I am processing HERMES data in Gannet 3.5.3 and have a few questions about specific cases. I want to understand if these are cases that can be corrected by reprocessing with different settings, or some other way you might recommend.
All examples below are from cerebellum using the alignment SpecRegHERMES.
CASE 1 - Good GABA peak but flipped
The GABA peak/model appears to have the expected shape, but it is inverted, and the quantified GABA+/Water value is negative.
I noticed that this subject does not show up in the GANNETLoad output, i was able to process this dataset in osprey so i don’t think there is something wrong with the dataset per se.
CASE 1: In GannetPreInitialise.m, try setting ON_OFF_order to 'ABCD'. If that doesn’t work, try 'CBAD'.
CASE 2: What was the error that was thrown when you tried to load the data?
CASE 3: That is a model fitting failure for the GABA peak. This one is a little trickier to address because the dataset is of good quality, so it feels unfair to discard it outright. That said, if you have a large sample size, removing one GABA measurement data point won’t (shouldn’t) change things. Otherwise, we would need to optimize the fitting routine, and that would mean refitting all your data. And that requires a little trial-and-error.
