Plot BOLD standard deviation before and after preprocessing

[oesteban]

I've just realized that nipreps/mriqc#625 is a bit out of the scope of MRIQC, but it definitely fits well within fMRIPrep's reports.

We would gauge first whether these s.d. maps can replace the bold reference on the BOLD ROIs reportlet. If so, we would not need one additional reportlet.