A glaucoma image detection method based on deep learning of fundus photography

Abstract

The invention discloses a glaucoma image detection method based on deep learning of fundus photography. Preprocessing the original fundus images that have been collected and marked in the database to obtain training example fundus images to form a training database; amplify to obtain an amplified training database; establish a convolutional neural network including a multi-layer neural network structure, The amplified training database is used to input the convolutional neural network for training; for the fundus image to be tested, the fundus image to be tested is input into the trained convolutional neural network to obtain the output value of the last layer of neural network structure, and then the Glaucoma is identified. The invention can continuously optimize the data features used for judgment and the parameters of the convolutional neural network, thereby greatly improving the accuracy, sensitivity and specificity of glaucoma image detection, and saving medical resources.

Description

technical field [0001] The invention relates to an image processing method, in particular to a glaucoma image detection method based on deep learning of fundus photography. Background technique [0002] Glaucoma is a group of eye diseases that threaten and damage optic nerve vision, mainly related to pathological elevated intraocular pressure, and its prevalence is increasing year by year. According to the World Health Organization, it is listed as the second most common cause of blindness worldwide. Because people with glaucoma often come to see a doctor in the advanced stage of the disease, glaucoma is also known as the silent thief. Although glaucoma cannot be cured and the vision damage it causes is irreversible, further damage to vision can be prevented with early diagnosis, appropriate medical treatment, and surgery. Therefore, early diagnosis of glaucoma is crucial to reduce the incidence of blindness. [0003] In the existing computer-aided diagnosis technology of...

AI Technical Summary

Problems solved by technology

The traditional computer-aided diagnosis of glaucoma based on morphological features has the following defects: first, the artificially defined features have limitations and cannot make full use of the information in the fundus image, resulting in limited accuracy in practical applications; second, the algorithm is static, accuracy cannot improve with more patient data

Images