Confirming the correctness of removal of confounds

Hi everyone!

I have written a script to regress out (using GLM) the confounds such as motion parameters, mean White matter signal etc from each voxel of fMRI scan.

Is there a way to confirm if the results are correct?


Hi @varunk,

I don’t have the answer to your question, but have you already looked into nilearn's signal.clean function? Not sure if this is comparable to a GLM approach, but by using the confounds parameter (whilst carefully setting the other parameters) might give you comparable results.

Alternatively, if you’re ok with it, you could also share your code. I personally would be interested in this and it would also allow others to give you a feedback on your code.