Alex,
The tutorial is treating the orig surface (gray-white boundary) as the
fiducial surface. The smoothing is just taking off the rough voxelly
look -- not projecting out along the normal. Although you can shrink or
expand a surface, Caret needs to know how much, and this varies along
the cortex, so the easiest way around this is to average the orig and
pial surfaces to get something that approximates midway between the
gray-white boundary and the pial surface. This is what we mean by a
fiducial surface, but evidently this step was not included in this very
old tutorial. (There may be cases where you really do want to stick
with the orig as the fiducial -- i.e., if you're comparing registration
results, but want to keep the starting surface equal for both methods.)
If you don't have a good reason not to, I'd essentially replace this
step preceding the step where you save the fiducial coord file:
Select Surface: Load AUXconfig Into REFconfig
... with these steps:
File: Import File: Freesurfer ASCII surface file: ../surf/rh.pial.asc
select right hemisphere
select coords, but deselect topology
File: Save Data File: Coord file: Macaque.falco.R.PIAL.12389.coord
File Type: FIDUCIAL; Coordinate Frame: Native; Orientation Left,
Posterior, Inferior
Surface: Create Average Coordinate File
Add: Macaque.falco.R.PIAL.12389.coord and
Macaque.falco.R.RAW.12389.coord
Output coordinate file: Macaque.falco.R.FIDUCIAL.12389.coord
After pressing Apply, select File: Open Data File: Coord file:
Macaque.falco.R.FIDUCIAL.12389.coord and specify that this is a fiducial
coord file.
File: Save Data File: Coord file: Macaque.falco.R.FIDUCIAL.12389.coord
and specify in the comments that this is the average of the rh.orig and
rh.pial surfaces.
Still do the steps to calculate the curvature and save the surface shape
file.
Rather than import the inflated surface,do Surface: Geometry: Generate
Inflated and Ellipsoid from Fiducial. Accept the default selections.
Then do Surface: Flatten Full or Partial Hemisphere
Type: Full Hemisphere (Ellipsoid) and Morph Sphere
At this point, page 30 and on of this tutorial should work:
http://brainmap.wustl.edu/caret/pdf/Caret_5_12_Tutorial_Segment.pdf
Even if you don't care about flat maps, it's the easiest way to draw the
registration borders, so we think of it as a means to registration (even
though registration is carried out on the sphere, using the closed
topology).
After flattening, this tutorial covers registration:
http://brainmap.wustl.edu/caret/pdf/Caret_5_12_Tutorial_Intro.pdf
But it hasn't yet been updated to reflect our PALS atlas target and
core6 landmark set. This page supersedes the registration tutorial,
where they conflict:
http://brainvis.wustl.edu/help/landmarks_core6/landmarks_core6.html
Sorry everything is so fractured.
Donna
On 12/16/2005 01:10 PM, Fornito, Alexander wrote:
Hi Donna,
Thanks for your quick response.
I gatehred from the tutorial that Caret generates a fiducial surface from the
freesurfer surfaces, which is what I'd like to do. Do I have to import the orig
and pial surfaces for this. What do I do after this?
Thanks again,
Alex
Alex Fornito
M.Psych/PhD (clin. neuro.) candidate
Melbourne Neuropsychiatry Centre and Department of Psychology
The University of Melbourne
[EMAIL PROTECTED]
________________________________
From: [EMAIL PROTECTED] on behalf of Donna Hanlon
Sent: Sat 17/12/2005 12:56 AM
To: Caret, SureFit, and SuMS software users
Subject: Re: [caret-users] Tutorial troubles
Hi Alex,
I recognize your name from other mailing lists, but don't recall you
posting on caret-users before -- welcome!
Yes, this post reminds me of our need to update the old Caret 4.6
Tutorial 11 (Freesurfer-to-Caret), because this is something several
users are going to want to do (e.g., so you can register your
Freesurfer-generated surfaces using Caret's landmark-based registration
algorithm, which is useful in ways that differ from Freesurfer's
intensity/convexity-based registration algorithm). At some point, both
tools may support multiple registration algorithms, but I'm guessing
this is a ways off.
Reading this passage, it's amazing how little I recall of the way Caret
4.6 used to work. ;-) But I don't think this step is necessary in
Caret5. The smoothing will operate on the surface in the main window,
whatever surface that is (presumably the rh.orig.asc surface).
One thing that bears attention from David, my boss, however, is whether
it is generally good practice to import both the rh.orig.asc and
rh.pial.asc surfaces, and then use the Surface: Create Average Coord
File, the result of which should be more like our SureFit-generated
fiducial surfaces. I suppose this depends on what your objectives are,
but David feels it is important to aim for layer 4, because then the
surface area along gyral crests is not underestimated and the area along
sulcal fundi is not overestimated. Since many of our analysis
techniques depend on surface area, this is an important consideration
for us.
Please bear with our current state of documentation. We're juggling
many priorities right now, but at some point we will turn our attention
to updating the tutorials. Users who update them for us get stickers.
;-) Seriously, if you run into more trouble, let us know, because other
users will soon feel your pain.
Donna
On 12/16/2005 01:55 AM, Fornito, Alexander wrote:
Hi,
I'm new to Caret and have just downloaded version 5.3.
I'm trying to work with surfaces generated in freesurfer, and have been
following the tutotial, but have become stuck at a point where I think the new
version differs to older ones.
After improting the surfaces and smoothing it, the tutorial (p. 71) says Select
Surface: Load AUXconfig into REFconfig.
The config files (?) did not come up before, so i don't know what this refers
to. Also, I can't seem to find anything that might resemble this option in the
usrface (or any other) pull-down menu.
Any help on how to proceed is much appreciated. My goal is to used caret
registration with freesurfer surfaces.
Thanks,
Alex
Alex Fornito
M.Psych/PhD (clin. neuro.) candidate
Melbourne Neuropsychiatry Centre and Department of Psychology
The University of Melbourne
[EMAIL PROTECTED]
_______________________________________________
caret-users mailing list
caret-users@brainvis.wustl.edu
http://pulvinar.wustl.edu/mailman/listinfo/caret-users
_______________________________________________
caret-users mailing list
caret-users@brainvis.wustl.edu
http://pulvinar.wustl.edu/mailman/listinfo/caret-users
_______________________________________________
caret-users mailing list
caret-users@brainvis.wustl.edu
http://pulvinar.wustl.edu/mailman/listinfo/caret-users
