Fieldmap gone wrong #22
Description
Hi,
We have been working on obtaining QSM maps using SEPIA, for which we used ROMEO for the unwrapping step. For a portion of the subjects we obtain fieldmaps that are noise, which also resulted in bad QSM maps: 
.
Our data is MEGRE data with monopolar readout. At first, we had the Eddy correction set "on". After putting this setting to "off":
- A couple of bad fieldmaps (and QSM maps) improved.
- A couple of bad fieldmaps stayed bad (sub-025).
- Some fieldmaps became bad that were not bad before (sub-096).
Using the SEPIA toolbox, we looked at other methods to obtain the fieldmaps and they managed to give us good fieldmaps. The configuration settings were as follows:
- For ROMEO:
algorParam=struct();
% General algorithm parameters
algorParam.general.isBET = 0;
algorParam.general.isInvert = 0;
algorParam.general.isRefineBrainMask = 1;
% Total field recovery algorithm parameters
algorParam.unwrap.echoCombMethod = 'ROMEO total field calculation';
algorParam.unwrap.offsetCorrect = 'On';
algorParam.unwrap.mask = 'SEPIA mask';
algorParam.unwrap.qualitymaskThreshold = 0.5;
algorParam.unwrap.useRomeoMask = 0;
algorParam.unwrap.isEddyCorrect = 0;
algorParam.unwrap.isSaveUnwrappedEcho = 0;
algorParam.unwrap.excludeMaskThreshold = 0.3;
algorParam.unwrap.excludeMethod = 'Weighting map';
- Method 2 (MEDI + SEGUE):
algorParam = struct();
algorParam.general.isBET = 0 ;
algorParam.general.isInvert = 0 ;
algorParam.general.isRefineBrainMask = 0 ;
% Total field recovery algorithm parameters
algorParam.unwrap.echoCombMethod = 'MEDI nonlinear fit' ;
algorParam.unwrap.unwrapMethod = 'SEGUE' ;
algorParam.unwrap.isEddyCorrect = 0 ;
algorParam.unwrap.excludeMaskThreshold = 0.5 ;
algorParam.unwrap.excludeMethod = 'Weighting map' ;
- Method 3 (Optimum weights + SEGUE):
algorParam = struct();
algorParam.general.isBET = 0 ;
algorParam.general.isInvert = 0 ;
algorParam.general.isRefineBrainMask = 0 ;
% Total field recovery algorithm parameters
algorParam.unwrap.echoCombMethod = 'Optimum weights' ;
algorParam.unwrap.unwrapMethod = 'SEGUE' ;
algorParam.unwrap.isEddyCorrect = 0 ;
algorParam.unwrap.isSaveUnwrappedEcho = 1 ;
algorParam.unwrap.excludeMaskThreshold = 0.5 ;
algorParam.unwrap.excludeMethod = 'Weighting map' ;
Do you have an idea what causes these bad fieldmaps and how we could obtain good fieldmaps using ROMEO for the maps that were bad. Here, I have data of 2 subjects (sub-025 & sub-096) that you could look at.
Thank you in advance!
Best,
Marianka