Time Series Data in MNI Space

Thanks to Cary for sorting this out. It appears that the process for obtaining mean time series from the mni305 volumes generated by preproc-sess is very similar to that for time series in surface space

After the data have been detrended, mri_segstats can be called on the detrended volume to extract the average waveform within each of the labeled segments. The next challenge would be to determine the identities of each segment so that you know which waveforms correspond to which subcortical or cortical region.

#Assume ${sub} contains the subject ID
#Assume ${r} contains the run number
#Assume ${FILEPAT} contains the name of the detrended volume: 
FILEPAT=fmcpr.siemens.sm7.mni305.2mm.mgh
mri_segstats --seg /usr/local/freesurfer/5.3.0/subjects/fsaverage/mri.2mm/aseg.mgz --i ${sub}/bold/${r}/${FILEPAT} \
--sum ${sub}/bold/${r}/sum_subcort.txt \
--avgwf ${sub}_${r}.${FILEPAT}.wav.txt

This will generate a *.wav.txt file in the current working directory with t=timepoints rows and s=43 columns (1 per segment).