Boundary enhanced convolutional neural network for OCT image cornea layer segmentation

A convolutional neural network and corneal layer technology, applied in the field of image segmentation, can solve problems such as underutilization, achieve accurate extraction, reduce segmentation errors, and improve detection sensitivity and effectiveness

Active Publication Date: 2021-07-23
THE EYE HOSPITAL OF WENZHOU MEDICAL UNIV
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, none of the above segmentation networks make full use of various boundary information in images to assist in the segmentation of objects of interest.

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
  • Boundary enhanced convolutional neural network for OCT image cornea layer segmentation
  • Boundary enhanced convolutional neural network for OCT image cornea layer segmentation
  • Boundary enhanced convolutional neural network for OCT image cornea layer segmentation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0025] A boundary-enhanced convolutional neural network for OCT image corneal layer segmentation is further described below in conjunction with the accompanying drawings;

[0026] refer to figure 1 , a kind of boundary enhancement convolutional neural network for OCT image corneal layer segmentation among the present invention, comprises the following steps:

[0027] Step 1, analyze the operational similarity between the Gaussian difference boundary detection algorithm and the convolution module in the existing segmentation network, and then improve the convolution module in the segmentation network based on the Gaussian difference algorithm and explain the feasibility and rationality of this improvement strategy .

[0028] Step 2, the design of the convolution module

[0029] Inspired by the Gaussian difference boundary detection algorithm, a pixel-based subtraction operation is introduced between the two convolutional layers of the existing convolution module to obtain a n...

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

A boundary enhanced convolutional neural network for OCT image cornea layer segmentation is characterized in that a to-be-designed convolution module is integrated into a BiO-Net network in a module replacement mode to serve as coding and decoding convolution modules at the same time, so that the boundary enhanced convolutional neural network is constructed. According to the network, boundary convolution features and non-boundary convolution features are transmitted to convolution modules of different levels by means of forward and reverse jump links in BiO-Net, through learning and detection of the two convolution features, the detection sensitivity and effectiveness of an interested target and a boundary region thereof in an image are jointly improved, the segmentation error of the target boundary region is reduced, different cornea layers in the OCT image can be accurately extracted at the same time, and the segmentation performance of the boundary enhanced convolutional neural network is superior to that of existing networks such as U-Net and BiO-Net.

Description

technical field [0001] The invention relates to the technical field of image segmentation, in particular to a boundary-enhanced convolutional neural network for corneal layer segmentation in OCT images. Background technique [0002] Image segmentation is an important image processing technology, which is used to accurately and clearly distinguish different regions of interest in the image, so that these regions have different imaging characteristics (such as gray scale distribution and tissue contrast), thus facilitating the quantitative analysis of regions of interest. With the help of this technology, the detection, location and measurement of morphological features of lesion areas in medical images can be performed, which can greatly reduce the image processing time of image analysts and improve the accuracy of clinical diagnosis of diseases, so it has very important research value. In order to accurately segment images, various types of segmentation algorithms have been ...

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 Applications(China)
IPC IPC(8): G06T7/136G06N3/04G06N3/08
CPCG06T7/136G06N3/08G06T2207/20081G06T2207/20084G06T2207/30041G06T2207/30204G06N3/045
Inventor 王雷常倩沈梅晓吕帆陈浩
Owner THE EYE HOSPITAL OF WENZHOU MEDICAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap