The report
Longer scans boost prediction and cut costs in brain-wide association studies used HCP resting state data, with preprocessing involving censoring frames based on FD and DVARS. From digging around
CBIG_preproc_fslmcflirt_outliers.sh, I'm guessing that the FD measure is what
fsl_motion_outliers calls FDRMS (and
not FD), and that reported DVARS thresholds apply to the scaling conventions from that FSL tool. Are those timeseries for DVARS and FD available in the CBIG repo? Or, are the particular frames that were censored available? If they're not already available, is it correct that they were generated from the "unprocessed" data (e.g., mcflirt to the sbref, and then fsl_motion_outliers)?