Dear fmriprep experts,
I collected two different T1-weighted images for my fMRI study (one raw image “T1_005” and the other corrected for coil sensitivity and normalized using a fast pre scan "T1_006). I was wondering whether fmriprep pipeline already includes automatic normalization of anatomic sequences or not. if it does, then it would make sense to use the raw image. However, if it does not I will use the corrected image.
Can anyone help me with this?
Thank you
Hi,
You can read more about the anatomical processing here: GitHub - nipreps/smriprep: Structural MRI PREProcessing (sMRIPrep) workflows for NIPreps (NeuroImaging PREProcessing tools).
I am not sure what kind of normalization you are referring to, but there is bias correction for intensity non-uniformity. There is also spatial normalization, but that doesn’t sound like what you are referring to.
Best,
Steven
Dear Steven,
Thank you for your reply.
Yes, I am referring to the anatomically irrelevant intensity variation throughout data that can be induced, for example, by the choice of the radio-frequency coil. Our fMRI scanner also provides an intensity non-uniformity correction for each participant’s T1. However, if fmriprep does it already in the pipeline, it is better to use the raw T1-weighted image so that the image is not corrected twice.
Loris