Method for positioning region of interest based on convolutional neural network significance graph

A region of interest and convolutional neural network technology, which is applied in medical automated diagnosis, medical informatics, instruments, etc., can solve the problems of heavy workload, high cost, and low accuracy of locating lesions, achieving less time-consuming work, low cost effect

Active Publication Date: 2016-12-07
杭州健培科技有限公司
View PDF4 Cites 42 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] The present invention provides a method for locating the region of interest based on the saliency map of the convolutional neural network, ai

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
  • Method for positioning region of interest based on convolutional neural network significance graph
  • Method for positioning region of interest based on convolutional neural network significance graph
  • Method for positioning region of interest based on convolutional neural network significance graph

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0063] In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in combination with specific examples and with reference to the detailed drawings. However, the described implementation examples are only intended to facilitate the understanding of the present invention, and do not have any limiting effect on it.

[0064] The automatic positioning of low-density lesions on a lung CT image is taken as an example below. The method for locating a region of interest based on a convolutional neural network saliency map in this embodiment includes the following steps.

[0065] Step 1: Label samples: Screen low-dose lung CT images with a size of 512*512, divide them into images with low-density lesions and images without low-density lesions, and establish sample libraries respectively.

[0066] Step 2: Train a deep convolutional neural network model until convergence:

[0067] (1) C...

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 method for positioning a region of interest based on a convolutional neural network significance graph. The method comprises the following five steps of marking a sample, training a deep convolutional neural network model until the convergence is realized, extracting a significance graph, generating a positioning atlas for a focus, and positioning the focus. According to the technical scheme of the invention, based on the combination of a qualitative-grade calibration sample, a convergent deep convolutional neural network model and a significance graph, the intelligent learn and analysis based on medical image big data can be realized. In this way, the region of interest having a particular attribute in a medical image can be positioned. The method reduces the workload and the cost of calibration samples. Meanwhile, the method is capable of efficiently and accurately positioning the focus position, so as to facilitate doctors to make diagnosis and treatment on medical images.

Description

technical field [0001] The invention belongs to the field of medical image intelligent diagnosis, and mainly relates to a method for locating a region of interest based on a convolutional neural network saliency map. Background technique [0002] With the rapid development of medical imaging technology and computer technology, more and more information such as human physiology, structure, and function are presented to doctors in the form of medical images to help doctors analyze and diagnose diseases. The primary goal to be achieved at this stage is to make rational use of the growing medical imaging data, combined with the most cutting-edge artificial intelligence technology, to provide doctors with faster and more accurate computer-aided diagnosis. [0003] Intelligent algorithms in existing computer-aided diagnosis rely on traditional machine learning classification models and cleverly designed feature extraction engineering for data dimensionality reduction. However, in...

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): G06K9/32G06F19/00
CPCG16H50/20G06V10/25
Inventor 程国华严超费菁媛季红丽
Owner 杭州健培科技有限公司
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