[Freesurfer] Detrending timeseries with yhat.mgh from mri_glmfit

Hello,

I have a question regarding the yhat.mgh file generated from some resting-state functional data after running: mri_glmfit --y fmcpr.siemens.sm0.self.?h.nii.gz --glmdir $dest_dirname --qa --save-yhat

I had assumed that the yhat file was the detrended data, but after plotting it in matlab and seeing perfectly smooth lines, I came to believe that the yhat.mgh file represents the trendline for each timeseries in the source .nii.gz? In this case, if I use mri_segstats to pull out the timeseries data from an ROI, is it then valid to calculate roi_timeseries/roi_trendline in order to arrive at a detrended timeseries for that ROI?

Thanks, Chris