Hi,
I think that it will be difficult to get betsurf to work on this data, and you are probably better off going for some alternative approach. Segmenting the skull, by taking your brain mask, dilating this and then classifying the tissue with FAST, should work reasonably well. From this you could then fit surfaces with marching cubes in python/matlab to get surfaces if you needed them. However, I do not work with this kind of data so it would be interesting to hear opinions from other people that do work with NHP data.
All the best,Mark
From: wolf zinke <[log in to unmask]>
Reply-To: FSL - FMRIB's Software Library <[log in to unmask]>
Date: Wednesday, 7 January 2015 23:08
To: "[log in to unmask]" <[log in to unmask]>
Subject: Re: [FSL] betsurf with NHP
Hi Bryson,
Thanks for the reply.
Getting a brain mask is not a problem. The tricky part with betsurf is the transformation into standard space :
http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/BET/UserGuide#betsurf
This transformation seems to be mandatory. I was wondering if it could be omitted (and hoping for the best that the results are getting good enough without these constraints), or if there are NHP data files available that could be used instead.Finally, it requires that you supply a transformation from the T1 to standard space (e.g., generated by flirt) as it uses standard-space masks in order to constraint some of the mesh fitting.
wolf
On 1/7/15 1:40 PM, Bryson Dietz wrote:
Hello,
As with bet (I believe) you can just "fool" the algorithm. Use: fsledithd vim (or whatever text editor you prefer) and modify the dx dy dz to be closer to human size. I do this often for rat studies (just using bet though).
Bryson
On Wed, Jan 7, 2015 at 1:24 PM, wolf zinke <[log in to unmask]> wrote:
Hi,
If I understand it correctly, betsurf is tailored to use (only) with humans because it utilizes data/assumptions based on the standard brain. Is it possible to adapt it in a way to get betsurf working with NHPs? I need to get the representation of the outter and inner skull as shown in red in the image. My current segementation I got with some nasty tricks and tweaks that I would like to replace with a 'better' method.
Has anyone done this with betsurf on NHPs or do you have other recommendations for tools that might be suitable for this task?
thanks,
wolf