Eye fundus image-based diabetes and related disease classification method and equipment

Abstract

The invention provides an eye fundus image-based diabetes and related disease classification method and equipment. The method comprises the steps that sample data are acquired, the sample data comprise fundus images and multiple types of information, and the multiple types of information are diabetes type information and at least one kind of diabetes related disease type information or at least two kinds of diabetes related disease type information; a machine learning model is trained by using a large amount of sample data, thus the evaluation results are output, wherein the evaluation resultat least comprises a classification result corresponding to the type information, the machine learning model comprises a feature extraction network and at least one output network, the feature extraction network is used for extracting feature information from the fundus image, and the at least one output network is used for outputting the evaluation result according to the feature information; andthe machine learning model at least adjusts own parameters according to the output evaluation result and the type information in the sample data.

Description

Technical field [0001] The invention relates to the field of medical image analysis, in particular to a method and equipment for classifying diabetes and related diseases based on fundus images. Background technique [0002] In recent years, machine learning technology has been widely used in the medical field. In particular, machine learning technology represented by deep learning has received widespread attention in the field of medical imaging. In terms of fundus image detection, deep learning technology can more accurately detect a certain feature of fundus images. For example, use a large number of fundus image samples of diabetic patients to train the deep learning model, and use the trained model to perform diabetes detection on the fundus image . [0003] The Chinese patent application number 201810387302.7 discloses a fundus image detection method based on machine learning. The method first detects the entire area of ​​the fundus image with a higher degree of feature dete...

AI Technical Summary

Problems solved by technology

This solution is suitable for identifying multiple diseases at the same time, such as detecting whether the fundus image has multiple irrelevant disease characteristics such as macular hole, glaucoma, and diabetes to obtain corresponding classifications, and is not suitable for the classification of related diseases

Images