Retinal image arteriovenous classification device and equipment and storage medium

Abstract

The invention discloses a retinal image arteriovenous classification device, and relates to the field of medical image processing. The device comprises an image obtaining module which is used for obtaining a to-be-classified image; a preprocessing module used for preprocessing the to-be-classified image to obtain a preprocessed image; an adversarial network module which comprises a pre-trained generator unit used for outputting the arteriovenous classification probability of each pixel point according to the preprocessed image; and a result output module used for generating an arteriovenous classification result according to the arteriovenous classification probability of each pixel point in the preprocessed image. The invention further provides retinal image arteriovenous classification equipment and a storage medium, the complexity of a classification algorithm can be reduced, the characteristics and metric functions of a retinal vessel topological structure do not need to be manually designed, and the classification speed and accuracy of the arteriovenous classification of retinal images are improved.

Description

technical field [0001] The invention relates to the field of medical image processing, in particular to a device, equipment and storage medium for classifying retinal image arteries and veins. Background technique [0002] The fundus has very important structures, including the optic nerve, retinal arteriovenous vessels, macula, etc. Through fundus examination, doctors can find ophthalmic diseases such as optic neuropathy and retinal vascular disease, as well as cardiovascular and cerebrovascular diseases such as hypertensive arteriosclerosis and diabetic retinopathy. systemic disease. Retinal blood vessels are the only deep-level microvascular system in the human body that can be directly observed without damage. The changes in the width and curvature of the arteriovenous vessels are closely related to the development of systemic diseases. The main cause of eye vision problems. Therefore, the fully automatic retinal vessel segmentation and calculation method has very prac...

AI Technical Summary

Problems solved by technology

However, based on the method of deep learning, the cross-entropy loss function is commonly used in image segmentation tasks, which is based on the loss of pixels, and does not take into account the topological structure information of blood vessels, resulting in segmental continuous segmentation results of arteries and veins. , rather than a complete segment of arterial or venous vessel

Images