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

Brain glioma image grading method and device and storage medium

A grading method and glioma technology, applied in the field of image processing, can solve the problems of large noise, strong dependence on sample data, and long network training time.

Active Publication Date: 2020-10-27
WUYI UNIV
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

There is also a technology for grading malignant tumors through computer technology, but there are still disadvantages that require a large number of labeled samples and are too dependent on sample data, resulting in a long training time for the network and being greatly affected by noise.

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
  • Brain glioma image grading method and device and storage medium
  • Brain glioma image grading method and device and storage medium
  • Brain glioma image grading method and device and storage medium

Examples

Experimental program
Comparison scheme
Effect test

example

[0040] Step S210, segmenting the second processed image to obtain multiple instances 50 of the same size;

[0041] Step S220, selecting a first key instance 51 and a second key instance 52 from a plurality of instances 50, using the first key instance 51 to train the first sub-classification network and using the second key instance 52 to train the second sub-classification network, wherein The first key instance 51 is the maximum value of the instance 50 marked as 1 and the maximum value of the instance 50 marked as 0, and the second key instance 52 is the maximum value of the instance 50 marked as 1 and the minimum value of the instance 50 marked as 0 value, the label of the first key instance 51 and the label of the second key instance 52 are the same as the image-level labels;

[0042] Step S230, input the non-key instance into the trained first sub-classification network to obtain the first label, and input the non-key instance into the trained second sub-classification n...

example 50

[0059] The instance 50 segmentation module 21 is used to segment the second processed image to obtain multiple instances 50 of the same size;

[0060] The classification network training module 22 is used to select the first key instance 51 and the second key instance 52 from a plurality of instances 50, utilize the first key instance 51 to train the first sub-classification network and utilize the second key instance 52 to train the second Sub-classification network, where the first key instance 51 is the maximum value of the instance 50 marked as 1 and the maximum value of the instance 50 marked as 0, and the second key instance 52 is the maximum value of the instance 50 marked as 1 and the maximum value of the instance 50 marked as 0 The minimum value of the instance 50 of , the label of the first key instance 51 and the label of the second key instance 52 are the same as the image-level labels;

[0061] The mask processing module 23 is used to input the non-key examples in...

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 glioma image grading method and device and a storage medium, and by carrying out the normalization of an inputted brain glioma image, and the enhancement of a region ofinterest, the robustness and grading accuracy are improved; mask processing based on a weak supervision principle can accurately add a mask to an area corresponding to the brain glioma in the brain glioma image, so that the robustness and the grading accuracy are further improved, and the dependence on a large amount of training sample data with true value tags is reduced; large-scale time-consuming network training is avoided through the width learning network, and the problems of gradient disappearance and slow training caused by increase of the number of network layers when higher precision is obtained are solved.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a brain glioma image grading method, device and storage medium. Background technique [0002] Brain glioma is a common primary brain malignant tumor, which is usually divided into four grades from grade I to grade IV according to the degree of malignancy. Artificially identifying and grading CT images taken by experienced physicians is currently the main means of grading brain glioma images. There are also technologies for grading malignant tumors through computer technology, but there are still shortcomings that require a large number of labeled samples and are too dependent on sample data, resulting in a long network training time and being greatly affected by noise. Contents of the invention [0003] The purpose of the present invention is to at least solve one of the technical problems in the prior art, and provide a brain glioma image grading method, device and storage med...

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): G06T7/11G06K9/62
CPCG06T7/11G06T2207/20104G06T2207/20081G06T2207/30016G06T2207/30096G06F18/241
Inventor 柯琪锐郭焕鑫陈凯炫周文略曾彩莲王金鑫翟懿奎秦传波甘俊英应自炉陈俊娟曾军英徐颖
Owner WUYI UNIV