Arteriovenous cross compression symptom detection method and device, electronic equipment and readable storage medium

Abstract

The invention discloses an arteriovenous cross compression symptom detection method and device, electronic equipment and a readable storage medium, and relates to the technical field of deep learning,the technical field of computer vision and the field of AI medical treatment. According to the specific implementation scheme, the method includes performing blood vessel segmentation on a to-be-detected image to obtain a blood vessel segmentation result that whether each pixel in the to-be-detected image has a blood vessel or not; determining a blood vessel intersection point according to the blood vessel segmentation result; acquiring a region of interest ROI image of the blood vessel intersection from the to-be-detected image; utilizing a classification model to classify the ROI image to obtain a classification result of whether the ROI image has arteriovenous cross compression characteristics or not, wherein the classification model is obtained by training in advance based on a supervised training mode. According to the invention, the accuracy of the arteriovenous cross compression symptom detection result can be improved.

Description

technical field [0001] The invention relates to the technical field of image processing, specifically to the technical field of deep learning, the technical field of computer vision and the medical field of AI, and in particular to a detection method, device, electronic equipment and readable storage medium for arteriovenous cross compression syndrome. Background technique [0002] Arteriovenous nicking (AV nicking) refers to the phenomenon that veins are compressed by hardened arteries at the intersection of veins and arteries due to elevated blood pressure (ie, hypertension). In retinal color photographs, the arteriovenous cross compression sign is manifested as a decrease in the caliber of the veins on both sides of the arteriovenous (arteriovenous, AV) intersection. Such as figure 1 As shown in (a), it is a schematic diagram of the normal arteriovenous intersection, figure 1 (b) is a schematic diagram of arteriovenous cross compression sign, figure 1 (c) is a retinal ...

AI Technical Summary

Problems solved by technology

[0005] However, due to the influence of shooting equipment and shooting operations, the quality of fundus images will vary greatly. At the same time, due to the large differences in the fundus structure of each person and the possibility of fundus diseases, the unsupervised Kmeans algorithm is used for arteriovenous segmentation. The robustness of arteriovenous vessels is poor, and the inaccurate information of arteriovenous vessels will directly affect the accuracy of the detection results of arteriovenous cross compression signs; in addition, because the gray values ​​of arteriovenous vessels in fundus images are relatively close, resulting in the appearance of arteriovenous vessels If the difference is low, it will also reduce the accuracy of the arteriovenous vessel segmentation results, thereby affecting the accuracy of the arteriovenous cross compression sign detection results

Images