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

Network model and segmentation method for endometrial neoplasm segmentation

A network model, endometrial technology, applied in biological neural network models, neural learning methods, applications, etc., can solve the problems of unsatisfactory and difficult segmentation of small objects, and achieve accurate segmentation and scale reduction. Effect

Inactive Publication Date: 2021-04-09
UNIV OF SHANGHAI FOR SCI & TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, U-net has a good effect on the segmentation of large objects, and the segmentation effect on small objects is not ideal.
It is very difficult to segment endometrial tumors directly using the U-net network

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
  • Network model and segmentation method for endometrial neoplasm segmentation
  • Network model and segmentation method for endometrial neoplasm segmentation
  • Network model and segmentation method for endometrial neoplasm segmentation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0027] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be further described below.

[0028] The present invention proposes a network model for endometrial tumor segmentation, and trains two network models based on the U-net network: the first network model and the second network model, and the first network model is used for segmenting images Uterine region, the second network model is used to segment the endometrial tumor region in the image;

[0029] The MR image of endometrial cancer and its corresponding segmentation map are used as the training set to train the network model, in which the energy function is used to calculate the soft-max value of each pixel, which is defined as:

[0030]

[0031] where a k (x) represents the activation value of the number of feature channel categories k at the position x of the pixel, p k (x) is the approximate maximum function;

...

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 provides a network model and segmentation method for endometrial neoplasm segmentation. The method mainly comprises the following processing steps: preprocessing an original image; inputting the image into a first network model, and training the input image to obtain a segmentation result I1 of a uterus region; and inputting a uterus region grayscale image into a second network model for training to obtain a segmentation result I2 of a neoplasm. According to the network model and the segmentation method, the endometrial neoplasm is directly segmented on the whole magnetic resonance (MR) image; for the conditions that the accuracy is low, the sensitivity is low and the like due to the fact that a target object is relatively small, the grayscale of the uterus region is recovered on the basis of uterus segmentation, and then the uterus region image with the grayscale recovered is segmented, and thus, the proportion of the neoplasm in the image is well reduced, thus being more conducive to accurate segmentation of the neoplasm.

Description

technical field [0001] The present invention relates to the technical field of tumor segmentation methods, in particular to a network model and segmentation method for endometrial tumor segmentation. Background technique [0002] Surgery is the mainstay of treatment for endometrial cancer. For early-stage patients, the purpose of surgery is surgical-pathological staging, accurate judgment of lesion extent and prognosis, resection of the diseased uterus and possible metastatic lesions, and decision on the choice of postoperative adjuvant therapy. When judging uterine lesions, medical MR imaging usually includes a lot of tissues and organs near the uterus, and the uterus only accounts for a small part of the whole image, while endometrial tumors exist inside the uterus. The proportion of the tumor is even smaller. At this time, the automatic segmentation of the tumor will face greater difficulties. As the number of patients increases year by year, doctors need to manually se...

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): A61B5/055A61B5/00G06N3/04G06N3/08G06T7/11
CPCA61B5/055A61B5/7267A61B5/004A61B5/4325G06N3/08G06T7/11A61B2576/02G06N3/045
Inventor 秦晨阳
Owner UNIV OF SHANGHAI FOR SCI & TECH
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