LWFIT code for GABA quantification from MEGA-PRESS magnetic resonance spectra. It is based on a peak integration approach in the frequency domain to estimate the concentration of the metabolites.
The code requires matlab to run. It was originally developed using matlab 2017b and has also been tested with Matlab R2021a (9.10.0.1602886) under Linux. To use the code, start with the Matlab live scripts Notebook-XX.mlx which show how to use the individual functions. Further details are in the comments of the relevant functions.
LWFIT aligns the edit-off and difference spectra so that the main NAA peak is precisely at 2.01 ppm and then calculates the areas for
- NAA: 1.91 - 2.11 ppm, difference,
- Cr: 2.90 - 3.10 ppm, edit-off,
- GABA: 2.90 - 3.12 ppm, difference,
- Glu and Gln, GLX1: 2.25 - 2.45 ppm, difference,
- and GLX2: 3.65 - 3.85 ppm, difference.
It uses the real frequency-domain spectra and numerically integrates the peaks using piecewise-linear functions over the indicated fixed ppm ranges, selected to minimise contamination from other signals. Lorentzian, Gaussian and spline fitting and several baseline fitting and filtering methods were explored but initial testing indicated that numerical integration with minimal pre-processing yielded the most accurate estimation results of the methods considered. While more aggressive noise filters aesthetically improve the quality of the spectra, their application tends to exacerbate underestimation of GABA, especially for weak signals, suggesting that filtering eliminates some of the GABA signal present in the data. So we only perform zero-filling (4096 points) and a frequency shift to align the NAA peak.
The code has been used in the following paper and more details are available there:
CW Jenkins, M Chandler, FC Langbein, SM Shermer. Benchmarking GABA Quantification: A Ground Truth Data Set and Comparative Analysis of TARQUIN, LCModel, jMRUI and Gannet. [arxiv:1909.02163]
The analysis results of LWFIT used in that paper for a specific dataset, using the Matlab live scripts above are in the wiki, with all results computed and also converted to PDF.
The code is developed and maintained on qyber\black at https://qyber.black/mrs/code-lwfit
This code is mirrored at
The mirrors are only for convenience, accessibility and backup.
- Sophie M Shermer, Physics, Swansea University
- Christopher W Jenkins, Physics and Centre for Nanohealth and Clinical Imaging Unit, Swansea University; Cardiff University Brain Research Imaging Centre (CUBRIC)
- Frank C Langbein, School of Computer Science and Informatics, Cardiff University; langbein.org
- Max Chandler, School of Computer Science and Informatics, Cardiff University
For any general enquiries relating to this project, send an e-mail.
SM Shermer, FC Langbein, CW Jenkins, M Chandler. Code - LWFIT. Version 1.0. FigShare, Software. May 2021. [DOI:10.6084/m9.figshare.14540757.v1] [DEV:https://qyber.black/mrs/code-lwfit] [MIRROR:https://github.com/qyber-black/Code-LWFIT/tree/v1.0]
Copyright (C) 2019-2021, Sophie M Shermer, Swansea University
This program is free software: you can redistribute it and/or modify it under the terms of the GNU Affero General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero General Public License for more details.
You should have received a copy of the GNU Affero General Public License along with this program. If not, see https://www.gnu.org/licenses/.