Bruker PRESS (ParaVision 3.5)

[peter-mri]

I have some MRS data from Bruker (ParaVision360 v 3.5), PRESS sequence. Both single coil and array.
Either way I'm getting this error:

spec2nii bruker 8/pdata/1/2dseq

Traceback (most recent call last):
File "/home/peter/fsl/lib/python3.11/site-packages/spec2nii/bruker.py", line 81, in yield_bruker
d.query(queries)
File "/home/peter/fsl/lib/python3.11/site-packages/brukerapi/dataset.py", line 713, in query
raise FilterEvalFalse
brukerapi.exceptions.FilterEvalFalse: FilterEvalFalse

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/home/peter/fsl/bin/spec2nii", line 10, in
sys.exit(main())
^^^^^^
File "/home/peter/fsl/lib/python3.11/site-packages/spec2nii/spec2nii.py", line 701, in main
spec2nii(*args)
File "/home/peter/fsl/lib/python3.11/site-packages/spec2nii/spec2nii.py", line 297, in init
args.func(args)
File "/home/peter/fsl/lib/python3.11/site-packages/spec2nii/spec2nii.py", line 670, in bruker
self.imageOut, self.fileoutNames = read_bruker(args)
^^^^^^^^^^^^^^^^^
File "/home/peter/fsl/lib/python3.11/site-packages/spec2nii/bruker.py", line 44, in read_bruker
for data, orientation, dwelltime, meta, name in yield_bruker(args):
File "/home/peter/fsl/lib/python3.11/site-packages/spec2nii/bruker.py", line 83, in yield_bruker
raise ValueError(f'Bruker dataset {d.path} is not suitable for conversion to mrs_nifti')
ValueError: Bruker dataset 8/pdata/1/2dseq is not suitable for conversion to mrs_nifti

I'm using spec2nii version 0.7.4 (can't figure how to update to current version...)

I'm happy to share the data if it helps.
Cheers,
Peter

[wtclarke]

Hi Peter,

Yes please to data, though I can't promise any quick fixes.

Also, is this the new version of PV360? I've heard that quite a few things have changed.

Will

[peter-mri]

Thanks Will! There's no major rush. I've attached file from PRESS data from a single channel (volume) coil. These are from a "fat" phantom, so don't look for familiar peaks...
We got version 3.5 a few months ago, so it should be out for at least half a year, if not longer. Unfortunately there seems to be changes from version to version... Unfortunately I don't know the details of differences between various v3.x versions.
There's no fid file (instead, there's a 64bit fid_proc.64). I've also loaded the data in Bruker's TopSpin software for viewing, which creates additional files (in pv2tsdata), those may not always be available...

The one piece of information from Bruker forum that may or may not help:
Group Delay parameter which is normally in the ACQP file as ##$GRPDLY. In PV360 this parameter is set to -1. The correct value (76.0833) is found in the ACQ_RxFilterInfo parameter in the ACQP file as:
##$ACQ_RxFilterInfo=( 2 )
(76.0833333333333, 0, 8, 19.8698903048866, 31) (0, 15.0833333333333, 27,19.8698903048866, 31)
The complex fid data must be shifted by the Group Delay (=round(76.0833)) indices and zeros appended to the array to keep the fid size fixed to the original value.

8.zip

[Microdeep-ZL]

Hi, I have problem converting 13C NSPECT spectra, too. I'm using ParaVision 6.0.1
It's the same error message "Bruker dataset 2dseq is not suitable for conversion to mrs_nifti"
I'm attempting to convert my 2dseq into nifti-mrs manually now. If you need the sample data to diagnose the problem, I can attach it here as well.

[Microdeep-ZL]

It seems the problem arises because one of the three conditions ["@type=='2dseq'", "@is_spectroscopy==True", "@is_complex==True"] is not met. In my case, my 2dseq is Magnitude, not complex. Is it possible to fix this?

[Microdeep-ZL]

unfortunately, after I reconstructed it to complex, I still got an error. But this time the error is different

$ spec2nii bruker -m 2DSEQ 2dseq
Traceback (most recent call last):
  File "/home/nmr/.conda/envs/fslmrs/bin/spec2nii", line 10, in <module>
    sys.exit(main())
             ^^^^^^
  File "/home/nmr/.conda/envs/fslmrs/lib/python3.12/site-packages/spec2nii/spec2nii.py", line 701, in main
    spec2nii(*args)
  File "/home/nmr/.conda/envs/fslmrs/lib/python3.12/site-packages/spec2nii/spec2nii.py", line 297, in __init__
    args.func(args)
  File "/home/nmr/.conda/envs/fslmrs/lib/python3.12/site-packages/spec2nii/spec2nii.py", line 670, in bruker
    self.imageOut, self.fileoutNames = read_bruker(args)
                                       ^^^^^^^^^^^^^^^^^
  File "/home/nmr/.conda/envs/fslmrs/lib/python3.12/site-packages/spec2nii/bruker.py", line 44, in read_bruker
    for data, orientation, dwelltime, meta, name in yield_bruker(args):
                                                    ^^^^^^^^^^^^^^^^^^
  File "/home/nmr/.conda/envs/fslmrs/lib/python3.12/site-packages/spec2nii/bruker.py", line 84, in yield_bruker
    yield from _proc_dataset(d, args)
  File "/home/nmr/.conda/envs/fslmrs/lib/python3.12/site-packages/spec2nii/bruker.py", line 121, in _proc_dataset
    d = FrameGroupMerger().merge(d, 'FG_COMPLEX')
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/nmr/.conda/envs/fslmrs/lib/python3.12/site-packages/brukerapi/mergers.py", line 21, in merge
    raise ValueError(f'Dataset does not contain {fg} frame group')
ValueError: Dataset does not contain FG_COMPLEX frame group

[nnitta-nobu]

Dear Dr.Will,

I am conducting research using a Bruker high-field MRI at a research institute in Japan.
Our primary focus is brain imaging, but we also conduct measurements on other tissues.

We recently upgraded our software to ParaVision 360 v3.6.
I am experiencing the same error as others, but I hope that FSL-MRS will be compatible with Bruker systems.

If needed, I am happy to provide data for further investigation.

Best regards,
Nobu

[cpugonna]

Hi all,
Just checking in to see if there has been any progress with understanding the scope of work required to make spec2nii work with Bruker PV360.3.5. I have some colleagues who are working on SVS on Bruker Biospec 7T and who would like to put their data into NIFTI-MRS format. My background is primarily in Human MRI and definitely not in spectroscopy so I am a little out of my depth. I would be happy to provide effort to make this work with you all but there will be some learning/familiarization required on my part.
Best regards
Chidi

[wtclarke]

@peter-mri @cpugonna @nnitta-nobu @Microdeep-ZL

We have a branch https://github.com/wtclarke/spec2nii/tree/enh/bruker that you might want to try. We'll also release it in the main package in due course, after some collaborators have had a quick test.