Automatic segmentation of brain tumor images based on convolution neural network

A convolutional neural network and automatic segmentation technology, applied in the field of medical image segmentation and deep learning, can solve the problems of long image preprocessing time, rough segmentation of brain tumor images, and unbalanced categories, and achieve shortened processing time and image preprocessing. Handle the effect of convenient operation and solve the category imbalance

Active Publication Date: 2018-12-18
UNIV OF ELECTRONIC SCI & TECH OF CHINA
View PDF3 Cites 26 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a method for automatic segmentation of brain tumor images based on convolutional neural networks

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
  • Automatic segmentation of brain tumor images based on convolution neural network
  • Automatic segmentation of brain tumor images based on convolution neural network
  • Automatic segmentation of brain tumor images based on convolution neural network

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0036] Example 1

[0037] The method for automatically segmenting brain tumor images based on convolutional neural network of the present invention includes the multi-modal MRI image of brain tumors, and further includes the following steps:

[0038] Step 1. Collect multi-modal MRI images of brain tumors and perform image preprocessing to obtain the original image set;

[0039] Step 2. Construct a framework for brain tumor segmentation based on multi-modal MRI images; the framework includes module one and module two, and module one includes a 3D convolutional neural network, residual unit and transposed convolution as the basis to form a parallel The deep deconvolution neural network is used to output the contour map of the brain tumor segmentation image; the second module includes a jump structure added to the deep deconvolution neural network structure in the module one, and it is used to output the lesion area segmentation of the brain tumor image Figure

[0040] Step 3. Input the...

Example Embodiment

[0047] Example 2

[0048] This embodiment further defines the following on the basis of embodiment 1: In step 1, the multimodal MRI image is four modal images Flair, T1, T2, T1C, and the two modalities Flair and T2 The image in the state is corrected by N4ITK, and the contrast of the image in the T1C and T1 modes is adjusted. Standardize the gray level of images between different individuals: first subtract the average value of the entire image and divide by the standard deviation of the brain area, adjust the pixel values ​​of all images to the interval [-5, 5], and return the entire image to Change to [0,1], and set the non-brain area to 0. Finally, translation transformation, distortion enhancement and elastic deformation are performed on the preprocessed data. The detailed process is as follows: First, it is necessary to acquire multi-modal MRI images of brain tumors. In this method, only four modal images of Flair, T1, T2, and T1C are used. Flair modal MRI images contain t...

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 brain tumor image automatic segmentation method based on a convolution neural network, which comprises the following steps: acquiring a multimodal MRI image of the brain tumor and carrying out image preprocessing to obtain an original image set; constructing the frame of brain tumor segmentation based on multimodal MRI images; the frame comprises a module 1 and a module 2. The module 1 comprises a parallel deep deconvolution neural network based on a 3d convolution neural network, a residual unit and a transposed convolution; the module 2 includes adding a hopping structure on the basis of the deep deconvolution neural network structure in the module 1; after several iterative training, the ideal weights are obtained and the segmentation images of brain tumors areoutput. Image segmentation results are tested and evaluated. The invention effectively solves the problem of low accuracy of brain tumor segmentation image segmentation. It can improve the recognition of tumors and make the image preprocessing more convenient. Using the loss function based on d coefficients in the segmentation module can effectively solve the problem of class imbalance.

Description

technical field [0001] The invention relates to the fields of medical image segmentation and deep learning, in particular to a method for automatic segmentation of brain tumor images based on convolutional neural networks. Background technique [0002] Medical image segmentation is a key technology in image analysis and processing. Separating relevant tissues of interest according to the similarity and specificity of the image area is of great significance to the clinical diagnosis and treatment process and is the main premise of all follow-up work. , the quality of the segmentation effect will directly affect the smooth progress of the information processing work. Medical image segmentation of brain tumors is an important branch in the field of image segmentation. Brain tumor segmentation technology plays an important role in the clinical diagnosis and treatment of brain tumors. Through the segmentation results of brain tumors, doctors can measure tumor Size and location, ...

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): G06T7/11G06T7/12G06N3/04
CPCG06T7/11G06T7/12G06T2207/30096G06T2207/30016G06T2207/20084G06T2207/20081G06T2207/10088G06N3/045
Inventor 程建郭桦苏炎洲高银星许轲
Owner UNIV OF ELECTRONIC SCI & TECH OF CHINA
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