Hi Jeff - read the thread below starting with my Sept 8 2008 email.
Bruce - what is the status on the insula label for aparc?
carl
Hello all,
Thank you for your quick replies to my previous posts. One more
question. I am wondering what the difference is and if I should be taking average thickness measurements for cortical regions from lh.aparc.a2005s.stats or lh.aparc.stats. I think the main difference is that lh.aparc.a2005s.stats was made with ?h.atlas2005_simple.gcs and lh.aparc.stats was made with ?h.curvature.buckner40.filled.desikan_killiany.2007-06-20gcs. Which one is more accurate? I am using freesurfer 3.0.5.
Thank you, Jeff Sadino
Bruce Fischl fischl at nmr.mgh.harvard.edu Tue Sep 9 10:03:31 EDT 2008 Previous message: [Freesurfer] annotation file and labels for fsaverage Next message: [Freesurfer] dicom image viewer Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] great, thanks Ron.
Bruce On Tue, 9 Sep 2008, Ron Killiany wrote:
I believe Sarah will be over at MGH on friday to begin finalizing this.
--ron
On Mon, Sep 8, 2008 at 9:31 PM, Bruce Fischl <fischl at
nmr.mgh.harvard.edu>wrote:
Hi Dara,
there is an insula label in the works (*lots* of people have asked this question). Ron: any updates on the insula label?
thanks Bruce
On Mon, 8 Sep 2008, Dara Manoach wrote:
Thanks very much for the detailed info Karl (and Doug and Bruce)!
My primary ROI is ACC, so we'll probably use Aparc with the cingulate divisions. I'll have to review the paper to see if the anatomic bases of the divisions are described. I'm surprised, though, that there is not an insula label, and I'm wondering why it would not be possible to use ROI labels from the two annotation schemes to examine activation in group
data
(i.e., ACC labels from aparc and insula from aparc2005).
best, dara
On Sep 8, 2008, at 8:18 PM, Bruce Fischl wrote:
for what it's worth we label the cc now explicitly in the aseg, which is
certainly more accurate than the aparc used to be.
Thanks for the detailed answer though Carl!
Bruce On Mon, 8 Sep 2008 carl at nmr.mgh.harvard.edu wrote:
Hi Dara,
You can certainly use either one as Doug says, but depending on your purposes, one or the other may have some virtues. 2005 consists of discrete sulcal and gyral labels with traditional anatomical names so there are a lot more labels (which is why Doug describes it as very detailed), whereas aparc in general collapses these into summary regions.
But it is not a a simple case of lumper vs splitter....For instance it seems that
Aparc divides the cingulate into rostral anterior, caudal anterior, posterior cingulate whereas in aparc 2005 it is lumped into one cingulate label
Aparc shows the corpus c. nicely (I think there may even be a subparcellation available in development) whereas the cc is lost in the medial wall in 2005.
Aparc differentiates the pars orbitalis nicely on the orbital surface from the rest of the ofc, aparc 2005 does not ? possibly of interest to someone who is looking at ventrolateral pfc and its neighbors
Aparc includes in its lateral ofc label everything lateral to the
rectus
gyrus, including the central ofc and the (unlabeled) transverse sulcus neither of which would be understood as being in the lateral ofc as
that
term is typically used today.
Aparc 2005 shows a view that nicely differentiates the rectus gyrus on sagital section from the adjoining, whereas aparc lumps a broad territory into ?medial OFC?
Aparc 2005 includes parcellation of the Insula, aparc does not and
it is
lost in the unknown medial wall (there has been talk of fixing this but is this real, Bruce or Doug?)
Perhaps what we need is a ?best of aparc and aparc 2005?.
For surfing purposes you can also ?fool? tksurfer into showing you the names of the (typically more more detailed) 2005 parcellations while viewing the brain labeled with the aparc labels which at times have
more
detail ? though I have never asked Doug if this will always work correctly.
If others have discovered pros and cons in other regions, please share.
Best,
Carl