I am new to the MRS field, and I am trying out Osprey. We scanned some oil phantoms with a 3T Siemens machine. I was able to load and preprocess .raw files to a get a frequency spectrum. My current goal is to get some sort of quantification via fitting using lipid basis sets.
However, I get this error in my workspace when I try to process for quantifying Lip09, Lip13, and Lip20.
Osprey is pretty heavily geared towards processing and modeling 1H brain spectra, so it’s certainly not the ideal solution. The parametrized lipid signals are also just taken straight from the default way that LCModel describes in-vivo lipids, which is a) probably not appropriate for phantom data (too broad), b) is only a really super-crude approximation of triglyceride systems.
May I ask what the exact goal of your study will be? Depending on the aims, you might need to explore more sophisticated ways of modeling the signals. Like, are you trying to determine saturation degrees? Are you scanning the body or in the brain?
I’m attaching some references for further reading that have helped me understand the ways in which different triglycerides will give rise to different signals and how that knowledge can be used to infer useful stuff:
Sorry for no getting back to you earlier.
I attached a link to an old 1975 paper that helped me in my attempts to simulate lipid spectra.
I never could get a good result for the spins close to the carboxyl groups, but the rest worked pretty good using information from this paper.
So I am keen to see better solutions.
