|
Click to view the video of the phase 1 Spectral Fitting Challenge Results & Discussion
from the ISMRM Workshop on MR Spectroscopy, 15 August 2016, Lake
Constance, Germany.
MRS Fitting Challenge
Fitting of the MRS data plays an important role in the quantification of metabolite concentrations. A number of commercial and home-built packages are available and used by the MRS community to fit spectra.
Our aim is to engage the MRS community in fitting of sample datasets that will help in comparing fitting packages and will allow identification of problems and improvements that could be made to make fitting more accurate and reproducible.
Therefore, we generated synthetic data. After the initial review of the data from the first phase of fitting challenge, we are reopening the fitting challenge for a second phase to have a better representation of fitting programs.
Description of the data:
28 datasets are provided in three formats:
- text file (column 1: real part of water suppressed FID; column 2: imaginary part of water suppressed FID; column 3: real part of water FID; column 4: imaginary part of water FID)
- LCModel (.RAW and .h2o)
- jMRUI (ASCII txt formats for both water suppressed [WS] and water spectra)
Also, the metabolite basis set with macromolecular baseline is provided in these four formats:
- text file (column 1: real part of FID; column 2: imaginary part of FID) for each metabolite and macromolecular baseline
(MMBL)
- text file with no reference peak at 0 ppm
(column 1: real part of FID; column 2: imaginary part of FID)
for each metabolite and MMBL
- LCModel (.RAW) for each metabolite and
MMBL and .BASIS
- jMRUI (ASCII txt file
format) for each metabolite and MMBL
An additional dataset, Cr_10mM_test_water_scaling (text, LCModel, jMRUI), is provided for testing water scaling. This dataset contains a Cr spectrum which - with correct water scaling - should come out to be 10 mM.
metabolite concentration = (relaxation corrected water concentration) * (metabolite relaxation correction) * metabolite area / water area
relaxation corrected water concentration = 1 mol / 18.015 g * 0.6 * 0.78 g/ml * exp(-30 ms/110 ms) + 1 mol/18.015 g * 0.4 *0.65 g/ml * exp(-30 ms/80 ms) = 29697 mM
metabolite relaxation correction = 1/(exp(-30 ms/160 ms)) = 1.206
The following parameters describe the simulated data:
Frequency = 123.2 MHz
Sequence: PRESS
TE = 30 ms, TE1 = 11 ms, TE2 = 19 ms
TR >> T1’s
Spectral width = 4000 Hz
Number of points = 2048
Tissue content: GM = 60%, WM = 40%
Water content: GM = 0.78 g/ml, WM = 0.65 g/ml
T2 of water: GM = 110 ms, WM = 80 ms
T2 of metabolites = 160 ms
The datasets were generated using ideal pulses except for MMBL which was obtained experimentally. Glucose was simulated using both anomers, 0.36 alpha-glucose and 0.64 beta-glucose. The datasets may contain some or all of the metabolites listed among the basis spectra. In addition, lipid resonances and artifacts may be present.
How to participate in the challenge:
- You can participate in the challenge with any one or multiple of the following three levels of involvement:
- Using everything provided in the challenge (i.e., basis
sets and MMBL)
- Simulating your own basis set, but using macromolecular information provided
- Simulating your own basis set and using your own approach for macromolecules/baseline.
- You download all the spectra for fitting
and the spreadsheet for
the results.
- You fit the spectra with one or multiple fitting approaches, involving one or multiple fit programs and/or strategies than may or may not include manual optimization steps or parameter choices (all of which you will specify in your results file).
- Finally, you submit the results from fitting using the
spreadsheet provided and also visual displays of the fit in either PS, PDF, jpg, tiff formats (preferably zoomed in to 0.0 to 4.3 ppm). For each variant of fitting you perform, please upload the results and spectra separately.
- Use following naming convention for the spreadsheets and figures: dataset#_(eight letter code for institution)_(two letter initials)_(running number based on number of different fits submitted).xlsx (e.g., dataset1_Minnesot_MM_1.xlsx).
- Please include a spreadsheet with the concentration and CRLB that you get for Cr in the dataset Cr_10mM_test_water_scaling using following spreadsheet format: dataset29_Cr_Minnesot_MM_1.xlsx
- Please make sure that you provide information in the correct format. Please do not change the order of metabolites when reporting information.
Submission Deadline has Passed
Organizing committee contacts:
Malgorzata Marjanska: gosia@cmrr.umn.edu
Roland Kreis:roland.kreis@insel.ch
The fitting challenge prepared by Malgorzata Marjanska, Dinesh Deelchand, and Roland Kreis.
|