Description

New flavor challenge!

See this branch: https://github.com/ExploreASL/FlavorDatabase/tree/bug-%231170_GE_Premier_HongKong
BRANCH

GE data Hong Kong has some issues, it is from a new GE Premier scanner.
The ASL deltaM & M0 are correctly concatenated in 4D. Structural module goes fine, if the wrong structural images matrix size is corrected. ASL module goes fine, if the M0 issue below is corrected.

Tasks

• Waiting for the correct studyPar.json settings (now I don't know the ASL protocol settings yet). JAN:No need to wait - all settings are correctly read from DICOM.
• studyPar.M0 = true doesn't work (although this seemingly used to work with older flavors). It should apparently be studyPar.M0Type = Included which will set an intermediate jsonLocal.M0 = true in xASL_imp_NII2BIDS_Subject_DefineM0Type. Something went wrong here recently? Same when testing with older GE flavor release DV22.0 (/FlavorDatabase/GE_PCASL_3Dspiral_22.0LX_1). This leads to "M0Type":"Absent" in rawdata and no m0scan in aslcontext.tsv. If I change this manually to "M0Type":"Included", and replace one deltaM by m0scan. JAN: That's correct, M0 is not a ASL-BIDS parameter and thus should not be provided in studyPar.json. You should provide only ASL-BIDS compliant parameters there. Besides, m0scan and deltaM positions are correctly read from the DICOMs.

HENK: OK then please provide a warning to the user whenever DICOM fields overwrite studyPar.json or vice versa (this should be a general comparison function, you have that already right?

• We get the following warning (also when running older GE flavor release DV22.0). JAN: Will be fixed in Master issue flavor processing #342

Warning: Field EchoTime in xASL structure has a value 0.0528, which is outside of the
recommended range <0.5,500> ms.
In xASL_bids_parms2BIDS (line 101)
In xASL_bids_Dicom2JSON (line 632)
In xASL_imp_DCM2NII_Subject_StoreJSON (line 38)
In xASL_imp_DCM2NII_ConvertScan (line 172)
In xASL_wrp_DCM2NII_Subject (line 126)
In xASL_wrp_DCM2NII (line 45)
In xASL_module_Import (line 147)
In xASL_init_Iteration>runIteration (line 382)
In xASL_init_Iteration (line 125)
In ExploreASL_Import (line 50)
In ExploreASL (line 91)

• The T1w and FLAIR NIfTIs are incorrectly sorted: each left/right part of the matrix is put in a 4rd dim,
  so [0.5*(X Y Z) 2] instead of [X Y Z 1]. JAN: this was reported as a dcm2nii issue already and should be solved in the new dcm2nii version - see Update dcm2nii version #435
• Is xASL_bids_Dicom2JSON fails in a GE deltam-m0scan situation #1133 slightly related? JAN: Yes, both will be fixed in Update dcm2nii version #435
• Anonymize this dataset. JAN: I will anonymize the DICOMs.
• Add formally to flavors. JAN: Yes, added the correct JSONs and DICOMs and set everything up.
• Add references to flavors (ASL). JAN: We can add this one already, as this works nicely, but we need to wait for the structural data.
• Add references to flavors (structural). JAN: We need to wait for the Update dcm2nii version #435 fix to have the perfect consistent flavors.
• Once done, delete the original branch in Flavors.

How to test

Flavor BRANCH

Release notes

Added a GE_RX28.0 to flavors.