Low-dose whole-body PET image enhancement method based on self-inverse convolution generative adversarial network

An image enhancement, low-dose technology, applied in image enhancement, biological neural network model, image analysis, etc., can solve problems such as image contrast reduction, image noise increase, affecting doctor's diagnosis, etc., to maintain image contrast and improve robustness , the effect of preserving image details and contrast

Active Publication Date: 2020-12-29
浙江明峰智能医疗科技有限公司
View PDF4 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, reducing the injection dose will lead to increased image noise, decreased contrast, and may eventually affect the doctor's diagnosis
Traditional image post-processing methods including BM3D, NLM, etc. usually lead to problems such as image over-smoothing and image contrast degradation

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
  • Low-dose whole-body PET image enhancement method based on self-inverse convolution generative adversarial network
  • Low-dose whole-body PET image enhancement method based on self-inverse convolution generative adversarial network
  • Low-dose whole-body PET image enhancement method based on self-inverse convolution generative adversarial network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0045] refer to Figure 1 to Figure 9 A specific implementation of a low-dose whole-body PET image enhancement method based on a self-inverse convolution generative adversarial network of the present invention will be further described.

[0046] The low-dose whole-body PET image enhancement method based on the self-inverse convolution generation confrontation network uses the collected low-dose PET images and full-dose PET images to train the model, uses the low-dose PET images and training results to test the model, and saves the test results to obtain Low-dose PET image enhancement results.

[0047] like Figure 5 As shown, the above training process is:

[0048] (1) Acquisition of low-dose and full-dose PET images;

[0049] (2) Divide the low-dose and full-dose PET image datasets into training, validation, and test sets;

[0050] (3) Normalize the low-dose and full-dose PET images between 0 and 1 in the training set and validation set;

[0051] (4) In the training set ...

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 low-dose whole-body PET image enhancement method based on a self-inverse convolution generative adversarial network, and the method comprises the steps: training a model through employing a collected low-dose PET image and a collected full-dose PET image, testing the model through the low-dose PET image and a training result, storing the testing result, and obtaining a low-dose PET image enhancement result. According to the method, the mutual conversion between the low-dose PET image and the full-dose PET image is realized by using the self-inverse neural network, andthe trained network has the self-inverse property, so that the noise in the images can be effectively reduced, the image contrast can be effectively maintained, and the robustness of the model is effectively improved. The network model is simple in structure and high in calculation efficiency, noise in the image can be effectively reduced, image details and contrast are effectively reserved, andthe quantitative error is lower than 10%.

Description

technical field [0001] The present invention relates to the technical field of medical imaging equipment, and more specifically relates to a low-dose whole-body PET image enhancement method based on self-inverse convolution generating an adversarial network. Background technique [0002] Positron Emission Tomography (Positron Emission Tomography) is a relatively advanced clinical examination imaging technology in the field of nuclear medicine. By injecting radiotracer drugs into the human body, a pair of gamma rays produced by the decay of radionuclide in the drug can obtain the distribution map of the radioactive drug in the human body. General radioactive tracers are selected from the substances necessary for the metabolism of biological life, such as: glucose, protein, nucleic acid, fatty acid, labeled with short-lived radionuclides (such as 18F, 11C, etc.). The aggregation in metabolism reflects the situation of life metabolic activities, so as to achieve the purpose of...

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): G06T5/00G16H30/20G06N3/04G06N3/08
CPCG06T5/007G06N3/08G16H30/20G06T2207/10104G06T2207/20081G06N3/045
Inventor 周龙叶宏伟
Owner 浙江明峰智能医疗科技有限公司
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