Folder contains codes for TCI submission 2022. We used same parameters as author provided.
=================================================================================
Comparison of ktslr, L+S-Otazo, L+S-Lin, altGDMin-MRI and altGDMin-MRI2.
To generate Table II results:
This code requires the Matlab version of the Michigan Image Reconstruction Toolbox (MIRT) from [http://web.eecs.umich.edu/~fessler/code/index.html] Please set up MIRT before running the examples. Run the mirt-main/setup.m: L+S-Lin code requires the Matlab version of the Michigan Image Reconstruction Toolbox (MIRT). Run Main_files_retrospective.m: This run all the 5 algorithms and calculate NMSE (Normalized Mean Square Error), Recon Time required and save these results in Comparison_error.txt [Error(Time)] amd Comparison_sim.txt [sim(Time)].
===================================================================================
For datasets and questions contact email sbabu@iastate.edu, namrata@iastate.edu.
If you are using our code please cite our paper: S. Babu, S. S. Nayer, S. G. Lingala and N. Vaswani, "Fast Low Rank Column-Wise Compressive Sensing For Accelerated Dynamic MRI," ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2022, pp. 1346-1350, doi: 10.1109/ICASSP43922.2022.9747549. : Babu S, Lingala SG, Vaswani N. Fast low rank column-wise compressive sensing for accelerated dynamic MRI. IEEE transactions on computational imaging. 2023 Apr 17.