ICA-AROMA after fmriprep

Hello experts, I have run fMRIprep on my data and I have just realized that I miss Aroma. Then some subjects have ica-aroma outputs and some other not (my bad). Since is taking too long for every subject running fmriprep I was thinking on apply ica-aroma with the fmriprep outputs.

For this, (if I’m not wrong) I would need to do:
1- Smooth the data first. But I would like to do it with the same parameters used by ica-aroma in fmriprep (where can I see this?)
2- Run ica-aroma with _desc-preproc_bold.nii.gz and the text file with the realignment parameters (which I can find it in the confounds_regressors.tsv the last 6 columns).

Am I right, or is a better way to do this
Thank you


I love how you stated this.

Are you reusing the work directory? did you keep it?

You could just wrap a command line around fMRIPrep’s AROMA workflow. If you do so, please send us a pull-request because it would be a very cool utility to have.

All the inputs and parameters are documented .

Yes I’m using the same

Sorry maybe I missed something, I don’t fully understand what should I have to add.

Then I assume I’m in the right way.
Thanks for your reply

Then fMRIPrep should reuse all cached results. Meaning, runtime should not be as long.

Sorry that was unclear. I’m suggesting you to reuse the existing code within fmriprep to achieve your goal. By doing so, you’ll be creating a new tool (basically to make AROMA user-friendly with fmriprep outputs) that we would be interested in adding to the distribution of fmriprep, to maximize the usage.

2 posts were split to a new topic: fMRIPrep: how to resample data from T1w space into MNI space