Hi Janosch,

yes, that works if longitudinal changes are really small. You should be careful with copying edits from base to the long as even with small atrophy there can be local deformations of the brain, and rigid alignment and copying may not make sense.

Yet, I believe that even without copying edits, it may be sufficient in this case to check and fix the surfaces in the base, and then (re-)run the long runs. Usually they should be fine even without any edits there.

Best, Martin

On 08/29/2014 01:47 PM, Janosch Linkersdörfer wrote:

Hi Martin and all,

I am about to process a large longitudinal dataset (normal development, no patients) and wanted to ask about the best (i.e., time efficient) strategy for troubleshooting.

If I understand correctly from https://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalEdits, it might be OK in my case (only little longitudinal changes expected) to only edit the base runs and transfer those edits to the long runs. So I thought of doing the following:

• use the QA tools to export detailed snapshots of all base runs
• check the snapshots for errors
• use Freeview to add control points / edit white matter in brainmask.mgz of the base runs (while simultaneously checking whether those edits would make sense in brainmask.mgz of all long runs of the same subject)
• rerun base
• rerun longs

Does that make sense?

Thanks,

Janosch