Hi Gabriel,

Just looking ahead, there's a good chance reviewers will complain about using 
an individual as an atlas target, if you are registering other individuals to 
this template.  But I don't want to presume to know what/how/why you are doing 
this.  I just felt better saying this before you did a lot of work, rather than 
after. ;-)

Note that the flattening tutorial might be helpful, too, as it is somewhat of a 
means to the spherical registration end, using the older methods.  In 
particular, drawing the medial wall is often easier on the compressed medial 
wall view you get when you flatten.  You don't have to use the flat map, but it 
can be handy as a border drawing substrate.

Note also that back when we used the methods described in these tutorials, we 
used only the "Core6 landmarks" for humans, because individual variability made 
other landmarks less reliable.  When you do an atlas-to-atlas registration 
(which is typically population to population, in our studies), you can often 
add more landmarks where there is clear correspondence between source and 
target (e.g., a mean baby midthickness surface and a mean adult midthickness 
surface).  You typically draw on something less folded (sphere, flat, 
ellipsoid), but project the borders and view on the anatomical/midthickness 
surface, to ensure accuracy.  You might have two caret sessions going -- one 
for source, one for target -- with a second window open in each.  The main 
window has the less folded border drawing substrate, while the midthickness is 
in window 2.

When in doubt about the correspondence between source and target borders, 
delete/truncate the border.

I think perhaps the LVD registration algorithm is preferred in these cases, but 
if your source is an individual rather than a population, I'm less sure.

Donna


On May 21, 2014, at 2:40 AM, Gabriel Gonzalez Escamilla <[email protected]> wrote:

> Dear Donna, 
> 
> Thanks so much for your answer,
> 
> To make my own surface-atlas from my template subject, I did followed the 
> steps for creation of individual surfaces in the same tutorial that you sent 
> me, but as suggested for saving time I didn't created the sphere, and I don't 
> ave any thickness info, so I might need to generate it!. And yes, I want the 
> Brodmann parcellation in the form of surface-based paint/label or border 
> form, just as shown in the Caret tutorials. So, I think I will go with the 
> surface-based registration to bring the paint/borders David van Essen 
> generated to my atlas. 
> 
> I think the part of this tutorial to perform such a thing is the one titled 
> as: Spherical Registration, right? If so, I think is just matter of follow 
> carefully all the steps.
> 
> Thanks so much,
> 
> Gabriel
> 
> 
> -- 
> --------------------------
> PhD. student Gabriel González-Escamilla
> Laboratory of Functional Neuroscience
> Department of Physiology, Anatomy, and Cell Biology
> University Pablo de Olavide
> Ctra. de Utrera, Km.1
> 41013 - Seville
> - Spain -
> 
> Email: [email protected]
> http://www.upo.es/neuroaging/es/ 
> _______________________________________________
> caret-users mailing list
> [email protected]
> http://brainvis.wustl.edu/mailman/listinfo/caret-users


_______________________________________________
caret-users mailing list
[email protected]
http://brainvis.wustl.edu/mailman/listinfo/caret-users

Reply via email to