Hello there!

I was wondering if I could get some help on an issue I am noticing in the segmentation of the left thalamus and left Pallidum on my scans. I’ve included pictures of tkmedit so you can see what I mean. The tkmedit uses this code:  tkmedit SZ2014 brainmask.mgz lh.white -aux norm.mgz -aux-surface rh.white -segmentation aparc+aseg.mgz

All of my scans have been through the entire recon-all process and , out of my sample of about 172 patients and controls, just about 29 of them show this odd segmentation on, almost always, the left hemisphere. The left thalamus seems to be over estimating the size and the pallidum is underestimating its size. I was wondering why this would be and if the community thought it would be okay to go back and manually edit the segementation to fit the subcorticals structures better?

Also, since I have never done such editing, is it advisable or even possible to do so?