Confounds regressors: which would you recommend to include for GLM with children's task fmri?

Hello experts,

I have collected children(6~12 year olds)‘s task fMRI data and just completed running fmriprep.
I included the 6 motion parameters, FD, aCompCor as nuisance regressors in a GLM with adults’ data.
Would you recommend any other additional parameters for children’s data?

For your info, I ran fmriprep with --ignore fieldmaps --output-spaces MNIPediatricAsym:res-2:cohort-3 T1w fsaverage:den-10k --skull-strip-template MNIPediatricAsym:cohort-3 --fd-spike-threshold 0.5 --dvars-spike-threshold 20.

I would appreciate any comments you could give me!


1 Like

No big suggestion on my side.
If you really want to be on the safe side, and have a lot of data, you could use the so-called “Friston’s 24 motion parameters”, that include time-shifted motion time course and the squared version of motion and time-shifted motion, but I wouldn’t take this as a requirement.

The most important thing is to make sure that your data are not corrupted e.g. by motion, through visual inspection.