Unlock instant, AI-driven research and patent intelligence for your innovation.

Low-dose X-ray image de-noising method based on cascaded convolutional neural network

A convolutional neural network, low-dose technology, applied in the field of low-dose X-ray image denoising based on cascaded convolutional neural network

Inactive Publication Date: 2017-11-28
SHENZHEN WEITESHI TECH
View PDF0 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at solving the problem of denoising medical images with low radiation dose, the purpose of the present invention is to provide a low-dose X-ray image denoising method based on cascaded convolutional neural network, and propose a convolutional neural network based on cascaded structure A New Framework for the Web

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
  • Low-dose X-ray image de-noising method based on cascaded convolutional neural network
  • Low-dose X-ray image de-noising method based on cascaded convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0032] It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other. The present invention will be further described in detail below in conjunction with the drawings and specific embodiments.

[0033] figure 1 It is a system flowchart of a low-dose X-ray image denoising method based on a cascaded convolutional neural network of the present invention. It mainly includes denoiser structure; cascade structure; parameter training.

[0034] Among them, the denoiser structure includes network structure and network parameters.

[0035] The network structure is the basic unit for constructing a convolutional network. According to the samples and quantity of the input image, the number required for stacking of basic units is determined. The basic unit structure is as follows:

[0036] 1) Convolution layer;

[0037] 2) Convolution layer, activation function;

[0038] 3) Convolution...

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 low-dose X-ray image de-noising method based on a cascaded convolutional neural network presented by the invention mainly includes a de-noising device structure, a cascade structure, and parameter training. First, independent basic function units fi are constructed using a convolutional neural network, Poisson noise is added according to an input normal-dose X-ray image XH to generate a low-dose version image XL, the image is input to the network fi to map XL to the residual of XH, and noise is reduced via the same fi. Then, N cascaded fi of the same structure is constructed in sequence, and end-to-end training is implemented through residual passing to remove noise in the image. An X-ray image generated below normal radiation dose can be processed. A cascaded convolutional neural network is provided to solve the problem on residual passing. The removal efficiency of blocky and streak marks caused by noise is improved.

Description

technical field [0001] The invention relates to the field of image denoising, in particular to a low-dose X-ray image denoising method based on a cascaded convolutional neural network. Background technique [0002] X-rays have been widely used in various imaging techniques, especially medical images, because of their non-deflection characteristics in electric or magnetic fields, which provide an accurate reference for modern diagnosis and treatment techniques. However, when X-ray scanning is performed on the human body and lesion parts are collected, a certain radiation dose will be generated, and these radiation doses can lead to the generation of cancer to a certain extent. Therefore, the use of low radiation dose to generate X-ray images has gradually attracted attention, but at the same time, the disadvantage of low-dose images is the addition of inevitable noise, which is a crucial factor affecting the quality of medical images. The existence of noise will reduce the p...

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): G06T5/00G06N3/08
CPCG06N3/08G06T2207/20081G06T2207/10116G06T5/70
Inventor 夏春秋
Owner SHENZHEN WEITESHI TECH