Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Hand bone key region acquisition method based on convolutional neural network and multi-granularity attention

A convolutional neural network and attention technology, applied in the field of hand bone key area acquisition based on convolutional neural network and multi-granularity attention, can solve problems such as expensive, insufficiently fine and diverse areas, and achieve the effect of improving accuracy

Pending Publication Date: 2022-05-27
HEFEI UNIV OF TECH +1
View PDF0 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, manual annotation not only requires expert knowledge, but is also expensive and subjective
Moreover, the regions located by the existing attention-based methods are not fine and diverse enough.

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
  • Hand bone key region acquisition method based on convolutional neural network and multi-granularity attention
  • Hand bone key region acquisition method based on convolutional neural network and multi-granularity attention
  • Hand bone key region acquisition method based on convolutional neural network and multi-granularity attention

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0045] In this embodiment, a method for acquiring key regions of hand bone based on convolutional neural network and multi-granularity attention can locate the most discriminative ossification region only by using the bone age annotation information of the hand bone picture, and locate the most differentiated ossification region. The area of ​​​​is consistent with the key area used by the TW3 method. Specifically, as figure 1 shown, proceed as follows:

[0046] Step 1: Normalize the size of all hand bone X-ray images, and apply random affine transformation and random horizontal flip for data enhancement to obtain an input image set, and any image in the input image set is marked as I;

[0047] In this example, the dataset used is the public bone age dataset provided by the Radiological Society of North America in 2017. All hand images contain bone age and gender information. The original spatial resolution of the hand bone images in this dataset is not the same. Since the 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 hand bone key region acquisition method based on a convolutional neural network and multi-granularity attention. The method comprises the following steps: 1, acquiring a data set containing bone age information of a hand bone; 2, constructing a bone age evaluation network containing multi-granularity attention; 3, performing off-line training on the established bone age evaluation network; and 4, acquiring a key area of the hand bone by utilizing attention in the trained network. According to the method, the key region of the hand bone can be obtained only through the bone age label of the hand bone, the obtained key region is consistent with a key region used by a TW3 method, and the difficulty that an existing key region obtaining method depends on manual labels is overcome.

Description

technical field [0001] The invention relates to the field of medical image analysis, in particular to a method for acquiring key areas of hand bones based on a convolutional neural network and multi-granularity attention. Background technique [0002] Bone age assessment on hand radiographs is often used to investigate growth in children. In clinical practice, the diagnosis and treatment of growth disorders can be aided and monitored by exploring the delayed or accelerated appearance of non-handed ossification centers. [0003] The TW method assesses the maturity level of specific bones of the hand and wrist. The maturity levels of different bones correspond to different maturity scores. The skeletal maturity is converted into scores and the scores of all key areas are added to calculate the total maturity score. The bone age is finally calculated from the total maturity score. The TW3 method is relatively more complicated and requires more time. Since the maturity of eac...

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): G06T7/00G06N3/04G06K9/62G06V10/80
CPCG06T7/0012G06T2207/30008G06N3/044G06N3/045G06F18/253
Inventor 王晓华范伟程峰胡敏盛海王宇航
Owner HEFEI UNIV OF TECH
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
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