Rookie question: preparing fMRI data from scanner to preprocessing

I have dicom data collected from a project. I converted the files from DICOM to NIFTI
Eg. These are some filed I have a finger and vocal task, my NIFTI files contain 4 Task_AP_topup, 4 VOCAL1_ REV_AP and 4 HAND_REV_AP files. I also have 3 T1w and T2wSPC files.

I have seen a lot of tutorials that only have files already prepared. Now I have my own files from a scanner and my rookie question is how do I prepare the files. More specifically do I combine the 4 Nifti files for vocal and hand into single vocal and hand nifty files? what do I do with the “Task_topup” files which was collected right before the hand and vocal tasks ? At what point may I need the “T2wSPC” files. Can I choose just 1 of my T1 images for co-registration?

Hi @Reubebe,

Your best bet would likely to be making a BIDS compatible dataset and then running a BIDS app such as fMRIprep.

You can use the tool dcm2bids (tutorial) to convert your dicoms directly into a BIDS valid dataset.

No, you most likely do not want to combine these data into a single run. Artifacts such as low frequency drift and scanner heat effects over the scanning session will cause jumps between data if you were to simply concatenate. In your BIDS data set, each scan should have a run-<X> label in the file name to designate which run number it is.

Assuming these are fieldmap files, they would go in the BIDS fmap folder, using the naming specified in the BIDS specification. dcm2bids will take care of the naming and organization for you.

T2 images can be helpful for many things, such as refining pial surfaces in brain surface reconstruction, and getting T1/T2 ratio images, which can be used to infer myelination levels.

You could, but you might get better results using all the ones available (that are of good quality).


Thank you for your response