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

Retinal blood vessel segmentation method based on deep-learning adaptive weight

A technology of retinal blood vessels and adaptive weights, which is applied in the field of medical image processing, can solve the problems of lack of robustness of textures, achieve the effects of improving segmentation accuracy, eliminating class imbalance problems, and avoiding interference

Active Publication Date: 2017-10-24
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF7 Cites 46 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The matched filter method can detect blood vessel-like objects very well, but the disadvantage of this method is that it is not robust to the texture in the background, and it is easy to extract the texture in the background as a blood vessel

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
  • Retinal blood vessel segmentation method based on deep-learning adaptive weight
  • Retinal blood vessel segmentation method based on deep-learning adaptive weight
  • Retinal blood vessel segmentation method based on deep-learning adaptive weight

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0027] The method of the present invention first expands the sample of the retinal blood vessel image, and groups the samples; aiming at the problem of small receptive field and slow convergence speed, a blood vessel segmentation full convolutional neural network structure is established, and the network is pre-trained with training samples to obtain The initial model parameters of retinal blood vessel segmentation; in order to solve the problem of the imbalance between the blood vessel pixel and the background pixel in the retinal image, a global adaptive weight method is proposed, which can update the weight of the pixel in the loss function as iteratively progresses, and promote the nerve The loss function of the network mainly comes from the mis-segmented area, and at the same time accelerates the network convergence; at the end of the network layer, a conditional random field layer is added to enhance the spatial constraints of the characteristics, and the network is tuned; ...

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 retinal blood vessel segmentation method based on the deep-learning adaptive weight, comprising: performing sample expansion on retinal blood vessel images and grouping the samples; constructing a full-convolution neural network for blood vessel segmentation; pre-training the network through the training samples; performing the global adaptive weight segmentation on the retinal blood vessel images; obtaining the initial model parameters for retinal blood vessel segmentation; adding a conditional random field layer to the last network layer and optimizing the network; using the rotation test method to input the testing samples to the network; and obtaining the retinal blood vessel segmentation result. The blood vessel segmentation full-convolution neural network structure and the adaptive weight method proposed by the invention can realize the human-eye level image segmentation and can test on two internationally published retinal image databases DRIVE and CHASE_DB1 with the average accuracy of 96.00 % and 95.17% respectively, both higher than the latest algorithm.

Description

Technical field [0001] The invention relates to the field of medical image processing, in particular to a method for segmenting retinal blood vessels based on deep learning adaptive weights. Background technique [0002] The retina can be used as an important detection index for common diseases, such as hypertension and diabetes, and has been a hot spot in domestic and foreign research for many years. Computer-based automatic extraction, measurement and analysis of fundus blood vessels have important application value in medical diagnosis. [0003] The segmentation methods of retinal blood vessel images are mainly divided into two categories: rule-based and learning-based. [0004] The rule-based method is mainly to use the characteristics of blood vessels in the retinal image, and design corresponding filters to complete the task of enhancing blood vessel features and suppressing background noise. It usually consists of three parts: preprocessing, segmentation, and postprocessing. ...

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
IPC IPC(8): G06T7/10G06N3/04G06N3/08
CPCG06T7/10G06T2207/20081G06T2207/20084G06T2207/30041
Inventor 程洪徐宏罗院生杨路
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA
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