Possible error in pre-processing step in Osprey for HERMES data

[PGMM]

Hi Georg and Helge, (and anyone else currently working on Osprey)

I have been successful writing a job file for a series of acqusitions (multi-subjects, multi-sessions), and running it (using the GUI for now, wil try the CLI method later). However, for one session in one subject, the pre-processing alingment step seems to have failed. (See image attached) Basically, the pre-allinged spectra look more allinged then the post allinged. The chemical shift/drift plot also shows this. This of course means the model fit, and other steps don’t really work properly for this data point.

Any idea why this may have happened?
Any way to salvage this data set? Turn of alingment perhaps?

[Helge]

Hi @PGMM,

quick questions:

1. Are you running version 2.4.0? I have seen similar issues in the prior version, which were related to the initial drift estimation.

2. Are the other datasets also upside down, and is it working for these? If this is only happening for this dataset you could try to flip the raw data before running OspreyProcess. You can use:

MRSCont.raw{7,1}=op_ampScale(MRSCont.raw{7,1},-1);

Best,
Helge

[PGMM]

Hi Helge, runnning 2.3.0 - did not realise it had been updated to 2.4. I will update via git hub and re-run. (before trying to data “flip”)

Note, this did not happen to any of the other data sets, even if “upside down”. (example below)

[PGMM]

Upgrading to the latest version fixed the issue Helge. Thanks for that. I will keep a closer eye on the versions going forward.