[Freesurfer] hippocampus segmentations

Hi, I had a question regarding the hippocampus labels in freesurfer (stable 3.0.2). In the course of trying shape analysis (Martin Styner's SPHARM tools) on them, I've noticed the tail regions tend to be "fatter" than would normally be expected. For example, the attached pictures show an average image for our sample, compared to the sample average included in the package, which I think was based on manual ROI's.

I was wondering if this was an expected attribute of how freesurfer classifies the region, and also if there was anything I could do to address this. I can also provide datasets if it'd be helpful.

Thanks,

-Aaron-