Hi Mellanie,
Actually, there are seven handles after you press FloodFill Vol2. But I
could spot only two looking at the inflated surface.
One was in the familiar trouble spot, the parahippocampal gyrus (medial
surface view), while the other was near the temporal pole (ventral
surface view).
This text file outlines some tips for patching those handles:
http://brainmap.wustl.edu/pub/donna/ROBARTS/dec31_2002.R.full.txt
login pub
password download
These edits reduce the handle count by three, leaving four handles, but
I couldn't see them on the surface. When this happens, sometimes I
re-run the segmentation through automatic error correction:
- use segmentation in vol2
- fill ventricles no
- correct errors yes
- identify sulci yes
At the end of this process, I check the Locate Objects output; error
correction generates an updated *errors*mnc file, which should be the
default selection when the resulting *segment_vent_corr*_corr.mnc is
loaded as vol2. If Locate Objects reports no objects, then there were
no crossover areas sufficiently large to be flagged as an error object.
Another result of re-running error correction is an updated
SURFACES/*segment_vent_corr*_corr.errors*RGB_paint file. When you paint
this file on your updated surface, and you can't find any red-painted
areas (you won't, if Locate Objects reports no error objects), then you
can safely proceed to caret.
But in this situation, once I launch caret, I switch to the ellipsoid
surface and do Surface:Measurements:Crossover Check, other type surface.
In this case, I got 2 node crossovers and 49 tile crossovers. This
tells me to play with some of the topology fixing features. I did this:
Surface: Topology:
Check for islands
Disconnect islands
Remove corners and stragglers
Orient tiles correctly
Then I saved the topo file as
dec31_2002.R.full.segment_vent_corr.patch8_clean.65582.topo (insert
clean somewhere to differentiate it from the original topology).
Surface: Measurements: Crossover check now reports 2 node crossovers, 1
tile crossover -- a worthwhile improvement.
Then on to flattening....
Donna
On 08/19/2004 07:19 PM, [EMAIL PROTECTED] wrote:
Hi Donna,
There are still 6 handles remaining on my surface. I have used the objects
list and the surface, but still have not been able to reduce the error count.
Could you please guide me to the trouble spots? Any help would be greatly
appreciated.
I have uploaded the volume. (dec31_2002.R.full)
Thanks,
Mellanie
_______________________________________________
caret-users mailing list
caret-users@brainvis.wustl.edu
http://pulvinar.wustl.edu/mailman/listinfo/caret-users
