add support for scrubbing.

[simexp]

add support for scrubbing.

[smeisler]

I can take a look at this. I know fMRIPrep uses both FD and DVARS to annotate motion outliers, but it seems like most scrubbing approaches only use FD. So I envision scrubbing to involve

1. set a FD threshold (e.g. 0.2 or 0.5 mm based on some literature)
2. set a num outliers threshold (e.g. 4 minutes of remaining data, or 20% data remaining) that defines if a subject would be excluded from the study due to motion
3. toggle between basic scrubbing (remove outliers) and "optimized scubbing" (iterative process that involves basic scubbing and then removing continuous data segments that are too short)
4. Perform the scrubbing
5. Return both the one-hot encoded motion outlier confounds and a binary variable for if subject passed motion outlier threshold (step 2)

[pbellec]

fixed!