Basis set needed: Philips 3T MEGAPress MMsuppr. GABA80

Analysis / Modeling Basis Sets
1

Dear all,
I’m currently taking baby steps in Osprey, and in order to analyse my own data I would like to ask the community

  1. Is there already a basis set to model MEGApress MM-suppressed data (Philips 3T Achieva, 32ch headcoil, GABA w/ TE=80ms, pulses at 1.5 & 1.9 ppm), and
  2. would you be willing to share it? :slight_smile:

I’m not familiar to basis set modelling at all, and having read the related posts it seems far from trivial…

Thank you very much for your help.

Cheers,
Stefanie

2

Hi @Stefanie,

Which sequence do you use? The universal sequence from JHU?

I could share a MM suppressed basis set for the universal sequence and I could also help with possible issues during the data analysis.

Best,
Helge

3

Hi @Helge,
thank you soooo much! Yes, we use JHU sequences only.

By now, I made my way through the included Osprey example data, and all I need to fix is one minor bug during coregistration (I guess it may be search path related, or likewise… so no worries).
In the meantime, I took a look at Jamie Near’s FID-A toolbox which is also very well documented, but it left me puzzled which simulation tool to adapt for an MM-suppressed MEGA-PRESS - so your response really made my day :slight_smile:

Last december’s online GABA workshop was very helpful and inspiring, and so is the whole MRS community! Loads of thanks to all of you :slight_smile: :smiley:

Cheers,
Stefanie

1 Like
4

Hi @MRSatBMH,

No worries!

I’ve just added the basis set into the latest develop branch on GitHub. For the jobFile definition I’d set the co-edited MM model to none:

opts.fit.coMM3              = 'none';
opts.fit.FWHMcoMM3              = 14;

Great to hear that the EDITINGSCHOOL was helpful. And I’m also very excited to see this forum thrive! :smiley:

Helge

5

Hi both,

Be mindful that the JHU patch allows two pulse sets (the default Philips pulses, and then what @Helge referred to as ‘universal’, which are custom pulses that @richardedden cooked up and that we have ported across vendors). Just make sure you’re using the correct basis set that corresponds to the pulse set you’ve used during acquisition.

Best,
Georg