff dimension, similarity dimension, box-counting dimension, capacity dimension, correlation dimension, and information dimension.

[0169] In a further embodiment, in step S31, based on the fundus blood vessel segmentation image obtained by the segmentation process, determining a first set of morphological parameters corresponding to any target sub-region within the ROI specifically includes:

[0170] S311: Taking the optic disc or macula as a reference, extracting a segmented image of the fundus vessels in a preset area around the optic disc or macula from the fundus vessel segmentation image, and determining a first set of morphological parameters corresponding to the preset area based on the segmented image of the fundus vessels in the preset area; or

[0171] S312: Determine, based on the multiple blood vessel branch points selected by the user, a first set of morphological parameters corresponding to the selected sub-region including the multiple blood vessel branch points.

[0172] In a further embodiment, step S311 specifically includes:

[0173] Taking the optic disc or macula as a reference, extracting the fundus arterial vessel image and the fundus venous vessel image in a preset area around the optic disc or macula from the fundus vessel segmentation image;

[0174] Determining a first set of morphological parameters associated with the fundus arteries corresponding to the preset area based on the fundus artery image within the preset area around the optic disc or macula;

[0175] Determining a first set of morphological parameters associated with the fundus venous vessels corresponding to the preset area based on the fundus venous vessel image within the preset area around the optic disc or macula;

[0176] In a further embodiment, step S312 specifically includes:

[0177] Determining, based on the plurality of vascular branch points selected by the user, a selected subregion including the plurality of vascular branch points;

[0178] extracting a fundus arterial vessel image and a fundus venous vessel image from the selected subregion;

[0179] Determining a first set of morphological parameters associated with the fundus arteries corresponding to the selected subregion based on the fundus artery image extracted from the selected subregion;

[0180] According to the fundus venous vessel image extracted from the selected sub-region, a first group of morphological parameters associated with the fundus venous vessels corresponding to the selected sub-region is determined.

[0181] In some embodiments, step S3 may further include:

[0182] S32: determining a fundus vessel centerline based on the fundus vessel segmentation image obtained by the segmentation process, or obtaining the fundus vessel centerline based on the preprocessed fundus image; determining a second set of fundus vessel morphological parameters or a second set of morphological parameters corresponding to any target subregion within the ROI based on the fundus vessel centerline;

[0183] The second group of morphological parameters includes any multiple of fundus vessel fractal dimension, fundus vessel length, fundus vessel density, and fundus vessel spacing.

[0184] In some embodiments, step S32 specifically includes:

[0185] Determine the fundus artery image and the fundus venous image from the fundus blood vessel segmentation image according to the fundus blood vessel segmentation image obtained by the segmentation process; or directly identify and determine the fundus artery image and the fundus venous image based on the preprocessed fundus image;

[0186] Determine a fundus artery centerline based on the fundus artery image, or obtain the fundus artery centerline based on the preprocessed fundus image, and determine a second set of morphological parameters associated with the fundus artery based on the fundus artery centerline, or determine the second set of morphological parameters associated with the fundus artery corresponding to any target sub-region within the ROI;

[0187] The fundus venous centerline is determined based on the fundus venous image, or the fundus venous centerline is obtained based on the preprocessed fundus image, and a second set of morphological parameters associated with the fundus venous vessels is determined based on the fundus venous centerline, or the second set of morphological parameters associated with the fundus venous vessels corresponding to any target sub-region within the ROI.

[0188] In some embodiments, the second set of morphological parameters of the fundus vessels determined according to the centerline of the fundus vessels in step S32 or the second set of morphological parameters corresponding to any target sub-region within the ROI specifically include:

[0189] S321: extracting a fundus vessel segmentation image within a preset area around the optic disc or macula from the fundus vessel segmentation image, taking the optic disc or macula as a reference; determining a fundus vessel centerline within the preset area based on the fundus vessel segmentation image within the preset area around the optic disc or macula, or obtaining the fundus vessel centerline within the preset area based on a preprocessed fundus image within the preset area; determining a second set of morphological parameters corresponding to the preset area based on the fundus vessel centerline within the preset area; or,

[0190] S322: Determine a selected sub-region including multiple vascular branch points based on the multiple vascular branch points selected by the user; determine the fundus vascular centerline within the selected sub-region based on the fundus vascular segmentation image within the selected sub-region, and determine a second set of morphological parameters corresponding to the selected sub-region based on the fundus vascular centerline within the selected sub-region.

[0191] In some embodiments, step S321 may specifically include:

[0192] Taking the optic disc or macula as a reference, extracting the fundus arterial vessel image and the fundus venous vessel image in a preset area around the optic disc or macula from the fundus vessel segmentation image;

[0193] Determining a centerline of the fundus artery according to an image of the fundus artery in a preset area around the optic disc or macula, or obtaining the centerline of the fundus artery in the preset area based on a preprocessed fundus image; determining a second set of parameters associated with the fundus artery corresponding to the preset area according to the centerline of the fundus artery in the preset area;

[0194] Determining a centerline of the fundus veins based on an image of the fundus veins in a preset area around the optic disc or macula, or obtaining the centerline of the fundus veins in the preset area based on a preprocessed fundus image; determining a second set of parameters associated with the fundus veins corresponding to the preset area based on the centerline of the fundus veins in the preset area;

[0195] In some embodiments, step S322 may specifically include:

[0196] Determining, based on the plurality of vascular branch points selected by the user, a selected subregion including the plurality of vascular branch points;

[0197] extracting a fundus arterial vessel image and a fundus venous vessel image from the selected subregion;

[0198] Determining a centerline of the fundus artery within the selected subregion based on the fundus artery image extracted from the selected subregion, and determining a second set of morphological parameters associated with the fundus artery corresponding to the selected subregion based on the centerline of the fundus artery within the selected subregion;

[0199] Based on the fundus venous vessel image extracted from the selected sub-region, the centerline of the fundus venous vessel in the selected sub-region is determined, and based on the centerline of the fundus venous vessel in the selected sub-region, a second set of morphological parameters associated with the fundus venous vessel corresponding to the selected sub-region is determined.

[0200] Specifically, the method for determining the proportion of fundus blood vessel length is as follows:

[0201] Obtaining the sum of the lengths of all fundus blood vessels in the fundus blood vessel segmentation image as the total length of the first fundus blood vessel;

[0202] Obtaining the sum of the lengths of all fundus blood vessels in the preset area as the total length of the second fundus blood vessel;

[0203] According to the ratio of the total length of the second fundus blood vessels to the total length of the first fundus blood vessels, the fundus blood vessel length ratio corresponding to the fundus blood vessels in the preset area is determined.

[0204] Specifically, the method for determining the proportion of fundus blood vessel length corresponding to the arterial blood vessel is as follows:

[0205] Obtaining the sum of the lengths of all fundus arterial vessels in the fundus vessel segmentation image as the total length of the first arterial vessel;

[0206] Obtaining the sum of the lengths of all fundus arteries in the preset area as the total length of the second artery;

[0207] According to the ratio of the total length of the second arterial blood vessel to the total length of the first arterial blood vessel, the proportion of the fundus blood vessel length corresponding to the arterial blood vessels in the preset area is determined.

[0208] Specifically, the method for determining the proportion of fundus blood vessel length corresponding to the venous vessels is as follows:

[0209] Obtaining the sum of the lengths of all fundus veins in the fundus vessel segmentation image as the total length of the first vein;

[0210] Obtaining the sum of the lengths of all fundus veins in the preset area as the total length of the second vein;

[0211] According to the ratio of the total length of the second vein to the total length of the first vein, the proportion of the fundus blood vessel length corresponding to the veins in the preset area is determined.

[0212] In a further embodiment, determining the fractal dimension of the fundus blood vessels based on the fundus blood vessel segmentation image obtained by the segmentation process specifically includes:

[0213] (1) First, the fundus image is gridded, and multiple grids with the same side length (εi) are formed on the fundus image each time, where i is a positive integer ranging from 1 to n;

[0214] (2) Detect the number of grids (Ni) that intersect with the fundus blood vessels among multiple grids with the same side length (εi);

[0215] (3) Repeat step (2) until the number of grids (Nn) corresponding to the grid with the side length (εn) is obtained;

[0216] (4) Then, the number of grids (Ni) that intersect with the fundus blood vessels obtained in each operation and the inverse of the side length of the grid corresponding to this operation (εi) are fitted with a straight line under logarithmic conditions. The slope of the straight line is the box counting dimension.

[0217] Specifically, this step calculates the ratio of the area of ​​the fundus vessels to the fundus area, thereby obtaining the vessel density. This embodiment of the present invention can calculate the area of ​​any region, as well as the vessel density within that region. Obtaining the vessel density depends on the region area, which can be calculated using existing operators.

[0218] In some embodiments, the fundus blood vessel density is calculated as follows:

[0219] (1) Extracting the region of interest (ROI) of the fundus image;

[0220] (2) performing threshold segmentation on the pre-processed fundus image to obtain the fundus vascular area;

[0221] (3) Calculate the area of ​​fundus vascular region;

[0222] (4) Calculate the area of ​​the region of interest (ROI);

[0223] (5) The blood vessel surface density is obtained according to the following formula:

[0224] Blood vessel surface density = area of ​​fundus blood vessels / area of ​​region of interest (ROI).

[0225] Preset area blood vessel density = preset area fundus blood vessel area / preset area fundus area;

[0226] Arterial blood vessel surface density in the preset area = fundus arterial blood vessel area in the preset area / fundus area in the preset area;

[0227] The venous blood vessel surface density of the preset area=the fundus venous blood vessel area of ​​the preset area / the fundus area of ​​the preset area.

[0228] In some embodiments, the fundus blood vessel density is calculated as follows:

[0229] (1) Extracting the region of interest (ROI) of the fundus image;

[0230] (2) performing threshold segmentation on the pre-processed fundus image to obtain the fundus vascular area;

[0231] (3) Separate the fundus vascular area to obtain each independent blood vessel;

[0232] (4) Perform morphological operations on each blood vessel to obtain the blood vessel skeleton line, which is recorded as the blood vessel line;

[0233] (5) Calculate the length of each blood vessel line and then obtain the total length of all blood vessel lines in the ROI area;

[0234] (6) Calculate the area of ​​the region of interest (ROI);

[0235] (7) The vascular linear density is obtained according to the following formula:

[0236] Vessel line density = total length of all vessel lines in the ROI region / area of ​​the region of interest (ROI).

[0237] The blood vessel line density of the preset area = the total length of the fundus blood vessels in the preset area / the fundus area of ​​the preset area;

[0238] Arterial blood vessel linear density in the preset area = total length of fundus arterial blood vessels in the preset area / fundus area in the preset area;

[0239] The venous blood vessel linear density in the preset area = the total length of the fundus venous blood vessels in the preset area / the fundus area in the preset area.

[0240] In some embodiments, the length of a fundus blood vessel is calculated as follows:

[0241] (1) Extract the centerline of the fundus blood vessel;

[0242] (2) Smoothing the obtained fundus blood vessel centerline;

[0243] (3) Calculate the length of the smoothed center line, which is the length of the current fundus blood vessel.

[0244] This embodiment of the present invention automatically identifies and segments retinal blood vessels, distinguishing between arteries and veins. It then calculates metrics such as fractal dimension, vascular density, and vessel length to comprehensively reflect vascular morphology. This embodiment uses digital language to comprehensively describe vascular morphology from various angles, enabling a detailed understanding of fundus vascular morphology and facilitating the precise assessment of abnormalities and progression of changes in fundus vascular structure.

[0245] Example 2

[0246] Figure 8 FIG. 1 is a block diagram of a device for evaluating fundus vascular morphology according to an embodiment of the present invention. Figure 8 As shown, the apparatus 400 includes:

[0247] The fundus image preprocessing module 410 is used to preprocess the fundus image to obtain a preprocessed fundus image;

[0248] The fundus blood vessel segmentation processing module 420 is used to perform blood vessel segmentation processing on the pre-processed fundus image to obtain a fundus blood vessel segmentation image;

[0249] The fundus vascular morphological characteristic parameter determination module 430 is configured to determine a plurality of parameters describing the fundus vascular morphological characteristics based on the fundus vascular segmentation image obtained through the segmentation process.

[0250] In some embodiments, the fundus image pre-processing module 410 is specifically configured to determine a region of interest (ROI) based on the fundus image;

[0251] The fundus vascular morphological characteristic parameter determination module 430 specifically includes:

[0252] A first group of morphological parameter determination submodule 432 is configured to determine a first group of morphological parameters of the fundus vessels or a first group of morphological parameters corresponding to any target subregion within the ROI based on the fundus vessel segmentation image obtained by the segmentation process;

[0253] The first group of morphological parameters includes any number of fundus vessel fractal dimension, fundus vessel area, fundus vessel density, fundus vessel area ratio, and fundus vessel spacing.

[0254] In some embodiments, the first set of morphological parameter determination submodule 432 is specifically configured to:

[0255] Determine the fundus artery image and the fundus vein image from the fundus blood vessel segmentation image obtained by the segmentation process; or directly identify and determine the fundus artery image and the fundus vein image based on the preprocessed fundus image;

[0256] Determining a first set of morphological parameters associated with the fundus arteries according to the fundus artery image, or a first set of morphological parameters associated with the fundus arteries corresponding to any target sub-region within the ROI;

[0257] A first set of morphological parameters associated with the fundus venous vessels is determined according to the fundus venous vessel image, or a first set of morphological parameters associated with the fundus venous vessels corresponding to any target sub-region within the ROI.

[0258] In some embodiments, the first set of morphological parameter determination submodule 432 specifically includes:

[0259] a first preset area morphological parameter determination unit, configured to extract, from the fundus vascular segmentation image, a segmented image of the fundus vascular within a preset area around the optic disc or macula, taking the optic disc or macula as a reference, and determine a first set of morphological parameters corresponding to the preset area based on the segmented image of the fundus vascular within the preset area; or

[0260] The first selected sub-region morphological parameter determining unit is configured to determine a first set of morphological parameters corresponding to the selected sub-region including the multiple blood vessel branch points according to the multiple blood vessel branch points selected by the user.

[0261] In some embodiments, the first preset area morphological parameter determination unit is specifically configured to:

[0262] Taking the optic disc or macula as a reference, extracting fundus arterial and venous images in a preset area around the optic disc or macula from the fundus vessel segmentation image;

[0263] Determining a first set of morphological parameters associated with the fundus arteries corresponding to the preset area based on an image of the fundus arteries within a preset area around the optic disc or macula;

[0264] Determining a first set of morphological parameters associated with the fundus venous vessels corresponding to the preset area based on an image of the fundus venous vessels in a preset area around the optic disc or macula;

[0265] In some embodiments, the first selected sub-region morphological parameter determination unit is specifically configured to:

[0266] Determining, based on the plurality of vascular branch points selected by the user, a selected subregion including the plurality of vascular branch points;

[0267] extracting fundus arterial vessel images and fundus venous vessel images from the selected sub-region;

[0268] Determining a first set of morphological parameters associated with the fundus arteries corresponding to the selected subregion based on the fundus artery image extracted from the selected subregion;

[0269] According to the fundus venous vessel image extracted from the selected sub-region, a first group of morphological parameters associated with the fundus venous vessels corresponding to the selected sub-region is determined.

[0270] The method for determining the proportion of fundus blood vessel area is as follows:

[0271] Obtaining the sum of the areas of all fundus blood vessels in the fundus blood vessel segmentation image as the first fundus blood vessel total area;

[0272] Obtaining the sum of the areas of the fundus blood vessels in the fundus blood vessel segmentation image within a preset area around the optic disc or macula as the second fundus blood vessel total area;

[0273] The ratio of the total area of ​​the second fundus vessels to the total area of ​​the first fundus vessels is determined as the proportion of the fundus vessels in the preset area around the optic disc or macula. The method for determining the proportion of the fundus arterial vessels is as follows:

[0274] Obtaining the sum of the areas of all fundus arteries in the fundus vessel segmentation image as the first fundus artery total area;

[0275] Obtaining the sum of the areas of the fundus arteries in the fundus artery image within a preset area around the optic disc or macula as the second total fundus artery area;

[0276] The ratio of the total area of ​​the second fundus artery to the total area of ​​the first fundus artery is determined as the proportion of the fundus artery area in the preset area around the optic disc or macula.

[0277] The method for determining the proportion of fundus venous area is as follows:

[0278] Obtaining the sum of the areas of all fundus veins in the fundus vessel segmentation image as the first fundus vein total area;

[0279] Obtaining the sum of the areas of the fundus venous vessels in the fundus venous vessel image within a preset area around the optic disc or macula as the second total fundus venous vessel area;

[0280] The ratio of the total area of ​​the second fundus veins to the total area of ​​the first fundus veins is determined as the proportion of the fundus vein area in the preset area around the optic disc or macula.

[0281] In some embodiments, the fundus vessel morphological characteristic parameter determination module 430 further includes a second group of morphological parameter determination submodule 434, which is specifically configured to: determine the fundus vessel centerline based on the fundus vessel segmentation image obtained by the segmentation process, or obtain the fundus vessel centerline based on the preprocessed fundus image; determine the second group of morphological parameters or the second group of morphological parameters corresponding to any target subregion within the ROI based on the fundus vessel centerline;

[0282] The second group of morphological parameters includes any multiple of fundus vessel fractal dimension, fundus vessel length, fundus vessel density, and fundus vessel spacing.

[0283] In some embodiments, the second set of morphological parameter determination submodule 434 is specifically configured to:

[0284] Determine the fundus artery image and the fundus venous image from the fundus blood vessel segmentation image according to the fundus blood vessel segmentation image obtained by the segmentation process; or directly identify and determine the fundus artery image and the fundus venous image based on the preprocessed fundus image;

[0285] Determine a fundus artery centerline based on the fundus artery image, or obtain the fundus artery centerline based on the preprocessed fundus image, and determine a second set of morphological parameters associated with the fundus artery based on the fundus artery centerline, or determine the second set of morphological parameters associated with the fundus artery corresponding to any target sub-region within the ROI;

[0286] The fundus venous centerline is determined based on the fundus venous image, or the fundus venous centerline is obtained based on the preprocessed fundus image, and a second set of morphological parameters associated with the fundus venous vessels is determined based on the fundus venous centerline, or the second set of morphological parameters associated with the fundus venous vessels corresponding to any target sub-region within the ROI.

[0287] In some embodiments, the second set of morphological parameter determination submodule 434 specifically includes:

[0288] a second preset area morphological parameter determination unit, configured to extract, from the fundus vascular segmentation image, a fundus vascular segmentation image within a preset area around the optic disc or macula, with the optic disc or macula as a reference; determine, based on the fundus vascular segmentation image within the preset area around the optic disc or macula, a fundus vascular centerline within the preset area, or obtain, based on a preprocessed fundus image within the preset area, the fundus vascular centerline within the preset area; determine, based on the fundus vascular centerline within the preset area, a second set of morphological parameters corresponding to the preset area; or

[0289] The second selected sub-region morphological parameter determination unit is used to determine a selected sub-region including multiple vascular branch points based on the multiple vascular branch points selected by the user; determine the fundus vascular centerline within the selected sub-region based on the fundus vascular segmentation image within the selected sub-region, and determine the second set of morphological parameters corresponding to the selected sub-region based on the fundus vascular centerline within the selected sub-region.

[0290] In some embodiments, the second preset area morphological parameter determination unit is specifically configured to:

[0291] Taking the optic disc or macula as a reference, extracting fundus arterial and venous images in a preset area around the optic disc or macula from the fundus vessel segmentation image;

[0292] Determining a centerline of the fundus artery based on an image of the fundus artery within a preset area around the optic disc or macula, or obtaining the centerline of the fundus artery within the preset area based on a preprocessed fundus image; determining a second set of parameters associated with the fundus artery corresponding to the preset area based on the centerline of the fundus artery within the preset area;

[0293] Determining a fundus venous centerline based on a fundus venous image within a preset area around the optic disc or macula, or obtaining the fundus venous centerline within the preset area based on a preprocessed fundus image; determining a second set of parameters associated with the fundus venous vessels corresponding to the preset area based on the fundus venous centerline within the preset area;

[0294] In some embodiments, the second selected sub-region morphological parameter determination unit is specifically configured to:

[0295] Determining, based on the plurality of vascular branch points selected by the user, a selected subregion including the plurality of vascular branch points;

[0296] extracting fundus arterial vessel images and fundus venous vessel images from the selected sub-region;

[0297] Determining a centerline of the fundus artery in the selected subregion based on the fundus artery image extracted from the selected subregion, and determining a second set of morphological parameters associated with the fundus artery corresponding to the selected subregion based on the centerline of the fundus artery in the selected subregion;

[0298] Based on the fundus venous vessel image extracted from the selected sub-region, the centerline of the fundus venous vessel in the selected sub-region is determined, and based on the centerline of the fundus venous vessel in the selected sub-region, a second set of morphological parameters associated with the fundus venous vessel corresponding to the selected sub-region is determined.

[0299] In some embodiments, the fundus image preprocessing module 410 can be specifically used to: extract the region of interest ROI of the fundus image; perform denoising on the ROI; perform normalization on the denoised image; perform enhancement on the normalized image to obtain a preprocessed fundus image.

[0300] In some embodiments, the fundus blood vessel segmentation processing module 420 may be specifically configured to:

[0301] The pre-processed fundus image is threshold segmented using the color and morphological features of the fundus blood vessels to obtain an initial fundus blood vessel segmentation image, which is then corrected to obtain a final sample image.

[0302] Based on the final sample image, a semantic segmentation network is used for model training. The trained model is used for forward processing to output a confidence probability map of the same size as the final sample image.

[0303] According to the set threshold, the confidence probability map is converted into a binary image;

[0304] According to the binary image, the fundus vascular area and the fundus vascular segmentation image are obtained.

[0305] In some embodiments, the fundus blood vessel density is determined based on the following method:

[0306] Extract the region of interest (ROI) of the fundus image;

[0307] Performing threshold segmentation on the pre-processed fundus image to obtain the fundus vascular area;

[0308] Calculate the area of ​​fundus vascular region;

[0309] Calculate the area of ​​the region of interest (ROI);

[0310] The blood vessel surface density is obtained according to the following formula: blood vessel surface density = area of ​​fundus blood vessel region / area of ​​region of interest (ROI).

[0311] Preset area blood vessel density = preset area fundus blood vessel area / preset area fundus area;

[0312] Arterial blood vessel surface density in the preset area = fundus arterial blood vessel area in the preset area / fundus area in the preset area;

[0313] The venous blood vessel surface density of the preset area=the fundus venous blood vessel area of ​​the preset area / the fundus area of ​​the preset area.

[0314] In some embodiments, the fundus blood vessel density is determined based on the following method:

[0315] Extract the region of interest (ROI) of the fundus image;

[0316] Performing threshold segmentation on the pre-processed fundus image to obtain the fundus vascular area;

[0317] Separate the fundus vascular area to obtain each independent blood vessel;

[0318] Perform morphological operations on each blood vessel to obtain the blood vessel skeleton line, which is recorded as the blood vessel line;

[0319] Calculate the length of each blood vessel line and then obtain the total length of all blood vessel lines in the region of interest (ROI);

[0320] Calculate the area of ​​the region of interest (ROI);

[0321] The vascular line density was obtained according to the following formula: vascular line density = total length of all vascular lines in the region of interest (ROI) / area of ​​the region of interest (ROI).

[0322] The blood vessel line density of the preset area = the total length of the fundus blood vessels in the preset area / the fundus area of ​​the preset area;

[0323] Arterial blood vessel linear density in the preset area = total length of fundus arterial blood vessels in the preset area / fundus area in the preset area;

[0324] The venous blood vessel linear density in the preset area = the total length of the fundus venous blood vessels in the preset area / the fundus area in the preset area.

[0325] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0326] Example 3

[0327] like Figure 9 As shown, an embodiment of the present invention further provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, any of the above-mentioned methods for evaluating fundus vascular morphological characteristics is implemented.

[0328] If the integrated module / unit is implemented in the form of 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 present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. Of course, there are other ways of readable storage media, such as quantum memory, graphene memory, etc. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practices in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practices, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0329] Example 4

[0330] Figure 10 This is a functional block diagram of a computer device according to an embodiment of the present invention. The present invention also provides a computer device, please refer to Figure 10 At the hardware level, the computer device includes a processor and, optionally, an internal bus, a network interface, and memory. The memory may include internal memory, such as high-speed random-access memory (RAM), or non-volatile memory, such as at least one disk drive. Of course, the computer device may also include other hardware required for the business.

[0331] The processor, network interface and memory can be connected to each other through an internal bus, which can be an ISA bus, a PCI bus or an EISA bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 10 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0332] The memory is used to store the program. Specifically, the program may include program code, and the program code includes computer operation instructions. The memory may include internal memory and non-volatile memory, and provides instructions and data to the processor. The processor reads the corresponding computer program from the non-volatile memory into the internal memory and then runs it. The processor executes the program stored in the memory and is specifically used to execute Figures 1 to 2. Figure 7 The illustrated embodiment discloses a method for evaluating fundus vascular morphological characteristics.

[0333] The above-mentioned method for measuring fundus vascular morphological characteristic indicators can be applied to or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above-mentioned method can be completed by hardware integrated logic circuits in the processor or by software instructions. The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be 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 various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in conjunction with the embodiments of the present invention can be directly implemented and executed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium well-known in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in the memory, and the processor reads the information in the memory and, in conjunction with its hardware, completes the steps of the above method.

[0334] Of course, in addition to software implementation, the computer device of the present invention does not exclude other implementation methods, such as logic devices or a combination of software and hardware, etc., that is to say, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices. The systems, devices, modules or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer can be, for example, a personal computer, a laptop computer, an in-vehicle human-computer interaction device, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0335] Although the present invention provides method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive means. The order of steps listed in the embodiments is only one way of executing the order of many steps and does not represent the only execution order. When an actual device or terminal product is executed, it can be executed in sequence or in parallel according to the method shown in the embodiments or the drawings (for example, in a parallel processor or multi-threaded processing environment, or even a distributed data processing environment).

[0336] The present invention is described with reference to the flowcharts and / or block diagrams of the methods, apparatus, and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the functions specified in one or more processes in the flowcharts and / or one or more blocks in the block diagrams.

[0337] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0338] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0339] It should be noted that, in the embodiments of the present invention, relational terms such as first and second, etc., are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element.

[0340] Each embodiment in this specification is described in a related manner. Similar portions between the various embodiments can be referenced to each other. Each embodiment focuses on the differences between the other embodiments. In particular, the device, computer device, and readable storage medium embodiments are generally similar to the method embodiments, so their descriptions are relatively simple. For related portions, reference can be made to the descriptions of the method embodiments.

[0341] The above description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention are included in the scope of protection of the present invention.

Claims

1. A method for evaluating fundus vascular morphology, characterized in that: The method comprises: S1: preprocessing the fundus image to obtain a preprocessed fundus image; S2: performing blood vessel segmentation processing on the pre-processed fundus image to obtain a fundus blood vessel segmentation image; S3: determining a plurality of parameters describing morphological characteristics of fundus blood vessels according to the fundus blood vessel segmentation image; In step S1, a region of interest (ROI) is determined based on the fundus image. Step S3 includes: S31: determining a first set of morphological parameters of the fundus blood vessels or a first set of morphological parameters corresponding to any target sub-region within the ROI according to the fundus blood vessel segmentation image obtained by the segmentation process; The first group of morphological parameters includes any one of fundus vascular fractal dimension, fundus vascular area, fundus vascular density, fundus vascular area ratio, and fundus vascular spacing; the fundus vascular density includes fundus vascular line density, which is obtained according to the following formula: fundus vascular line density = total length of all vascular lines in the region of interest (ROI) / area of ​​the region of interest (ROI); the formula for calculating vascular spacing is as follows: ; represents the average vascular spacing, where σ is the standard deviation of the pixel value of each grid pixel obtained by gridding the fundus vessel segmentation image with a given grid side length ε, μ is the average pixel value of all pixels in each grid obtained by gridding the fundus vessel segmentation image with a given grid side length ε, and n is the number of grids; S32: determining a fundus vessel centerline based on the fundus vessel segmentation image obtained by the segmentation process, or obtaining the fundus vessel centerline based on the preprocessed fundus image; and determining a second set of morphological parameters or a second set of morphological parameters corresponding to any target subregion within the ROI based on the fundus vessel centerline; The second group of morphological parameters includes any multiple of fundus vessel fractal dimension, fundus vessel length, fundus vessel density and fundus vessel spacing.

2. The method according to claim 1, characterized in that Step S31 specifically includes: determining a fundus artery image and a fundus vein image from the fundus blood vessel segmentation image according to the fundus blood vessel segmentation image; or directly identifying and determining the fundus artery image and the fundus vein image based on the preprocessed fundus image; Determining a first set of morphological parameters associated with the fundus artery vessels according to the fundus artery image, or a first set of morphological parameters associated with the fundus artery vessels corresponding to any target sub-region within the ROI; A first set of morphological parameters associated with the fundus venous vessels is determined according to the fundus venous image, or a first set of morphological parameters associated with the fundus venous vessels corresponding to any target sub-region within the ROI.

3. The method according to claim 1, characterized in that In step S31, based on the fundus blood vessel segmentation image obtained by the segmentation process, a first set of morphological parameters corresponding to any target sub-region within the ROI is determined, specifically including: S311: Taking the optic disc or macula as a reference, extracting a segmented image of the fundus vessels in a preset area around the optic disc or macula from the segmented image of the fundus vessels, and determining a first set of morphological parameters corresponding to the preset area based on the segmented image of the fundus vessels in the preset area; or S312: Determine, based on the multiple blood vessel branch points selected by the user, a first set of morphological parameters corresponding to the selected sub-region including the multiple blood vessel branch points.

4. The method according to claim 3, characterized in that Step S311 specifically includes: Taking the optic disc or macula as a reference, extracting the fundus arterial vessel image and the fundus venous vessel image in a preset area around the optic disc or macula from the fundus vessel segmentation image; Determining a first set of morphological parameters associated with the fundus arteries corresponding to the preset area based on the fundus artery image within the preset area around the optic disc or macula; Determining a first set of morphological parameters associated with the fundus venous vessels corresponding to the preset area based on the fundus venous vessel image within the preset area around the optic disc or macula; Step S312 specifically includes: determining, according to a plurality of blood vessel branch points selected by a user, a selected subregion including the plurality of blood vessel branch points; extracting a fundus arterial vessel image and a fundus venous vessel image from the selected subregion; Determining a first set of morphological parameters associated with the fundus arteries corresponding to the selected subregion based on the fundus artery image extracted from the selected subregion; According to the fundus venous vessel image extracted from the selected sub-region, a first group of morphological parameters associated with the fundus venous vessels corresponding to the selected sub-region is determined.

5. The method according to claim 1, wherein Step S32 specifically includes: Determining a fundus artery image and a fundus venous image from the fundus blood vessel segmentation image obtained by segmentation processing; or directly identifying and determining the fundus artery image and the fundus venous image based on the preprocessed fundus image; Determine a fundus artery centerline based on the fundus artery image, or obtain the fundus artery centerline based on the preprocessed fundus image, and determine a second set of morphological parameters associated with the fundus artery based on the fundus artery centerline, or determine the second set of morphological parameters associated with the fundus artery corresponding to any target sub-region within the ROI; The fundus venous centerline is determined based on the fundus venous image, or the fundus venous centerline is obtained based on the preprocessed fundus image, and a second set of morphological parameters associated with the fundus venous vessels is determined based on the fundus venous centerline, or the second set of morphological parameters associated with the fundus venous vessels corresponding to any target sub-region within the ROI.

6. The method according to claim 1, characterized in that In step S32, a second set of morphological parameters or a second set of morphological parameters corresponding to any target sub-region within the ROI is determined based on the fundus blood vessel centerline, specifically including: S321: extracting a fundus vessel segmentation image within a preset area around the optic disc or macula from the fundus vessel segmentation image, taking the optic disc or macula as a reference; determining a fundus vessel centerline within the preset area based on the fundus vessel segmentation image within the preset area around the optic disc or macula, or obtaining the fundus vessel centerline within the preset area based on a preprocessed fundus image within the preset area; determining a second set of morphological parameters corresponding to the preset area based on the fundus vessel centerline within the preset area; or, S322: Based on the multiple vascular branch points selected by the user, determine a selected sub-region including the multiple vascular branch points; based on the fundus vascular segmentation image within the selected sub-region, determine the fundus vascular centerline within the selected sub-region; based on the fundus vascular centerline within the selected sub-region, determine a second set of morphological parameters corresponding to the selected sub-region.

7. The method according to claim 6, characterized in that Step S321 specifically includes: Taking the optic disc or macula as a reference, extracting the fundus arterial vessel image and the fundus venous vessel image in a preset area around the optic disc or macula from the fundus vessel segmentation image; Determining a centerline of the fundus artery according to an image of the fundus artery in a preset area around the optic disc or macula, or obtaining the centerline of the fundus artery in the preset area based on a preprocessed fundus image; determining a second set of parameters associated with the fundus artery corresponding to the preset area according to the centerline of the fundus artery in the preset area; Determining a centerline of the fundus veins based on an image of the fundus veins in a preset area around the optic disc or macula, or obtaining the centerline of the fundus veins in the preset area based on a preprocessed fundus image; determining a second set of parameters associated with the fundus veins corresponding to the preset area based on the centerline of the fundus veins in the preset area; Step S322 specifically includes: determining, according to a plurality of blood vessel branch points selected by a user, a selected subregion including the plurality of blood vessel branch points; extracting a fundus arterial vessel image and a fundus venous vessel image from the selected subregion; determining a centerline of the fundus artery within the selected subregion based on the fundus artery image extracted from the selected subregion, and determining a second set of morphological parameters associated with the fundus artery corresponding to the selected subregion based on the centerline of the fundus artery within the selected subregion; Based on the fundus venous vessel image extracted from the selected sub-region, the centerline of the fundus venous vessels in the selected sub-region is determined; based on the centerline of the fundus venous vessels in the selected sub-region, a second set of morphological parameters associated with the fundus venous vessels corresponding to the selected sub-region is determined.

8. A device for evaluating fundus vascular morphology, characterized in that: The device comprises: A fundus image preprocessing module is used to preprocess the fundus image to obtain a preprocessed fundus image; a fundus blood vessel segmentation processing module, configured to perform blood vessel segmentation processing on the fundus image to obtain a fundus blood vessel segmentation image; a fundus vascular morphological characteristic parameter determination module, configured to determine a plurality of parameters describing the fundus vascular morphological characteristics based on the fundus vascular segmentation image; The fundus image preprocessing module is specifically used to determine the region of interest ROI based on the fundus image; The fundus vascular morphological characteristic parameter determination module specifically includes: The first group of morphological parameter determination submodule is used to determine the first group of fundus vascular morphological parameters or the first group of morphological parameters corresponding to any target subregion within the ROI based on the fundus vascular segmentation image obtained by the segmentation process; wherein the first group of morphological parameters includes: any multiple of the fundus vascular fractal dimension, fundus vascular area, fundus vascular density, fundus vascular area ratio, and fundus vascular spacing; the fundus vascular density includes the fundus vascular line density, which is obtained according to the following formula: fundus vascular line density = total length of all vascular lines within the region of interest (ROI) / area of ​​the region of interest (ROI); the vascular spacing is calculated as follows: ; represents the average vascular spacing, where σ is the standard deviation of the pixel value of each grid pixel obtained by gridding the fundus vessel segmentation image with a given grid side length ε, μ is the average pixel value of all pixels in each grid obtained by gridding the fundus vessel segmentation image with a given grid side length ε, and n is the number of grids; The second group of morphological parameter determination submodule is specifically used to determine the fundus vascular centerline based on the fundus vascular segmentation image obtained by the segmentation process, or to obtain the fundus vascular centerline based on the preprocessed fundus image; determine the second group of morphological parameters or the second group of morphological parameters corresponding to any target sub-region within the ROI based on the fundus vascular centerline; wherein the second group of morphological parameters includes any multiple of the fundus vascular fractal dimension, fundus vascular length, fundus vascular density, and fundus vascular spacing.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, a method for evaluating fundus vascular morphological characteristics as described in any one of claims 1 to 7 is implemented.

10. A computer device, characterized in that: It includes: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement a method for evaluating fundus vascular morphological characteristics as described in any one of claims 1 to 7.

Images