Skip to content

A Multiple Self-Similarity Network Based Plug-and-Play Prior for MRI Reconstruction

Notifications You must be signed in to change notification settings

wustl-cig/PnP-MSSN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Multiple Self-Similarity Network (MSSN) Based Plug-and-Play Prior for MRI Reconstruction

This is a Tensorflow implementation of the SPL2020 paper A New Recurrent Plug-and-Play Prior Based on the Multiple Self-Similarity Network

Abstract

Recent work has shown the effectiveness of the plug-and-play priors (PnP) framework for regularized image reconstruction. However, the performance of PnP depends on the quality of image denoiser used as a prior. In this letter, we design a novel PnP denoising prior, called multiple self-similarity net (MSSN), based on the recurrent neural network (RNN) with self-similarity matching using multi-head attention mechanism. Unlike traditional neural net denoisers, MSSN exploits different types of relationships among non-local and repeating features to remove the noise in the input image. We numerically evaluate the performance of MSSN as a module within PnP for solving magnetic resonance (MR) image reconstruction. Experimental results show the stable convergence and excellent performance of MSSN for reconstructing images from highly compressive Fourier measurements.

Usage

python main.py
  • One sample data in fastMRI dataset named 'MRI_Knee_58.mat' is in folder './data/'
  • The settings of plug-and-play algorithm and neural network are in file settings.py
  • The MSSN is trained on BSD500 dataset, and the checkpoints are in folder './models/checkpoints/'
  • 36- and 48-line radial sampling of k-space are used in our experiments.

PnP-BM3D vs. PnP-DnCNN vs. PnP-MSSN

visualExamples

Citation

@article{song2019new,
    title={A New Recurrent Plug-and-Play Prior Based on the Multiple Self-Similarity Network},
    author={Song, Guangxiao and Sun, Yu and Liu, Jiaming and Wang, Zhijie and Kamilov, Ulugbek S},
    journal={IEEE Signal Processing Letters},
    year={2020},
    doi={10.1109/TCI.2019.2893568},
}

About

A Multiple Self-Similarity Network Based Plug-and-Play Prior for MRI Reconstruction

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages