Learned Low Rank

Learned Low Rank Prior: The easiest implementation of the deep unrolling/unfolding network for MRI reconstruction.

• Using only the low rank Casorati matrix property and do not using any CNN Net, just an unfolding version of the algorithm which using ADMM to solve the following optimization problem:

• referred from Keziwen/SLR-Net: Code for our work: "Learned Low-rank Priors in Dynamic MR Imaging" (github.com)

• the paper of SLR-Net is

Ke, Z., Huang, W., Cui, Z. X., Cheng, J., Jia, S., Wang, H., ... & Liang, D. (2021). 
Learned Low-rank Priors in Dynamic MR Imaging. 
IEEE Transactions on Medical Imaging, DOI: 10.1109/TMI.2021.3096218.

The Files of this project

• main.py is the training code, and test.py is the testing code
• model.py is the unfolding network code
• dataset_tfrecord.py is the code for loading data from *.tfrecord files
• WriteTFRecord.py & WriteTFRecord_singleCoil.py are the codes for making the *.tfrecord files from OCMR dataset BY MYSELF
• requirements.txt: pip install -r requirements.txt install all the requirements

NOTE

• The derivation of the algorithm which using ADMM to solve the low-rank optimization problem can be find in derivation.md