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

Arterial spin labeling image synthesis method based on generative adversarial hybrid model network

A technology of arterial spin labeling and mixed models, applied in the field of computer vision, can solve the problems of scarcity of arterial spin labeling image data sets, in-depth research on arterial spin labeling technology is hindered, and arterial spin labeling image data sets have not yet been released. , to achieve the effect of alleviating unstable training, clear details and contours, and small voxel differences

Pending Publication Date: 2022-02-15
NANCHANG UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Although arterial spin-labeled images have many advantages in Alzheimer's disease classification, there is no publicly available arterial spin-labeled image data set for Alzheimer's disease due to the scarcity of arterial spin-labeled image data sets , Deep study of arterial spin labeling technology in Alzheimer's disease hindered

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
  • Arterial spin labeling image synthesis method based on generative adversarial hybrid model network
  • Arterial spin labeling image synthesis method based on generative adversarial hybrid model network
  • Arterial spin labeling image synthesis method based on generative adversarial hybrid model network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] Below in conjunction with accompanying drawing, the present invention will be further described.

[0037] The structure of the generative adversarial hybrid model network used to synthesize arterial spin-labeled images in the present invention is as follows: figure 1 As shown, the area inside the dotted line is the autoencoder, and the area inside the solid line is the generative adversarial mixture model network. First, preprocessed structural MRI images are fed into the encoder. Second, the reparameterized public eigenvectors are input into the decoder to synthesize arterial spin-labeled images. Finally, the synthesized arterial spin-labeled image and the real arterial spin-labeled image are fed into the discriminator. This process can be expressed as:

[0038] Z~Encoder(X)=p(Z|X), Y′~Decoder(Z)=q(Y|Z)

[0039] Among them, X is the structural MRI data, Z is the common eigenvector of the structural NMR data and the arterial spin-labeled data, Y is the real arterial...

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 an arterial spin labeling image synthesis method based on a generative adversarial hybrid model network. The method comprises the following steps: 1, obtaining a structural magnetic resonance image and an arterial spin labeling image which are needed for generation, and executing a preprocessing operation; 2, inputting structural magnetic resonance data and arterial spin labeling data into an automatic coding machine for generation training; 3, inputting the corresponding arterial spin labeling image and an image synthesized by the automatic coding machine into a discriminator for training; 4, repeating the steps 2 and 3, and training a generative adversarial hybrid model network to obtain a trained generator; and 5, using the trained generator to iteratively generate predicted arterial spin labeling data by randomly inputting eight pieces of real structural magnetic resonance data. According to the invention, a synthesized arterial spin labeling image has clearer details and contours, the research of an arterial spin labeling technology in Alzheimer's disease can be promoted, and a high-quality supplementary data set is provided.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to an arterial spin-labeled image synthesis method based on a generative confrontational mixed model network. Background technique [0002] Alzheimer's disease is a common senile disease. In the early stage, the elderly suffering from this disease will have symptoms of forgetfulness, dull eyes, and slow speech. In the late stage, it will affect the sensory nerves and make the elderly lose their normal self-care ability. According to the World Health Organization, one out of every 20 people over the age of 65 will suffer from Alzheimer's disease. Today, Alzheimer's disease has become the fourth leading cause of death for the elderly after cardiovascular disease, cerebrovascular disease and malignant tumors. Therefore, the treatment of Alzheimer's disease is particularly important. With no real cure yet, the real hope for Alzheimer's lies in early intervention. ...

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): G06T17/00G06V10/764G06V10/774G06K9/62G16H50/20G06N3/04
CPCG06T17/00G16H50/20G06N3/045G06F18/217G06F18/241
Inventor 李菲红胡长江黄伟
Owner NANCHANG UNIV
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