iag iag-
you may want to try increasing the preflooding height and use the -more option. i've found the seed points don't work very well, if at all. if all else fails, you may have to manually fix it in tkmedit with the clone tool. open tkmedit with the brain and T1 volumes as main and aux, respectively. select edit voxels tool, then Tools->Configure Volume Brush. in the dialog box, select 'Clone' under the Mode section, and 'Aux Volume' under the Clone Source section. now when you make edits to the main volume, it will fill in with the values from the T1.
good luck, brian t. quinn
On Wed, 2 Mar 2005, iag iag wrote:
Hello all.
I am having a problem with skull stripping that I can't figure out a way around. One subject out of 10 is having the problem. The intensity normalization works fine and all values hover around 110 just like in the other 9 subjects. However, during skull stripping the pons is very neatly cut out in this one subject. Of course this causes problems later when WMFill can't find the pons.
I tried to change the flood height, but that didn't change anything (actually it changed NOTHING which kind of suprised me). I tried adding a seed point, but that didn't change the segmentation either. I even tried to add a seed point in Volume Index and Volume RAS coords just in case I was doing something wrong with the coord system expected by the routine. I also tried not using the watershed. But no matter what I try the pons is always neatly chopped out so I can't get past WMFill on this subject.
Does anyone have any ideas on what I might be able to do to allow the pons to be left during RmSkull?
Thanks for any ideas.