Processing Metabolite Cycling Data

Data Processing
1

Hello experts,

I am new to MRS. For a new protocol, I am using sLASER with metabolite cycling. However, I am running into issues processing these data, mostly due to my lack of experience. I tried FSL-MRS and the practical only provides instruction for processing using water suppressed data. Most of what I can find in published literature that incorporates metabolite cycling is quite vague (“eddy currents were corrected”), so I am really not sure where to start with figuring this out.

I guess my questions are:

  1. Is this possible to do with FSL-MRS? Or is there a better program for processing these data?
  2. Where does one get started with learning how to do this? There is no one at my institution who has used metabolite cycling before (which I recognize now is my hubris at thinking it would not be so difficult to figure out from published literature…). I wanted to try to get the best possible data for this study, and it seems in doing so I have painted myself into a corner.

Any direction would be sincerely appreciated.

2

Hi @marterin , Yes this is possible with FSL-MRS, and don’t worry I’m sure we can help you figure out what to do with your data. Could you tell me what format your data is in (and anything more about the sequence) and what shape it is after conversion with spec2nii? You can run mrs_tools info path/to/data.nii.gz.

3

Wonderful, thank you so much! The data is TWIX obtained using this sequence from CMRR (Spectroscopy - SEMI-LASER (SVS)). Happy to provide more details there if you would like.

This is the output from mrs_tools after conversion:
Data shape (1, 1, 1, 4096, 32, 34)
Dimension tags: [‘DIM_COIL’, ‘DIM_DYN’, None]
Spectrometer Frequency: 123.234553 MHz
Dwelltime (Spectral bandwidth): 8.330E-05 s (12005 Hz)
Nucleus: 1H
Field Strength: 2.89 T

4

Great. And we expect the 34 to actually be 17 of one cycle and 17 of the other cycle? If so, you can use mrs_tools reshape:

 mrs_tools reshape
usage: mrs_tools reshape [-h] --file FILE --shape SHAPE [SHAPE ...] [--d5 D5]
                         [--d6 D6] [--d7 D7] [--output OUTPUT]
                         [--filename FILENAME]

with a command like mrs_tools reshape --file data.nii.gz --shape 32 -1 2 --d6 DIM_DYN --d7 DIM_METCYCLE --filename data_reshaped. You can then run mrs_tools vis data_reshaped.nii.gz --display_dim DIM_METCYCLE --no_mean and you will hopefully see the two cases with the metabolite peaks pointing in different directions. If something looks odd then try the other order mrs_tools reshape --file data.nii.gz --shape 32 2 -1 --d7 DIM_DYN --d6 DIM_METCYCLE --filename data_reshaped

5

The first option worked! Thank you.