Identification method and device, computer device and storage medium

Abstract

The application relates to an identification method and device, computer equipment and a storage medium. The method comprises the following steps: obtaining a blood vessel image of an eye to be examined, the blood vessel image being generated by optical coherence tomography blood vessel imaging; obtaining statistical features of a macular region in the blood vessel image; and identifying a non-perfusion region in the blood vessel image based on the statistical features. The method can effectively improve the flexibility of identifying the non-perfusion region.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a recognition method, apparatus, computer device, and storage medium. Background Technology

[0002] Non-perfusion areas are ischemic areas in the fundus. They are caused by the complete occlusion of capillaries, which stops blood flow to certain areas of the retina. Therefore, there is a need to identify non-perfusion areas.

[0003] Traditional techniques typically rely on retinal contrast agents to identify non-perfused areas; however, this method is not very flexible. Summary of the Invention

[0004] Therefore, it is necessary to provide a more flexible identification method, device, computer equipment, and storage medium to address the aforementioned technical problems.

[0005] Firstly, this application provides an identification method. The method includes:

[0006] A vascular image of the examined eye is acquired, generated by optical coherence tomography angiography; statistical features of the macular region in the vascular image are obtained, and non-perfusion areas in the vascular image are identified based on these statistical features.

[0007] In one embodiment, the statistical feature includes at least two of the following: the average value of the pixels in the macula region, the median value of the pixels in the macula region, the standard deviation of the pixels in the macula region, and the squared difference of the pixels in the macula region.

[0008] In one embodiment, identifying the non-perfusion region in the vascular image based on the statistical features includes: determining a statistical feature threshold based on the statistical features of the macular region; performing binarization processing on the vascular image based on the statistical feature threshold to obtain a non-perfusion region mask; and determining the non-perfusion region based on the non-perfusion region mask.

[0009] In one embodiment, after binarizing the blood vessel image according to the statistical feature threshold to obtain a non-perfusion area mask, the method further includes: performing expansion and erosion processing on the non-perfusion area mask.

[0010] In one embodiment, the method further includes: receiving a selected region in the vascular image, and if the selected region contains the non-perfusion region, marking the non-perfusion region in the selected region and displaying the marked non-perfusion region.

[0011] In one embodiment, the method further includes: determining the number of pixels in the marked non-inflation region; and determining area information of the marked non-inflation region based on the number of pixels.

[0012] In one embodiment, before binarizing the blood vessel image according to the statistical feature threshold, the method further includes: performing convolution processing on the blood vessel image to obtain multiple mapping values ​​of the blood vessel image; and performing binarization processing on the blood vessel image based on the multiple mapping values ​​and the statistical feature threshold.

[0013] In one embodiment, before acquiring the statistical features of the macular region in the vascular image, the method further includes: acquiring an initial macular region in the vascular image; calculating the area of ​​the initial macular region; displaying a prompt message for the initial macular region if the area of ​​the initial macular region is greater than a preset area threshold; receiving a target operation triggered by a user based on the prompt message for the initial macular region; determining the macular region based on the target operation and the initial macular region, and calculating the statistical features of the macular region.

[0014] Secondly, this application also provides an identification device. The device includes:

[0015] The acquisition module is used to acquire vascular images of the examined eye, which are generated by optical coherence tomography angiography.

[0016] The first execution module is used to obtain statistical features of the macular region in the vascular image and identify non-perfusion areas in the vascular image based on the statistical features.

[0017] In one embodiment, the statistical feature includes at least two of the following: the average value of the pixels in the macula region, the median value of the pixels in the macula region, the standard deviation of the pixels in the macula region, and the squared difference of the pixels in the macula region.

[0018] In one embodiment, the first execution module is specifically used to: determine a statistical feature threshold based on the statistical features of the macular region; perform binarization processing on the blood vessel image based on the statistical feature threshold to obtain a non-perfusion area mask; and determine the non-perfusion area based on the non-perfusion area mask.

[0019] In one embodiment, the first execution module is specifically used to: expand and etch the mask in the non-injection area.

[0020] In one embodiment, the identification device further includes a second execution module, which is configured to: receive a selected region in the vascular image; if the selected region contains the non-perfusion region, mark the non-perfusion region in the selected region and display the marked non-perfusion region.

[0021] In one embodiment, the identification device further includes a third execution module, which is configured to: determine the number of pixels in the marked non-inflation region; and determine the area information of the marked non-inflation region based on the number of pixels.

[0022] In one embodiment, the first execution module is specifically used to: perform convolution processing on the blood vessel image to obtain multiple mapping values ​​of the blood vessel image; and perform binarization processing on the blood vessel image based on the multiple mapping values ​​and the statistical feature threshold.

[0023] In one embodiment, the recognition device further includes a fourth execution module, which is configured to: acquire an initial macular region in the blood vessel image; calculate the area of ​​the initial macular region; display a prompt message for the initial macular region if the area of ​​the initial macular region is greater than a preset area threshold; receive a target operation triggered by the user based on the prompt message for the initial macular region; determine the macular region based on the target operation and the initial macular region, and calculate the statistical characteristics of the macular region.

[0024] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform any of the steps described in the first aspect above.

[0025] The aforementioned identification method, apparatus, and computer equipment acquire vascular images of the examined eye, generated by optical coherence tomography (OCT) angiography. Statistical features of the macular region within these vascular images are obtained, and non-perfusion regions are identified based on these statistical features. The identification method provided in this application leverages the fact that neither the macular region nor the non-perfusion region contains blood vessels and shares the same image features. It utilizes the statistical features of the macular region in the vascular image to identify non-perfusion regions. Furthermore, this method eliminates the need for fundus contrast agents and effectively improves the flexibility of identifying non-perfusion regions. Attached Figure Description

[0026] Figure 1 This is a flowchart illustrating the identification method in one embodiment;

[0027] Figure 2 This is a flowchart illustrating a method for identifying non-perfusion regions in a blood vessel image based on this statistical feature in one embodiment.

[0028] Figure 3 This is a flowchart illustrating the steps prior to binarizing the blood vessel image based on the statistical feature threshold in one embodiment.

[0029] Figure 4This is a flowchart illustrating the steps following the identification of non-perfusion regions in a vascular image using statistical features in one embodiment.

[0030] Figure 5 This is a flowchart illustrating the identification method in another embodiment;

[0031] Figure 6 This is a structural block diagram of the identification device in one embodiment;

[0032] Figure 7 This is a structural block diagram of the identification device in another embodiment;

[0033] Figure 8 This is an internal structural diagram of a computer device in one embodiment;

[0034] Figure 9 This is a schematic diagram of the inner retinal blood vessels in one embodiment;

[0035] Figure 10 This is a schematic diagram of the macular region in a blood vessel image in one embodiment;

[0036] Figure 11 This is a schematic diagram of a non-perfusion region in a blood vessel image in one embodiment;

[0037] Figure 12 This is a schematic diagram of a non-inflated area within a selected region in one embodiment;

[0038] Figure 13 This is a flowchart illustrating a method for determining the macular region in one embodiment. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0040] Non-perfusion areas are ischemic areas in the fundus. They are caused by the complete occlusion of capillaries, which stops blood flow to certain areas of the retina. Therefore, there is a need to identify non-perfusion areas.

[0041] Traditional techniques typically rely on retinal contrast agents to identify non-perfused areas; however, this method is not very flexible.

[0042] In view of this, the embodiments of this application provide a highly flexible identification method.

[0043] The identification method provided in this application embodiment can be executed by a computer device, which can be a server.

[0044] In one embodiment, such as Figure 1 As shown, an identification method is provided, including the following steps:

[0045] Step 101: Obtain a vascular image of the eye being examined, which is generated by optical coherence tomography angiography.

[0046] Optionally, Optical Coherence Tomography Angiography (OCTA) is a non-contact, fast-scanning, and high-resolution vascular imaging technique. It compares cross-sectional images of multiple structural OCT images of the same imaging location of a sample to detect motion differences in backscattered light, reflecting the movement of red blood cells in the blood vessels, thereby enabling non-invasive imaging of the microvascular structures of the retina and choroid.

[0047] Optionally, the vascular image can be an image of the inner retinal vessels or an image of the superficial vascular complex.

[0048] In one possible implementation, taking the vascular image as an example of an inner retinal image, an Optical Coherence Tomography (OCT) scanner includes a light source and an optical probe. Based on the measurement light emitted by the light source, the examined eye is repeatedly scanned to obtain at least two sets of scan data. Optical Coherence Tomography Angiography (OCTA) then determines the blood flow motion data in the retina based on the at least two sets of data obtained from the repeated scans. The blood flow motion data is processed to obtain projection data, which is the vascular image. This can be illustrated as follows: Figure 9 As shown, Figure 9 Image of the blood vessels in the inner layer of the retina.

[0049] Step 102: Obtain the statistical features of the macular region in the vascular image, and identify the non-perfusion region in the vascular image based on the statistical features.

[0050] Optionally, the macula is located in the center of the retina, which is the most sensitive area for vision, and the cone cells responsible for vision and color perception are distributed in this area.

[0051] Optionally, the non-perfusion area refers to the ischemic area in the fundus.

[0052] In one possible implementation, the non-perfusion area is created due to complete capillary occlusion, which stops blood flow to certain areas of the retina.

[0053] Optionally, the statistical feature can be an image feature of the macula region or a noise feature of the macula region.

[0054] In one possible implementation, since neither the macula nor the non-perfusion region has blood vessels, their image features are identical, mainly consisting of noise. Therefore, the non-perfusion region in the blood vessel image can be identified based on the statistical features of the macula in the blood vessel image. Moreover, since the statistical features of the macula are determined by numerous pixels, slight increases or decreases in the number of pixels have little impact on the statistical features. Thus, using the statistical features of the macula to identify the non-perfusion region has high reliability and accuracy.

[0055] The aforementioned identification method acquires a vascular image of the examined eye, generated by optical coherence tomography (OCT) angiography. It then obtains statistical features of the macular region within this vascular image and identifies non-perfusion regions based on these features. The identification method provided in this application leverages the fact that neither the macular region nor the non-perfusion region contains blood vessels and shares similar image features. It utilizes the statistical features of the macular region in the vascular image to identify non-perfusion regions. Furthermore, this method eliminates the need for fundus contrast agents and effectively improves the flexibility of identifying non-perfusion regions.

[0056] In an optional embodiment of this application, before identifying the non-perfusion region in the vascular image based on the statistical features, the vascular image needs to be filtered and denoised.

[0057] In one possible implementation, the vascular image can be filtered and denoised based on a median filtering method.

[0058] In another possible implementation, the blood vessel image can be filtered and denoised based on a Gaussian filtering method.

[0059] In another possible implementation, the blood vessel image can be filtered and denoised based on median filtering and Gaussian filtering methods.

[0060] In one embodiment, such as Figure 2 As shown, the method for identifying non-perfusion regions in a vascular image based on these statistical features includes the following steps:

[0061] Step 201: Determine the statistical feature threshold based on the statistical characteristics of the macular region.

[0062] Optionally, the statistical feature threshold can be a specific value or it can be used to indicate a range.

[0063] In one embodiment, the statistical feature includes at least two of the following: the average value of the pixels in the macula region, the median value of the pixels in the macula region, the standard deviation of the pixels in the macula region, and the squared difference of the pixels in the macula region.

[0064] In an optional embodiment of this application, a correlation calculation is performed based on the statistical characteristics of at least two macular regions to obtain a threshold for the statistical characteristics. The correlation calculation includes, but is not limited to, multiplication and differentiation.

[0065] In one possible implementation, the average value of the pixels in the macula region and the standard deviation of the pixels in the macula region can be correlated and calculated to determine the threshold of the statistical feature.

[0066] In another possible implementation, the average value of the pixels in the macula region and the median value of the pixels in the macula region can be correlated in a calculation to determine the threshold of the statistical feature.

[0067] In another possible implementation, the average value of the pixels in the macula region and the squared difference of the pixels in the macula region can be correlated in the statistical feature calculation to determine the threshold of the statistical feature.

[0068] In another possible implementation, the median of the pixels in the macula region and the standard deviation of the pixels in the macula region can be correlated in the statistical feature calculation to determine the threshold of the statistical feature.

[0069] In another possible implementation, the median of the pixels in the macula region and the squared difference of the pixels in the macula region can be correlated in the statistical feature calculation to determine the threshold of the statistical feature.

[0070] In another possible implementation, the standard deviation of the pixels in the macula region and the squared difference of the pixels in the macula region can be correlated in the statistical feature calculation to determine the threshold of the statistical feature.

[0071] Step 202: Binarize the blood vessel image according to the statistical feature threshold to obtain a mask for the non-perfusion area.

[0072] Optionally, the binarization process refers to an image segmentation method. Binarization can set the grayscale value of pixels in an image to 0 or 255, that is, to present the entire image with a clear visual effect of only black and white.

[0073] Optionally, the mask refers to using a selected image, graphic, or object to occlude (all or part) the image being processed, thereby controlling the area or process of image processing.

[0074] In one possible implementation, the vascular image is binarized according to the statistical feature threshold, setting the grayscale value of pixels in the non-perfusion region to 0, and the grayscale value of pixels in the non-perfusion region to 255.

[0075] In another possible implementation, the vascular image is binarized according to the statistical feature threshold, so that the gray value of the pixels in the non-perfusion area is 255, and the gray value of the pixels in the non-perfusion area is 0.

[0076] Step 203: Determine the non-inflation area based on the non-inflation area mask.

[0077] In an optional embodiment of this application, since the statistical feature threshold is determined based on the macular region, when binarizing the vascular image based on this statistical feature threshold, the macular region may also be identified as a non-perfusion region. Therefore, after obtaining the non-perfusion region mask, the macular region needs to be removed, and then the non-perfusion region is determined based on the non-perfusion region mask with the macular region removed, such as... Figure 11 As shown, Figure 11 The gray area represents the identified non-inflation areas.

[0078] The method described above, which determines a statistical feature threshold based on the statistical characteristics of the macular region, then binarizes the vascular image based on the statistical feature threshold to obtain a non-perfusion area mask, and finally determines the non-perfusion area based on the non-perfusion area mask, can identify the non-perfusion area without the use of contrast agents, effectively improving the flexibility of non-perfusion area identification. Moreover, by using the statistical features of at least two macular regions to obtain the statistical feature threshold, the reliability and accuracy of non-perfusion area identification are effectively improved.

[0079] In one embodiment, such as Figure 3 As shown, before binarizing the blood vessel image according to the statistical feature threshold, the method further includes the following steps:

[0080] Step 301: Perform convolution processing on the blood vessel image to obtain multiple mapping values ​​of the blood vessel image.

[0081] Optionally, the size of the target convolution kernel and the target stride used in the convolution process can be preset by the user. For example, the size of the target convolution kernel can be 11×11, 9×9, or 13×13.

[0082] In one possible implementation, if the statistical feature threshold is calculated by correlating the average pixel value and the standard deviation of the pixels in the macular region, then multiple first mapping values ​​of the blood vessel image can be determined firstly based on the convolution process. For example, first mapping values ​​for region A, region B, and region C. These multiple first mapping values ​​refer to the average pixel values ​​corresponding to multiple regions of the blood vessel image obtained by performing convolution processing on the blood vessel image based on the target convolution kernel and the target stride. Then, multiple second mapping values ​​of the blood vessel image are determined based on the convolution process, such as the second mapping value for region A. The second mapping value for region B, the second mapping value for region C, and these multiple second mapping values ​​refer to the pixel standard deviations corresponding to multiple regions of the blood vessel image obtained by performing convolution processing on the blood vessel image based on the target convolution kernel and the target stride. These multiple first mapping values ​​and these multiple second mapping values ​​are regionally corresponding. For example, the first mapping value for region A and the second mapping value for region A are regionally corresponding. The multiple first mapping values ​​and the multiple second mapping values ​​are correlated and calculated to obtain multiple mapping values. For example, one of the multiple mapping values ​​can be obtained by correlating and calculating the first mapping value for region A and the second mapping value for region A.

[0083] In another possible implementation, if the statistical feature threshold is calculated by correlating the average pixel value of the macular region with the squared difference of the pixels in the macular region, then multiple first mapping values ​​of the blood vessel image can be determined first based on the convolution process, such as the first mapping value of region A, the first mapping value of region B, and the first mapping value of region C. These multiple first mapping values ​​refer to the average pixel values ​​corresponding to multiple regions of the blood vessel image obtained by performing convolution processing on the blood vessel image based on the target convolution kernel and the target stride. Then, multiple second mapping values ​​of the blood vessel image are determined based on the convolution process, such as the second mapping value of region A. The second mapping value for region B, the second mapping value for region C, and the multiple second mapping values ​​refer to the pixel squared differences corresponding to multiple regions of the blood vessel image obtained by convolution processing based on the target convolution kernel and the target stride. The multiple first mapping values ​​and the multiple second mapping values ​​are regionally corresponding. For example, the first mapping value for region A and the second mapping value for region A are regionally corresponding. The multiple first mapping values ​​and the multiple second mapping values ​​are correlated and calculated to obtain multiple mapping values. For example, one of the multiple mapping values ​​can be obtained by correlating and calculating the first mapping value for region A and the second mapping value for region A.

[0084] In another possible implementation, if the statistical feature threshold is calculated by correlating the average pixel value and the median pixel value in the macular region, then multiple first mapping values ​​of the blood vessel image can be determined first based on the convolution process. For example, first mapping values ​​for region A, region B, and region C. These multiple first mapping values ​​refer to the average pixel values ​​corresponding to multiple regions of the blood vessel image obtained by performing convolution processing on the blood vessel image based on the target convolution kernel and the target stride. Then, multiple second mapping values ​​of the blood vessel image are determined based on the convolution process, such as the second mapping value for region A. The second mapping value for region B, the second mapping value for region C, and these multiple second mapping values ​​refer to the pixel median values ​​corresponding to multiple regions of the blood vessel image obtained by performing convolution processing on the blood vessel image based on the target convolution kernel and the target stride. These multiple first mapping values ​​and these multiple second mapping values ​​are region-corresponding. For example, the first mapping value for region A and the second mapping value for region A are region-corresponding. Multiple mapping values ​​are obtained by performing association calculation on the corresponding multiple first mapping values ​​and the multiple second mapping values. For example, one of the multiple mapping values ​​can be obtained by performing association calculation on the first mapping value for region A and the second mapping value for region A.

[0085] In another possible implementation, if the statistical feature threshold is calculated by correlating the median of pixels in the macular region with the standard deviation of pixels in the macular region, then multiple first mapping values ​​of the blood vessel image can be determined firstly based on the convolution process. For example, first mapping values ​​for region A, region B, and region C. These multiple first mapping values ​​refer to the median of pixels corresponding to multiple regions of the blood vessel image obtained by performing convolution processing on the blood vessel image based on the target convolution kernel and the target stride. Then, multiple second mapping values ​​of the blood vessel image are determined based on the convolution process, such as the second mapping value for region A. The second mapping value for region B, the second mapping value for region C, and these multiple second mapping values ​​refer to the pixel standard deviations corresponding to multiple regions of the blood vessel image obtained by performing convolution processing on the blood vessel image based on the target convolution kernel and the target stride. These multiple first mapping values ​​and these multiple second mapping values ​​are regionally corresponding. For example, the first mapping value for region A and the second mapping value for region A are regionally corresponding. The multiple first mapping values ​​and the multiple second mapping values ​​are correlated and calculated to obtain multiple mapping values. For example, one of the multiple mapping values ​​can be obtained by correlating and calculating the first mapping value for region A and the second mapping value for region A.

[0086] In another possible implementation, if the statistical feature threshold is calculated by correlating the median of pixels in the macular region with the squared difference of pixels in the macular region, then multiple first mapping values ​​of the blood vessel image can be determined firstly based on the convolution process. For example, first mapping values ​​for region A, region B, and region C. These multiple first mapping values ​​refer to the median of pixels corresponding to multiple regions of the blood vessel image obtained by performing convolution processing on the blood vessel image based on the target convolution kernel and the target stride. Then, multiple second mapping values ​​of the blood vessel image are determined based on the convolution process, such as the second mapping value for region A. The second mapping value for region B, the second mapping value for region C, and these multiple second mapping values ​​refer to the pixel squared differences corresponding to multiple regions of the blood vessel image obtained by convolution processing based on the target convolution kernel and the target stride. These multiple first mapping values ​​and these multiple second mapping values ​​are regionally corresponding. For example, the first mapping value for region A and the second mapping value for region A are regionally corresponding. The multiple first mapping values ​​and the multiple second mapping values ​​are correlated and calculated to obtain multiple mapping values. For example, one of the multiple mapping values ​​can be obtained by correlating and calculating the first mapping value for region A and the second mapping value for region A.

[0087] In another possible implementation, if the statistical feature threshold is calculated by correlating the standard deviation of pixels in the macular region with the squared difference of pixels in the macular region, then multiple first mapping values ​​of the blood vessel image can be determined firstly based on the convolution process. For example, first mapping values ​​for region A, region B, and region C. These multiple first mapping values ​​refer to the pixel standard deviations corresponding to multiple regions of the blood vessel image obtained by performing convolution processing on the blood vessel image based on the target convolution kernel and the target stride. Then, multiple second mapping values ​​of the blood vessel image are determined based on the convolution process, such as the second mapping value for region A. The second mapping value for region B, the second mapping value for region C, and the multiple second mapping values ​​refer to the pixel squared differences corresponding to multiple regions of the blood vessel image obtained by convolution processing based on the target convolution kernel and the target stride. The multiple first mapping values ​​and the multiple second mapping values ​​are regionally corresponding. For example, the first mapping value for region A and the second mapping value for region A are regionally corresponding. The multiple first mapping values ​​and the multiple second mapping values ​​are correlated and calculated to obtain multiple mapping values. For example, one of the multiple mapping values ​​can be obtained by correlating and calculating the first mapping value for region A and the second mapping value for region A.

[0088] It should be noted that the statistical features corresponding to the first and second mapping values ​​are determined based on the statistical features on which the statistical feature threshold is based. Since the statistical feature threshold is calculated based on the association of at least two statistical features, that is, the statistical feature threshold can also be obtained based on three or four statistical features. Therefore, there will also be a third or fourth mapping value. Specifically, it can be determined based on the statistical features on which the statistical feature threshold is based.

[0089] Step 302: Binarize the blood vessel image based on the multiple mapping values ​​and the statistical feature threshold.

[0090] In one possible implementation, assuming the statistical feature threshold is a range, it is determined whether multiple mapping values ​​fall within this range. For the regions corresponding to mapping values ​​that fall within this range, i.e., non-injection regions, the grayscale value of the pixels in these regions is set to 255. For the regions corresponding to mapping values ​​that do not fall within this range, i.e., non-non-injection regions, the grayscale value of the pixels in these regions is set to 0.

[0091] In another possible implementation, assuming that the statistical feature threshold is a range, it is determined whether multiple mapping values ​​fall within the range. For the regions corresponding to the mapping values ​​that fall within the range, i.e., the non-injection regions, the grayscale value of the pixels in the region is set to 0. For the regions corresponding to the mapping values ​​that do not fall within the range, i.e., the non-non-injection regions, the grayscale value of the pixels in the region is set to 255.

[0092] The method described above, which first performs convolution processing on the vascular image to obtain multiple mapping values ​​of the vascular image, and then performs binarization processing on the vascular image based on the multiple mapping values ​​and the statistical feature threshold, has high reliability and high accuracy in obtaining the non-perfusion region by binarizing the vascular image based on the multiple mapping values ​​and the statistical feature threshold. This is because the multiple mapping values ​​are also calculated based on the correlation of at least two statistical features of multiple regions of the vascular image.

[0093] In one embodiment, after binarizing the vascular image according to the statistical feature threshold to obtain a non-perfusion area mask, the method further includes: performing expansion and erosion processing on the non-perfusion area mask.

[0094] Optionally, the dilation process refers to a technique that expands the boundary points of a binary object by merging all background points in contact with the object into the object, thereby expanding the boundary outward.

[0095] In one possible implementation, the expansion process can be a horizontal expansion process.

[0096] In another possible implementation, the expansion process can also be a vertical expansion process.

[0097] In another possible implementation, the expansion process can also be a full-scale expansion process.

[0098] Alternatively, erosion refers to a technique that eliminates the boundary points of an object, causing the boundary to shrink inward, and can remove objects smaller than the structural element.

[0099] In one possible implementation, the corrosion treatment can be a horizontal corrosion treatment.

[0100] In another possible implementation, the corrosion treatment can also be vertical corrosion.

[0101] In another possible implementation, the corrosion treatment can also be omnidirectional corrosion.

[0102] In one embodiment, after identifying the non-perfusion region in the vascular image based on the statistical features, the method further includes: receiving a selected region in the vascular image; if the selected region contains the non-perfusion region, marking the non-perfusion region in the selected region and displaying the marked non-perfusion region.

[0103] Optionally, the marker can be a rendering process, a label process, or an arrow indicator.

[0104] In one possible implementation, the vascular image can be displayed on a screen, and the user can perform a click operation based on the vascular image. The area clicked by the user is the selected area. If it is determined that the selected area includes the non-perfusion area, then the non-perfusion area within the selected area is rendered. Figure 12 As shown, Figure 12 The gray area represents the non-injected area within the selected region.

[0105] In another possible implementation, the vascular image can be displayed on a screen, and the user can perform a click operation based on the vascular image. The area clicked by the user is the selected area. If it is determined that the selected area does not include the non-perfusion area, no response is made.

[0106] In an optional embodiment of this application, after determining the non-perfusion area of ​​the vascular image, the non-perfusion area can be directly marked and displayed on the screen for easy viewing by the user. The user can perform a click operation based on the vascular image, and the area clicked by the user is the selected area. If it is determined that the selected area includes the non-perfusion area, the non-perfusion area included in the selected area is enlarged by a preset ratio.

[0107] The method described above, which receives a selected area in the vascular image and marks the non-perfusion area within the selected area if the selected area includes the non-perfusion area, and displays the marked non-perfusion area, can more intuitively show the non-perfusion area to the user, effectively improving flexibility. Moreover, in practical applications, such as when formulating a target plan for laser photocoagulation, the above method can help technicians determine the target range. That is, technicians can select the non-perfusion area of ​​interest for display and not display other non-perfusion areas, which can effectively improve the work efficiency of technicians and facilitate technicians to quickly formulate target plans based on non-perfusion areas.

[0108] In one embodiment, such as Figure 4 As shown, after identifying the non-perfusion region in the vascular image based on this statistical feature, the method further includes the following steps:

[0109] Step 401: Determine the number of pixels in the marked non-inflation region.

[0110] In one possible implementation, the number of pixels in the marked uninfused region can be determined based on the attribute information of the marked uninfused region.

[0111] Step 402: Determine the area information of the marked non-inflation region based on the number of pixels.

[0112] In one possible implementation, the area information of the marked non-inflation region is determined based on the number of pixels and the physical area of ​​the image pixels.

[0113] In an optional embodiment of this application, after determining the area information of the marked non-inflation area based on the number of pixels, the area information of the non-inflation area can be displayed when the area selected by the user includes the non-inflation area.

[0114] In one embodiment, such as Figure 13 As shown, before obtaining the statistical features of the macular region in the vascular image, the method further includes:

[0115] Step 701: Obtain the initial macular region in the blood vessel image.

[0116] In one possible implementation, the macular region in the blood vessel image can be obtained based on a macular region neural network recognition algorithm, or it can be obtained by delineating the boundaries of the macular region. Figure 10 As shown, Figure 10 The black area within the Chinese box represents the macular region in the identified blood vessel image.

[0117] Step 702: Calculate the area of ​​the initial macular region.

[0118] In one possible implementation, the number of pixels in the initial macula region can be obtained first, and then the area of ​​the initial macula region can be determined based on the number of pixels and the physical area of ​​the image pixels.

[0119] Step 703: If the area of ​​the initial macula is greater than a preset area threshold, display a prompt message for the initial macula.

[0120] Optionally, the preset area threshold can be set by a technician. The preset area threshold can be the area of ​​the macula of a normal human eye, which is 0.38±0.11mm2, or it can be customized by a technician.

[0121] Optionally, the prompt information can be the area information of the initial macula region, or it can be the boundary information of the initial macula region.

[0122] Step 704: Receive the target operation triggered by the user based on the initial macular region prompt information.

[0123] In one possible implementation, the target action triggered by the user based on the area information of the initially displayed macula region could be confirmation or modification.

[0124] In another possible implementation, the target action triggered by the user based on the boundary information of the initially displayed macula region could be confirmation or modification.

[0125] Step 705: Determine the macular region based on the target operation and the initial macular region, and calculate the statistical characteristics of the macular region.

[0126] In one possible implementation, if the target operation is confirmed, the initial macular region is identified as the macular region, and the statistical characteristics of the macular region are calculated.

[0127] In another possible implementation, if the target operation is modification, the user-triggered modification operation is received, the initial macular region is modified based on the modification operation, the modified initial macular region is determined as the macular region, and the statistical characteristics of the macular region are calculated.

[0128] The above-mentioned method first acquires the initial macular region in the vascular image, then calculates the area of ​​the initial macular region, and displays a prompt message for the initial macular region when the area of ​​the initial macular region is greater than a preset area threshold. Then, it receives a target operation triggered by the user based on the prompt message for the initial macular region, determines the macular region based on the target operation and the initial macular region, and calculates the statistical features of the macular region. This method lays the groundwork for the accuracy of determining the statistical features of the macular region in the subsequent process, thereby improving the reliability of the statistical feature threshold and thus improving the accuracy of identifying non-perfusion areas.

[0129] In one embodiment, such as Figure 5 As shown, another identification method is provided, which includes the following steps:

[0130] Step 501: Obtain a vascular image of the eye being examined, which is generated by optical coherence tomography angiography.

[0131] Step 502: Determine the statistical feature threshold based on the statistical characteristics of the macular region, perform convolution processing on the blood vessel image to obtain multiple mapping values ​​of the blood vessel image, and perform binarization processing on the blood vessel image based on the multiple mapping values ​​and the statistical feature threshold to obtain a non-perfusion area mask.

[0132] Step 503: Expand and etch the mask in the unfilled area.

[0133] Step 504: Determine the non-inflation area based on the non-inflation area mask.

[0134] Step 505: Receive a selected region in the vascular image. If the selected region contains the non-perfusion region, mark the non-perfusion region in the selected region and display the marked non-perfusion region.

[0135] Step 506: Determine the number of pixels in the marked non-inflation region, and determine the area information of the marked non-inflation region based on the number of pixels.

[0136] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0137] Based on the same inventive concept, this application also provides an identification device for implementing the identification method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more identification device embodiments provided below can be found in the limitations of the identification method described above, and will not be repeated here.

[0138] In one embodiment, such as Figure 6 As shown, an identification device 600 is provided, including: an acquisition module 601 and a first execution module 602, wherein:

[0139] The acquisition module 601 is used to acquire a vascular image of the eye being examined, which is generated by optical coherence tomography angiography.

[0140] The first execution module 602 is used to obtain statistical features of the macular region in the vascular image and identify non-perfusion regions in the vascular image based on the statistical features.

[0141] In one embodiment, the statistical feature includes at least two of the following: the average value of the pixels in the macula, the median value of the pixels in the macula, the standard deviation of the pixels in the macula, and the squared difference of the pixels in the macula.

[0142] In one embodiment, the first execution module 602 is specifically configured to: determine a statistical feature threshold based on the statistical characteristics of the macular region; perform binarization processing on the vascular image based on the statistical feature threshold to obtain a non-perfusion area mask; and determine the non-perfusion area based on the non-perfusion area mask.

[0143] In one embodiment, the first execution module 602 is specifically used to: expand and etch the mask of the non-injection area.

[0144] In one embodiment, another identification device 700 is provided, which, in addition to the modules in the identification device 600, also includes a second execution module 603, a third execution module 604 and a fourth execution module 605.

[0145] In one embodiment, the second execution module 603 is configured to: receive a selected region in the vascular image; if the selected region contains the non-perfusion region, mark the non-perfusion region in the selected region and display the marked non-perfusion region.

[0146] In one embodiment, the third execution module 604 is configured to: determine the number of pixels in the marked non-inflation region; and determine the area information of the marked non-inflation region based on the number of pixels.

[0147] In one embodiment, the first execution module 602 is specifically configured to: perform convolution processing on the blood vessel image to obtain multiple mapping values ​​of the blood vessel image; and perform binarization processing on the blood vessel image based on the multiple mapping values ​​and the statistical feature threshold.

[0148] In one embodiment, the fourth execution module 605 is specifically configured to: acquire an initial macular region in the blood vessel image; calculate the area of ​​the initial macular region; display a prompt message for the initial macular region if the area of ​​the initial macular region is greater than a preset area threshold; receive a target operation triggered by the user based on the prompt message for the initial macular region; determine the macular region based on the target operation and the initial macular region, and calculate the statistical characteristics of the macular region.

[0149] Each module in the aforementioned identification device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0150] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 8 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements an identification method.

[0151] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0152] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to perform the following steps: acquiring a vascular image of an eye under examination, the vascular image being generated by optical coherence tomography angiography; acquiring statistical features of the macular region in the vascular image; and identifying non-perfusion regions in the vascular image based on the statistical features.

[0153] In one embodiment, the statistical feature includes at least two of the following: the average value of the pixels in the macula, the median value of the pixels in the macula, the standard deviation of the pixels in the macula, and the squared difference of the pixels in the macula.

[0154] In one embodiment, when the processor executes a computer program to identify the non-perfusion region in the vascular image based on the statistical features, it further implements the following steps: determining a statistical feature threshold based on the statistical features of the macular region; performing binarization processing on the vascular image based on the statistical feature threshold to obtain a non-perfusion region mask; and determining the non-perfusion region based on the non-perfusion region mask.

[0155] In one embodiment, after binarizing the vascular image according to the statistical feature threshold to obtain a non-perfusion area mask, the processor further performs the following steps when executing the computer program: expanding and eroding the non-perfusion area mask.

[0156] In one embodiment, the processor, when executing a computer program, further performs the following steps: receiving a selected region in the vascular image, and if the selected region contains the non-perfusion region, marking the non-perfusion region in the selected region and displaying the marked non-perfusion region.

[0157] In one embodiment, when the processor executes the computer program, it further performs the following steps: determining the number of pixels in the marked non-inflated region; and determining the area information of the marked non-inflated region based on the number of pixels.

[0158] In one embodiment, before binarizing the vascular image according to the statistical feature threshold, the processor executes the computer program to further perform the following steps: convolving the vascular image to obtain multiple mapping values ​​of the vascular image; and binarizing the vascular image based on the multiple mapping values ​​and the statistical feature threshold.

[0159] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0160] A vascular image of the examined eye is acquired, generated by optical coherence tomography angiography; statistical features of the macular region in the vascular image are obtained, and non-perfusion areas in the vascular image are identified based on these statistical features.

[0161] In one embodiment, the statistical feature includes at least two of the following: the average value of the pixels in the macula, the median value of the pixels in the macula, the standard deviation of the pixels in the macula, and the squared difference of the pixels in the macula.

[0162] In one embodiment, when the computer program is executed by the processor to identify the non-perfusion region in the vascular image based on the statistical features, it further performs the following steps: determining a statistical feature threshold based on the statistical features of the macular region; performing binarization processing on the vascular image based on the statistical feature threshold to obtain a non-perfusion region mask; and determining the non-perfusion region based on the non-perfusion region mask.

[0163] In one embodiment, after the blood vessel image is binarized according to the statistical feature threshold to obtain a non-perfusion area mask, the computer program, when executed by the processor, further performs the following steps: expanding and eroding the non-perfusion area mask.

[0164] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: receiving a selected region in the vascular image, and if the selected region contains the non-perfusion region, marking the non-perfusion region in the selected region and displaying the marked non-perfusion region.

[0165] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: determining the number of pixels in the marked non-inflated region; and determining the area information of the marked non-inflated region based on the number of pixels.

[0166] In one embodiment, before binarizing the vascular image according to the statistical feature threshold, the computer program, when executed by the processor, further performs the following steps: convolving the vascular image to obtain multiple mapping values ​​of the vascular image; and binarizing the vascular image based on the multiple mapping values ​​and the statistical feature threshold.

[0167] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0168] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0169] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A recognition method, characterized in that, The method includes: Acquire vascular images of the examined eye, the vascular images being generated by optical coherence tomography angiography; Statistical features of the macular region in the vascular image are obtained; correlation calculation is performed based on at least two statistical features of the macular region to obtain a statistical feature threshold. The blood vessel image is binarized according to the statistical feature threshold to obtain a non-perfusion area mask; the non-perfusion region is determined based on the non-perfusion area mask.

2. The method according to claim 1, characterized in that, The statistical characteristics of the macular region include at least two of the following: the average value of the pixels in the macular region, the median value of the pixels in the macular region, the standard deviation of the pixels in the macular region, and the squared difference of the pixels in the macular region.

3. The method according to claim 1, characterized in that, After binarizing the vascular image according to the statistical feature threshold to obtain a non-perfusion area mask, the method further includes: The mask in the non-injection area is subjected to expansion and corrosion treatment.

4. The method according to claim 1, characterized in that, The method further includes: A selected region in the vascular image is received. If the selected region contains the non-perfusion region, the non-perfusion region in the selected region is marked and displayed.

5. The method according to claim 4, characterized in that, The method further includes: Determine the number of pixels in the marked non-inflation region; The area information of the marked non-inflation region is determined based on the number of pixels.

6. The method according to claim 1, characterized in that, Before binarizing the blood vessel image according to the statistical feature threshold, the method further includes: The blood vessel image is convolved to obtain multiple mapping values ​​for the blood vessel image; The blood vessel image is binarized based on the multiple mapping values ​​and the statistical feature threshold.

7. The method according to claim 1, characterized in that, Before obtaining the statistical features of the macular region in the vascular image, the method further includes: Obtain the initial macular region from the vascular image; Calculate the area of ​​the initial macular region; If the area of ​​the initial macular region is greater than a preset area threshold, a prompt message about the initial macular region will be displayed. Receive the target operation triggered by the user based on the initial macular region prompt information; The macular region is determined based on the target operation and the initial macular region, and the statistical characteristics of the macular region are calculated.

8. The method according to claim 1, characterized in that, The associated computational processing includes multiplication and differentiation.

9. An identification device, characterized in that, The device includes: The first acquisition module is used to acquire a vascular image of the eye being examined, the vascular image being generated by optical coherence tomography angiography. The second acquisition module is used to acquire statistical features of the macular region in the vascular image; perform correlation calculation processing based on at least two statistical features of the macular region to obtain a statistical feature threshold; perform binarization processing on the vascular image according to the statistical feature threshold to obtain a non-perfusion area mask; and determine the non-perfusion area based on the non-perfusion area mask.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

Images