Deep-learning-based classification and grading method for diabetes retinopathy

Abstract

The invention relates to a deep-learning-based classification and grading method for diabetes retinopathy. For each kind of diabetes retinopathy, lots of ophthalmoscope photographs are prepared; a deep convolution neural network including multi-level neural network architecture is established; the deep convolution neural network is trained based on the lots of ophthalmoscope photographs, so that the final output value of the deep convolution neural network meets a grading result of the ophthalmoscope photographs; and thus disease grading can be carried out by using the trained deep convolution neural network. According to the method provided by the invention, with application of the lots of ophthalmoscope photographs including diagnosis marks, needed features are learned automatically from a training example base by deep learning and grading diagnosis is carried out; and data features for determination and the deep convolution neural network parameters are corrected continuously during the training process, so that the grading accuracy and reliability in a real application scene can be improved substantially.

Description

technical field [0001] The present invention relates to a method for grading each type of diabetic retinopathy. Background technique [0002] In the prior art, the grading of various types of diabetic retinopathy is generally based on a number of manually defined features. For example, a Chinese invention patent with publication number CN105513077A discloses a system for screening diabetic retinopathy. It uses a classifier to identify and judge a number of manually defined target features, such as vessel outlines, red lesions (microhemorrhagic tumors), and brightness lesions (exudation, cotton wool spots), etc., so as to predict the grade of diabetic retinopathy Purpose. The current diabetic retinopathy grading technology basically belongs to this technical school. The shortcomings of the above grading methods are: first, the manually defined features have limitations, and cannot make full use of the information in medical images, resulting in limited accuracy in practical...

AI Technical Summary

Problems solved by technology

The shortcomings of the above grading methods are: first, the manually defined features have limitations, and cannot make full use of the information in medical images, resulting in limited accuracy in practical applications; second, the algorithm is static, and the accuracy cannot be obtained with Increased patient data