QC of MRS data with Osprey, focus on Myo-inositol

Dear reader

I have acquired dlPFC (Isotropic 20 mm voxel) MRS data with a Siemens 3T scanner in group of older adults. I am specifically interested in myo-inositol, but also in choline and creatine. To analyse my data, I runned Osprey (v 2.4.0) with the default settings. However, I’m struggling a bit with the quality control of my data, also specifically for the myo-inositol peaks.

From a previous colleague in the field, I learned that visual inspection is most important.
Besides, there are the reported QC measures by Osprey: Cr_SNR, Cr_FWHM, water_FWHM, residual_water_ampl, freqShift, and relResA. However, I read a lot of papers, and many of them mention CRLBs as measure for quality of a specific metabolite. Is it just my (older) version of Osprey, or are these CRLBs not reported? Can I calculate these myself? Or is it still the best solution to only visually inspect the spectrum, and the specific peak/residuals around Myo-inositol, and not use some numerical cut-off?

Furthermore, I also try to interpret the previous mentioned QC values that Osprey does report. Are there some numerical cut-offs that could be used there, to leave out the data that is so bad it does not need a visual inspection for myo-inositol?
To illustrate:

Cr_SNR - mean = 67, range: 37-93
Cr_FWHM - mean = 7.25, range: 5-13
water_FWHM - mean = 8.18, range: 6-12

I’m still new to the field, so I’d be happy if someone could help me.

Hi! I can’t answer all your questions, but that last I heard, Osprey did not provide CRLB. From my own previous experience, the SNR values seem low but the FWHM are pretty normal, although I know some people recommend >11Hz as a cutoff for linewidth. Hope that’s at least somewhat helpful!