NODDI data in BIDS

[Thomas Shaw]

Hi experts,

I'm trying to convert some DICOMS to BIDS but I'm unsure how my data should look.
I have two acquisitions using NODDI with 33 and 66 diffusion directions - how should this look in BIDS? Or should I be combining these images first? 

Thank!

[James Gholam]

Hi there,

Normally you'd arrange each acquisition into a separate nifti, with each nifti being a time series of volumes (i.e dim[0] = 4). Each file should have a name that represents the acquisition scheme, but both should lay in the same subject/session/modality folder (assuming these were actually the same session). The names could be:

"sub-01_ses-01_acq-33dir_run-01_dwi.nii",  and 

"sub-01_ses-01_acq-66dir_run-01_dwi.nii" 

These should have corresponding .bval and .bvec files specifying the directions and b values, as well as json sidecars. The sidecars would ideally use the "PulseSequenceDetails" key to specify the sequence used in detail. I'd put each file through dcm2niix if possible, with the BIDS sidecar mode on. See what comes out.

Let me know if I can help

James