Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A cup-and-disk segmentation method for fundus images based on generative adversarial mechanism

An image and mechanism technology, applied in image analysis, image enhancement, image data processing, etc., can solve the problems of high accuracy rate of glaucoma, long training time of segmentation model, inability to segment optic disc and optic cup, etc., to achieve enhanced features , speed up training time, and maintain stability

Active Publication Date: 2021-06-25
GUANGDONG POLYTECHNIC NORMAL UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the segmentation model training time of existing machine learning and deep learning methods is long
[0007] To sum up, the existing computer-aided technology for optic disc and optic cup segmentation still cannot meet the requirements of high accuracy and low computing time for glaucoma detection at the same time, nor can it realize the segmentation of optic disc and optic cup on fundus images of different imaging technologies. segmentation

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A cup-and-disk segmentation method for fundus images based on generative adversarial mechanism
  • A cup-and-disk segmentation method for fundus images based on generative adversarial mechanism
  • A cup-and-disk segmentation method for fundus images based on generative adversarial mechanism

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0034] A cup-and-disc segmentation method for fundus images based on generative confrontation mechanism, please refer to figure 1 , including the following steps:

[0035] S1: Carry out data enhancement operation on the original image, increase the number of fundus images and extract the target image to narrow the range of feature extraction, and enhance the characteristic contrast of the optic disc and optic cup through rotation transformation, positioning and cropping, and contrast enhancement;

[0036] S2: Automatically extract the features of the optic disc and optic cup through the U-Net network, and train the optic disc segmentation model and the optic cup segmentation model respectively through a large amount of data-enhanced optic disc images, using convolution, batch normalization, and pooling in the contraction path Alternately reduce the size of the optic disc image with downsampling, and extract more and more optic disc and optic cup features; when the number of fe...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a cup-and-disc segmentation method for fundus images based on a generative confrontation mechanism. By performing data enhancement processing on single-channel or multi-channel color fundus images, and then using U-Net network to segment the fundus images, the prediction obtained The segmented image will be sent to the discriminator network for authenticity identification, and the obtained authenticity judgment loss will be returned to adjust the model generated by the U-Net network. After multiple runs of the generation confrontation network, the optimal optic disc segmentation model and optic cup will be finally obtained. Segment the model to realize the segmentation detection of optic disc and optic cup.

Description

technical field [0001] The invention relates to the field of ocular medical image detection, and more specifically, to a cup-and-disc segmentation method for fundus images based on a generative confrontation mechanism. Background technique [0002] Glaucoma is a chronic eye disease in which the optic nerve is gradually damaged, and it is the second leading cause of blindness after cataract. The cup-to-disk ratio is the most commonly used detection method for glaucoma at home and abroad. Pressure and other measurement methods, the calculation of cup-to-disk ratio is relatively stable, and the carriers used for cup-to-disk ratio detection are generally fundus images such as fundus maps and OCT, which can be saved and used for sorting and analysis of medical big data. [0003] Use computer technology to analyze and process digital images to detect and segment the optic disc and cup of the fundus image, and the results are used as auxiliary opinions for diagnostic physicians to ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06T7/00G06T7/10
CPCG06T7/0012G06T2207/10024G06T2207/20081G06T2207/20132G06T2207/30041G06T7/10
Inventor 贾西平黄锦丽刘少鹏方刚陈桂君林智勇陈荣军柏柯嘉廖秀秀张倩
Owner GUANGDONG POLYTECHNIC NORMAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com