Screening method and system for fundus image of glaucoma based on deep learning

Abstract

Provided are a screening method and system for a fundus image of glaucoma based on deep learning. The screening method comprises the steps that the fundus image to be identified is acquired; an FCN model is extracted through an optic disk to position and cut the optic disk, and optic disk sub-images are extracted and obtained; the optic disk sub-images are identified through a classification CNN model, and a classification result of the fundus image of glaucoma is obtained. According to the screening method and system for the fundus image of the glaucoma based on deep learning, identificationis only conducted on the cut and extracted optic disk, so that information processing amount is reduced, and the screening method and system can assist a doctor in improving the detection efficiency of glaucoma.

Description

technical field [0001] The present invention relates to a technology in the field of artificial intelligence, in particular to a glaucoma fundus image screening method and system based on deep learning. Background technique [0002] Glaucoma is known as the silent stealer of light, because except for a few patients with red eyes, eye pain, and decreased vision during acute glaucoma attacks, the vast majority of early glaucoma patients do not have any prominent symptoms. When it is abnormal, glaucoma has caused irreversible damage to vision, so early detection and early treatment are effective means to control blindness caused by glaucoma. [0003] At present, most medical examination items do not include glaucoma examination, mainly because the diagnosis of glaucoma requires the examiner's rich clinical experience in ophthalmology, and general medical examination centers do not have professional and rich clinical experience in ophthalmology, so glaucoma examination is diffic...

AI Technical Summary

Problems solved by technology

[0003] At present, most of the physical examination items do not include glaucoma examination, mainly because the diagnosis of glaucoma requires the examiner's rich clinical experience in ophthalmology, and general physical examination centers do not have professional and rich clinical experience in ophthalmology, so glaucoma examination is difficult to popularize

[0004] The use of computer-aided examination is an ideal choice for universal glaucoma screening, but the current implementation of computer-aided examination is relatively complicated

In the traditional clinical examination, the main basis for judging glaucoma is the cup-to-disk ratio. However, clinical practice has proved that it is inaccurate to judge glaucoma only by the cup-to-disk ratio. Many funduses with physiologically large cups also have a large cup-to-disk ratio; Auxiliary judgment criteria need to be introduced, such as 1) whether the edge of the plate conforms to the ISNT rule, 2) the principle that the lower edge is thicker than the upper edge, 3) the principle that the upper edge is thicker than the nasal side, 4) the principle that the side is the thinnest, etc. On the one hand, these standards are converted It is difficult to quantitatively describe the characteristics of computer programs. On the other hand, comprehensive utilization of these characteristics requires a large number of branch judgment conditions, which greatly increases the complexity of the model, and the stability of the model and test results is not easy to guarantee.

Images