Nuisance regressors after ICA-Aroma

[shaymaa]

Dear NeuroStars people,
I’m trying to figure out how to use fmriprep non-aggressive ICA-Aroma output.
In Prime at el., 2015 paper they suggest using CSF,WM and linear trend regressors after ICA-Aroma.

My questions:

1. First of all should these regressors be recalculated? As I saw in previous posts it is still an open issue, how can I decide?
2. Second, what regressor in the tsv file is the linear trend regressor?
3. Third, ICA-Aroma non-aggressive bold image’s template is different than the regular preproc-bold image template, can I use the confounds in the tsv file with both templates?

Warm regard,
Shaymaa

[Steven]

@shaymaa

• First of all should these regressors be recalculated? As I saw in previous posts it is still an open issue, how can I decide?

This is still an open issue, and I think eventually the confounds will be recalculated on the AROMA output, but for now the confounds you have should work fine.

@shaymaa

• Second, what regressor in the tsv file is the linear trend regressor?

The cosine_XX regressors are the high pass filters which account for low frequency trends

@shaymaa

Third, ICA-Aroma non-aggressive bold image’s template is different than the regular preproc-bold image template, can I use the confounds in the tsv file with both templates?

Yes, these confounds do not relate to the final resampling space of the image.

Best,
Steven