support for DTI

[grandjeanlab]

I would like to see implemented DTI support in addition to fMRI / structural.
Can provide test dataset

[TheChymera]

Please archive and send me a minimal dataset (2-3 scan sessions, raw Bruker directories) as a .zip, and declare here that you agree to have it made publicly available as part of a CC-BY-licensed example data archive. Currently this is the samri_bindata archive, which will list you, me, and possibly others as collective authors.

[TheChymera]

@joanesg DWI support should be available after 9abbe2f :

chymera@zenhost ~/zsofia $ ag acq- 20180515_092658_ZK_SCI_TG_B10_20180515__1_1/ScanProgram.scanProgram 
19065:      <displayName>acq-FISPhor (E20)</displayName>
19810:      <displayName>acq-FISPcor (E21)</displayName>
20589:      <displayName>acq-T1mapRARE (E34)</displayName>
21537:      <displayName>acq-T2mapMSME (E23)</displayName>
22926:      <displayName>acq-T1mapRAREcor (E24)</displayName>
23983:      <displayName>acq-T2mapMSMEcor (E25)</displayName>
26009:      <displayName>acq-PDT2MSMEhor (E26)</displayName>
26841:      <displayName>acq-PDT2MSMEcor (E27)</displayName>
29457:      <displayName>acq-DTIEPI30dirsat (E41)</displayName>
31163:      <displayName>acq-T1FLASH3Diso (E42)</displayName>
chymera@zenhost ~/zsofia $ SAMRI bru2bids -s '{"acquisition":["FISPhor","FISPcor","T1mapRARE","T2mapMSME","T2mapMSMEcor","T1mapRAREcor","PDT2MSMEhor","PDT2MSMEcor","T1FLASH3Diso"]}' -d '{"acquisition":["DTIEPI30dirsat"]}' .
chymera@zenhost ~/zsofia $ tree bids/
bids/
├── dataset_description.json
└── sub-ZKB10
    └── ses-1
        ├── anat
        │   ├── sub-ZKB10_ses-1_acq-FISPcor.json
        │   ├── sub-ZKB10_ses-1_acq-FISPcor.nii
        │   ├── sub-ZKB10_ses-1_acq-FISPhor.json
        │   ├── sub-ZKB10_ses-1_acq-FISPhor.nii
        │   ├── sub-ZKB10_ses-1_acq-PDT2MSMEcor_T2w.json
        │   ├── sub-ZKB10_ses-1_acq-PDT2MSMEcor_T2w.nii
        │   ├── sub-ZKB10_ses-1_acq-PDT2MSMEhor_T2w.json
        │   ├── sub-ZKB10_ses-1_acq-PDT2MSMEhor_T2w.nii
        │   ├── sub-ZKB10_ses-1_acq-T1FLASH3Diso_T1w.json
        │   ├── sub-ZKB10_ses-1_acq-T1FLASH3Diso_T1w.nii
        │   ├── sub-ZKB10_ses-1_acq-T1mapRAREcor_T1w.json
        │   ├── sub-ZKB10_ses-1_acq-T1mapRAREcor_T1w.nii
        │   ├── sub-ZKB10_ses-1_acq-T1mapRARE_T1w.json
        │   ├── sub-ZKB10_ses-1_acq-T1mapRARE_T1w.nii
        │   ├── sub-ZKB10_ses-1_acq-T2mapMSMEcor_T2w.json
        │   ├── sub-ZKB10_ses-1_acq-T2mapMSMEcor_T2w.nii
        │   ├── sub-ZKB10_ses-1_acq-T2mapMSME_T2w.json
        │   └── sub-ZKB10_ses-1_acq-T2mapMSME_T2w.nii
        └── dwi
            ├── sub-ZKB10_ses-1_acq-DTIEPI30dirsat.json
            └── sub-ZKB10_ses-1_acq-DTIEPI30dirsat.nii

4 directories, 21 files

[grandjeanlab]

DWI/DTI images should be accompanied with bvec and bval, else useless. This is seemingly not the case in the tree output shown above.

See 8.3.4 Diffusion imaging data  in http://bids.neuroimaging.io/bids_spec.pdf

bvals can be found under PVM_DwEffBvals in the corresponding method file for the scan.
bvec are found under PVM_DwGradRead, PVM_DwGrandPhase, and PVM_DwGradeSlice in the same file.

Let me know when this is done. I can help you check if the bvec are in the correct order by looking at fiber orientations following FSL's dtifit function.