Low Rank Denoising Tools

Tools for low-rank denoising of MRSI.

This package contains functions to carry out:

• Global and local spatio-temporal low-rank denoising ⁴
• Global and local LORA ⁴
• Linear-predictability denoising ^{1, 4}
• SURE optimised local soft thresholding (SURE-SVT) ²
• SURE optimised local hard thresholding (SURE-SVHT) ⁶

Command line script

Spatio-temporal

mrsi_denoise st [--mask MASK] [-r RANK | -mp] [-p PATCH PATCH PATCH] [-s STEP] input output noise [noise]

LORA

mrsi_denoise lora ...

LP

mrsi_denoise lp ...

SVT/SVHT

mrsi_denoise svt... / mrsi_denoise svht ...

Python library

Denoising functions can be found in the mrs_denoising.denoising module.

Citation

If you use these tools please cite:
Clarke WT and Chiew M. ISMRM 2021

References

1: Cadzow JA. Signal enhancement-a composite property mapping algorithm. IEEE Transactions on Acoustics, Speech, and Signal Processing 1988;36:49–62 doi: 10.1109/29.1488.

2: Candès EJ, Sing-Long CA, Trzasko JD. Unbiased Risk Estimates for Singular Value Thresholding and Spectral Estimators. IEEE Transactions on Signal Processing 2013;61:4643–4657 doi: 10.1109/TSP.2013.2270464.

3: Chen Y, Fan J, Ma C, Yan Y. Inference and uncertainty quantification for noisy matrix completion. PNAS 2019;116:22931–22937 doi: 10.1073/pnas.1910053116.

4: Nguyen HM, Peng X, Do MN, Liang Z. Denoising MR Spectroscopic Imaging Data With Low-Rank Approximations. IEEE Transactions on Biomedical Engineering 2013;60:78–89 doi: 10.1109/TBME.2012.2223466.

5: Song J, Xia S, Wang J, Patel M, Chen D. Uncertainty Quantification for Hyperspectral Image Denoising Frameworks based on Low-rank Matrix Approximation. arXiv:2004.10959 [cs, eess] 2021.

6: Ulfarsson MO, Solo V. Selecting the Number of Principal Components with SURE. IEEE Signal Processing Letters 2015;22:239–243 doi: 10.1109/LSP.2014.2337276.