Hi Ben

if the unlababeled vertices are in non-cortical regions, e.g. on the midline, then this is normal. We also only enforce connectivity by removing small regions. If there are multiple large parts to a parcel, we don't relabel it as this can occur using some anatomical definitions.

We can't tell if your case is right without seeing it though

cheers Bruce

On Mon, 10 Oct 2011, Ben Hamida Cammoun Leila Faiza wrote:

Hello,

1)      I reiterate my question, because i cannot continue processing. Some vertices are not labeled in the rh.aparc.annot in bert as 35643,36746,37788.....

2)      The parcels of my subject are not connex.

Below is the detail on my first email. Thank you very much for answering me

De : Ben Hamida Cammoun Leila Faiza Envoyé : jeudi 6 octobre 2011 14:53 À : Objet : nonconnexity and holes in parcels

Hello, I am using freesurfer v5.1.0 (on ubuntu). when processing reconall on my data I notice on the parcellation aparc.annot two strange points: some isolated small regions not annotated, as well as the non connexity of some parcels. I checked the bert aparc.annot on inflated surface and I found the same problem of non connexity, for instance some medialorbitofrontal labeled vertices (vertex index 108562) in rostralanteriorcingulate, or vertex num 103429 labeled as insula in lateralorbitofrontal parcel. In lingual there is also a small hole without annotation like vertices: 35643,36746,37788.....

is this normal?  if yes, how can we correct that? thanks in advance