I was wondering if it is possible to get both minimally preprocessed and ICA-AROMA denoised data output in CIFTI space. I ask because I would like to compare various post-preprocessing strategies.
I’m not sure if it is worth implementing for everyone (would probably generate unneeded files for most cases) but perhaps there is some python code I could call with the AROMA-denoised nii.gz as input?
Alternatively, perhaps a tool like ciftity is the best way to go (although it seems quite involved and seems to register the spaces differently to fmriprep).
I should add - this only applies to non-aggressive AROMA denoising. For aggressive, I could just regress the generated noise confounds from the CIFTI timeseries, correct?
I was wondering about this as well, and it doesn’t look like the question was ever answered. From the workflow outline here (https://fmriprep.readthedocs.io/en/stable/workflows.html#epi-sampled-to-freesurfer-surfaces), it appears that the motion-corrected data is projected to the freesurfer surface, but the fully preprocessed data is only saved in volumetric space. I assume the same is true for the CIFTI output. Is this correct?
That is not possible at this point. But it would be fairly easy if we undertake the big refactor of ICA-AROMA @jdkent has been promising
Right now, ciftify is you only easy route forward.
Correct
The data sampled on surfaces have STC (slice-timing correction, optional), HMC (head-motion correction) and SDC (susceptibility-derived distortion correction, optional).
