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

Quick imaging model training method and device and server

A technology of imaging model and training method, which is applied in the direction of instruments, image data processing, character and pattern recognition, etc., and can solve the problems of inability to optimize imaging effect and poor undersampling mask

Active Publication Date: 2020-05-01
SHENZHEN INST OF ADVANCED TECH CHINESE ACAD OF SCI
View PDF11 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In view of this, the embodiment of the present invention provides a fast imaging model training method, device and server to solve the problems that the undersampling mask cannot be optimized and the imaging effect is not good

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
  • Quick imaging model training method and device and server
  • Quick imaging model training method and device and server
  • Quick imaging model training method and device and server

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0050] Such as figure 1 As shown, it is a schematic flow chart of the training method for the fast imaging model provided by Embodiment 1 of the present invention. This embodiment is applicable to the application scene of magnetic resonance scanning imaging, and the method can be executed by a training device for a fast imaging model, which can be a server, an intelligent terminal, a tablet or a PC, etc.; The training device of the model is described as the execution subject, and the method specifically includes the following steps:

[0051] S110. During each model iteration training, under-sampling the images scanned by the magnetic resonance according to the under-sampling mask to obtain training data;

[0052] In the process of magnetic resonance scanning imaging, scanning data, that is, full-sampled K-space data, is obtained. Magnetic resonance scanners need to sample scan data at the Nyquist sampling frequency to generate images to ensure that the data can be recovered ...

Embodiment 2

[0070] Such as Figure 5Shown is a schematic flow chart of the training method for the fast imaging model provided by Embodiment 2 of the present invention. On the basis of the first embodiment, this embodiment also provides a method of embedding the neural network for learning the under-sampling mask into the fast imaging model for iterative training to realize the learning of the under-sampling mask. The method specifically includes:

[0071] S210. During each model iteration training, under-sampling the images scanned by the magnetic resonance according to the under-sampling mask to obtain training data;

[0072] In related technologies, a fast imaging model constructed through deep learning can generate images from under-sampled data. If the imaging effect is not good, the fast imaging model parameters can be optimized through multiple iterations of training. However, no matter how the fast imaging model is optimized, the undersampled data of the input model is always o...

Embodiment 3

[0089] Such as Figure 7 Shown is a schematic structural diagram of the training device for the rapid imaging model provided by Embodiment 3 of the present invention. On the basis of Embodiment 1 or 2, the embodiment of the present invention also provides a training device 7, which includes:

[0090] The training data generation module 701 is used for undersampling the image scanned by the magnetic resonance according to the undersampling mask during each model iteration training to obtain the training data;

[0091] In an implementation example, during each iterative training of the model, the image scanned by the magnetic resonance is under-sampled according to the under-sampling mask, and when the training data is obtained, the training data generation module 701 includes:

[0092] An undersampling unit, configured to undersample the image scanned by the magnetic resonance according to the undersampling mask to obtain undersampled K-space data;

[0093] A data processing ...

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 belongs to the technical field of magnetic resonance scanning imaging, and provides a quick imaging model training method and device, and a server. The method comprises the steps: carrying out the undersampling of an image obtained through magnetic resonance scanning according to an undersampling mask during each model iterative training, and obtaining training data; inputting the training data into a quick imaging model, performing feature extraction on the training data through N multi-granularity attention modules according to multi-scale information and an attention mechanismof an image, and fusing feature maps extracted by each multi-granularity attention module; carrying out image reconstruction on the fused feature map, and outputting imaging data; reversely calculating a gradient according to the imaging data and a target label so as to update parameters of the quick imaging model and the under-sampling mask through the gradient; and carrying out forward calculation by adopting the updated parameters and the under-sampling mask. According to the embodiment of the invention, the problems that the under-sampling mask cannot be optimized and the imaging effect is poor are solved.

Description

technical field [0001] The invention relates to the technical field of magnetic resonance scanning imaging, in particular to a training method, device and server for a fast imaging model. Background technique [0002] Due to its powerful functions, magnetic resonance imaging can provide rich anatomical and functional information, making magnetic resonance imaging widely used in the medical field. To perform MRI, an MRI scan is performed on a patient in a clinical setting. During scanning, the patient needs to maintain a constant posture for a long time, resulting in poor patient experience. Therefore, the MRI speed needs to be accelerated. In real-world scenarios, MRI scanners need to sample data at the Nyquist sampling frequency to ensure that the data can be recovered completely without distortion. [0003] In the prior art, a fast magnetic resonance imaging method is mainly constructed based on deep learning. The steps of imaging with this method are mainly as follows...

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): G06T11/00G06K9/62A61B8/13
CPCG06T11/008A61B8/13G06F18/253
Inventor 王珊珊郑海荣梁皓云刘新梁栋
Owner SHENZHEN INST OF ADVANCED TECH CHINESE ACAD OF SCI
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