Fundus retina blood vessel recognition and quantification method, device and equipment and storage medium

Abstract

The invention provides a fundus retinal vessel recognition and quantification method, device and equipment and a storage medium, and the method comprises the steps: inputting an original fundus imageinto a pre-trained U-shaped convolutional neural network model for processing, and obtaining a target feature map; performing optic disk segmentation based on the target feature map; segmenting the original fundus image to obtain an arteriovenous blood vessel recognition result; carrying out region-of-interest positioning based on the optic disk segmentation result; extracting a blood vessel center line according to the arteriovenous blood vessel recognition result, detecting key points in the blood vessel center line, removing the key points to obtain a plurality of mutually independent bloodvessel sections, and correcting arteriovenous category information on each blood vessel section; and based on the extracted blood vessel center line, obtaining the blood vessel diameter of each bloodvessel section after category information correction, and then quantifying arteriovenous blood vessels in the region of interest. According to the embodiment of the invention, the fundus retina artery and vein vessel identification precision is improved, and the quantization precision is further improved.

Description

technical field [0001] The present application relates to the technical field of image processing, and in particular to a fundus retinal blood vessel identification and quantification method, device, equipment and storage medium. Background technique [0002] The fundus retinal arteries and veins have always been the focus of medical research, especially those within 1pd-1.5pd (Papillary Diameter, optic disc diameter) from the center of the optic disc. Evidence for early diagnosis of systemic and blood diseases, such as cardiovascular disease, diabetes, hypertension, etc. Calculating the diameter ratio of retinal arteriovenous vessels requires accurate classification of arteriovenous vessels. Traditional fundus retinal diagnosis relies on doctors to observe fundus images and obtain diagnostic results based on their own medical experience. This method takes a long time and is The workload is heavy, and the identification and classification of arteriovenous vessels are greatl...

AI Technical Summary

Problems solved by technology

Calculating the diameter ratio of retinal arteriovenous vessels requires accurate classification of arteriovenous vessels. Traditional fundus retinal diagnosis relies on doctors to observe fundus images and obtain diagnostic results based on their own medical experience. This method takes a long time and is The workload is heavy, and the identification and classification of arteriovenous vessels are greatly affected by subjectivity

With the development of computer image processing technology, at present, fundus color photos are used to extract fundus retinal blood vessels. Because the fundus color photos have the characteristics of uneven brightness, complex interlacing of blood vessels and background colors, and small difference between arteries and veins, it is very important for arteriovenous blood vessel identification and recognition. Classification causes certain difficulties, and the existing research on automatic identification of arteriovenous vessels mainly realizes local arteriovenous vessel identification based on color or partial structure of blood vessels, but the identification accuracy is still limited, which leads to low quantification accuracy

Images