Retinopathy recognition model generation method, recognition device and equipment

Abstract

The invention provides a retinopathy recognition model generation method, a recognition device and equipment. The method comprises the steps: training a preset network model by employing fundus imagesin a training set, obtaining a trained retinopathy recognition model, and enabling the fundus images in the training set to carry classification mark information corresponding to the fundus images, wherein the classification mark information is retina feature classification information corresponding to the eye fundus picture, and the preset network model learns the classification mark informationcontained in the eye fundus picture in the training process. Thus, the trained retinal lesion recognition model can be used for recognizing the retina image contained in the detection picture, and anaccurate classification result of the retinal features is given. According to the method, the device and the equipment provided by the embodiment of the invention, the network model for retinopathy recognition is trained in a deep learning mode, so that the accuracy and the detection efficiency of fundus image recognition are improved.

Description

[0001] technology neighborhood [0002] The invention relates to the technical field of medical image processing, in particular to a method for generating a retinal lesion recognition model, a recognition device and equipment. Background technique [0003] Acutely progressive posterior pole retinopathy of prematurity (AP-ROP) is a specific type of retinopathy. Different from conventional retinopathy (ROP), AP-ROP may be accompanied by plus lesions, retinal hemorrhage, and flat neovascularization. For the diagnosis of AP-ROP, there are several clinical problems: First, the course of AP-ROP progresses Unconventional, does not follow the regular progression of ROP from stage 1 to stage 5. At the same time, the incidence of AP-ROP is very low and its symptoms are not typical. Many ophthalmologists do not have enough experience in symptom diagnosis, which increases the possibility that children with AP-ROP cannot be diagnosed in time. [0004] Deep learning has been applied in th...

AI Technical Summary

Problems solved by technology

[0006] In view of the deficiencies in the above-mentioned prior art, the object of the present invention is to provide a method for generating a retinal lesion recognition model, a recognition device and equipment to overcome the lack of neural network for feature recognition of fundus images in the prior art. The model can only rely on the human eye to identify the information contained in the fundus image, resulting in the defects of low image feature recognition efficiency and strong subjectivity

Images