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

Ultrasonic and nuclear magnetic image registration fusion method and device based on hybrid supervised learning

A nuclear magnetic image and supervised learning technology, applied in the field of medical image processing, can solve the problems of manual labeling and difficult anatomical labeling, and achieve the effect of reducing difficulty and time cost, reducing quantity and accuracy requirements, and reducing the number of parameters.

Pending Publication Date: 2020-02-25
ARIEMEDI SCI SHIJIAZHUANG CO LTD
View PDF6 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, for the multimodal image registration task of liver MR and US, some anatomical labels (such as septum and small blood vessels) are difficult to manually annotate in 3D US images

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
  • Ultrasonic and nuclear magnetic image registration fusion method and device based on hybrid supervised learning
  • Ultrasonic and nuclear magnetic image registration fusion method and device based on hybrid supervised learning
  • Ultrasonic and nuclear magnetic image registration fusion method and device based on hybrid supervised learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0019] For the multimodal image registration task of liver MR and US, some anatomical labels (such as septum and small blood vessels) are difficult to manually annotate in 3D US images. Fortunately, the septum is usually the brightest part of an ultrasound image and has a strong gradient response, so some high-level similarity measures can also be used to handle the registration of these regions.

[0020] Inspired by these ideas, applicants propose a registration network based on hybrid supervision. First, a rigid network driven by vessel labels ensures that the two images are rigidly aligned in space, then we use the training of an elastic registration network co-supervised by vessel labels and the MIND similarity metric to predict the difference between MR and 3D ultrasound images. deformation relationship. Experiments show that this method greatly improves the performance.

[0021] Such as Figure 4 As shown, this method of registration and fusion of ultrasound and nucle...

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

According to an ultrasonic and nuclear magnetic image registration fusion method and device based on hybrid supervised learning, the task requirements are met, the training difficulty and time cost are greatly reduced, and meanwhile the requirements for the number and precision of artificial tags are reduced. The method comprises the steps that (1) a global rigid registration network is connectedwith a local B spline registration network in series, rigid deformation serves as a global deformation model, and a B spline serves as an elastic deformation model; wherein the rigid network is drivenby the vessel label to ensure that the ultrasound and nuclear magnetic images are rigidly aligned spatially; and (2) the deformation relationship between the MR and 3D ultrasonic images is predictedby using the training of an elastic registration network jointly supervised by the blood vessel label and MIND similarity measurement, wherein the MIND similarity measurement is mainly used, and theartificially segmented blood vessel label is used as auxiliary supervision.

Description

technical field [0001] The present invention relates to the technical field of medical image processing, in particular to a method for registration and fusion of ultrasonic and nuclear magnetic images based on hybrid supervised learning, and a registration and fusion device for ultrasonic and nuclear magnetic images based on hybrid supervised learning. Background technique [0002] Registration and fusion of MR (ie, Magnetic Resonance Imaging, Magnetic Resonance Imaging) and US (ie, ultrasound scan, ultrasound scan) liver images can provide high-quality guidance for liver ablation procedures. However, accurate MR-US registration is still a challenging task, and the main difficulties come from the large changes in the appearance of tissues or organs when imaged by different physical principles, including: gray-scale changes between two modalities, ultrasound contrast Low imaging quality, and large-scale deformation of the liver due to the compression of the ultrasound probe, ...

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): G06T7/33G06T7/00G06T5/50G06K9/62
CPCG06T7/33G06T5/50G06T7/0012G06T2207/10132G06T2207/10088G06T2207/30101G06T2207/20221G06T2207/20081G06T2207/20084G06F18/22
Inventor 丛伟建武潺邓巧玲
Owner ARIEMEDI SCI SHIJIAZHUANG CO LTD
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