Unlock instant, AI-driven research and patent intelligence for your innovation.

Multi-channel MRI reconstruction parallel imaging technology based on PI-GAN

An image and magnetic resonance image technology, applied in the field of medical image processing, can solve the problems of artificial reconstruction misjudgment, missed diagnosis, and traditional reconstruction methods cannot meet the real-time performance of medical image reconstruction.

Inactive Publication Date: 2021-12-03
HENAN UNIVERSITY OF TECHNOLOGY
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0009] In order to solve the problem of misjudgment and missed diagnosis in manual reconstruction in medical image reconstruction, and the traditional reconstruction method cannot meet the real-time performance of medical image reconstruction, the present invention provides a method for automatically and accurately reconstructing brain magnetic resonance images using a method of generative confrontation network

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
  • Multi-channel MRI reconstruction parallel imaging technology based on PI-GAN
  • Multi-channel MRI reconstruction parallel imaging technology based on PI-GAN

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0017] In order to verify the reconstruction performance of the present invention in 3D brain magnetic resonance images, we selected the BraTS public dataset for training, verification and testing.

[0018] Step 1: 40 brain magnetic resonance images were preprocessed, and FreeSurfer software was used to realize the affine and normalization of the data.

[0019] Step 2: Train the improved U-shaped convolutional network in the Pycharm development software, use the Adam optimizer to optimize, and set the learning rate to 0.0001. The batch size is set to 2, every 200 training is an epoch, and the training is 500 epoch. Randomly divide 200 images into a ratio of 8:1:1 for training, validation and testing. Train the network until the network converges and stop training.

[0020] Step 3: This experiment uses the untrained (Unseen) data in the BraTS public data set for testing, and the experimental results are evaluated by coincidence coefficients (NMSE, PSNR, SSIM).

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

An existing brain nuclear magnetic resonance image (MRI) has the problems of overlong sampling time, poor image quality and the like in imaging reconstruction, and the effect of eliminating aliasing artifacts in the reconstructed image is poor. In order to improve the quality and efficiency of brain MRI reconstruction and assist in improving the success rate of medical diagnosis, a multi-channel MRI reconstruction parallel imaging based on PI-GAN is provided. Multi-channel magnetic resonance imaging reconstruction is accelerated by combining parallel imaging and generative adversarial network architecture, and thus, image details can be better reserved in the reconstruction process. By using antagonistic loss and pixel-level loss in the image and frequency domain, it is possible to efficiently reconstruct a multi-channel MR image at a low noise level and to improve the structural similarity of the reconstructed image. Using multi-coil information and providing 'end-to-end' reconstruction may reduce aliasing artifacts and restore tissue structures. According to the invention, rapid and accurate reconstruction can be realized.

Description

technical field [0001] The invention belongs to the technical field of medical image processing, and in particular relates to a method for reconstructing brain magnetic resonance images. Background technique [0002] Magnetic resonance imaging (MRI) and computed tomography (CT) both use non-invasive imaging methods to display high-definition images of patients without skull artifacts. CT and MRI images are generally used in the diagnosis of brain diseases. CT images can clearly observe subtle changes in bone tissue morphology, but are not sensitive to changes in soft tissues. Medical images generated by MRI have high contrast and spatial resolution of soft tissues, so MRI is generally used for brain functional imaging. MRI is an indispensable auxiliary diagnostic tool in the diagnosis of clinical brain diseases, and provides an important basis for doctors' clinical diagnosis. [0003] Magnetic resonance imaging (MRI) is different from CT imaging, which has radiation damage...

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): G06T7/00G06T7/10G06N3/04G06N3/08
CPCG06T7/0012G06T7/10G06N3/084G06T2207/10088G06T2207/20081G06T2207/20084G06T2207/20132G06N3/045
Inventor 赵祥张鑫杨铁军李冰洁
Owner HENAN UNIVERSITY OF TECHNOLOGY