Dear Amanda
The longitudinally processed time points are already in base space. Those transforms you used map base to the original, cross sectionally processed volumes.
Regarding individual subfields: you can use mri_extract_label to extract individual labels. You can also obtain probabilistic segmentations, but it requires some extra work (and rerunning your subfield segmentation). Let me know if you are interested.
Cheers
Eugenio

From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> on behalf of Worker, Amanda <amanda.worker@kcl.ac.uk>
Sent: Wednesday, March 1, 2017 12:14:50 PM
To: Freesurfer support list
Subject: [Freesurfer] Longitudinal hippocampal subfield registered to base

Hi All,

I have applied the hippocampal subfield segmentation algorithm to each of my time points that have been longitudinally processed. I am now trying to register these back into base space using:

mri_convert -T BASE/mri/transforms/TP_to_base.lta -rl BASE/mri/norm.mgz -rt nearest --out_type mgz TP.long/mri/lh.hippoSfLabels-T1.v10.mgz TP.long/lh.hipp.base.mgz

Firstly, when I try to overlay lh.hipp.base.mgz for each timepoint onto BASE/mri/T1.mgz they do not align. Do you have any idea why this is?

Secondly, the output seems to be for the whole hippocampus, is there a way to do this for each subfield individually?

Any advice would be much appreciated,

Amanda