Hi Steve et al,
where in the BET code are the skull coordinates processed (line numbers in
the code might be handy)? If I could hack that section a little, it could
output XYZ locations for the skull, probably in the same coordinate
framework as the "brain" surface vertices.
I assume the following. BET finds vertices on the "brain" surface. For
each of these vertices it calculates a surface normal or the vector from the
MRI volume origin through the vertex point. It then traverses along this
normal or "radial" direction until it hits an image intensity that
corresponds to skull tissue. This point is located in XYZ and tagged
somehow.
If this assumption is correct, then there are as many of the skull XYZ
vertices as there are brain vertices. A close approximation to their
tesselation is the .dat file output for the brain vertices (although
probably not very good).
An alternative might be to load the skull .img into matlab and find all
voxels equal to 100. If the volume is loaded into an xyz matrix, the
indices of these voxels could be used as vertex points and tesselated with a
delaunay or convhull function. However, it would be useful to know exactly
what the coordinate system is for the brain vertices.
Also, does BET and FAST work better with 16 bit (ie 12 bit) images than with
uint8 images?
Take care, Darren
BTW, The download /doc/index.html has a broken link for BET (ie,
http://bet/... rather than just the relative /bet...). It would be handy to
have the research html pages in the download too.
----- Original Message -----
From: "Stephen Smith" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Thursday, February 28, 2002 8:23 AM
Subject: Re: [FSL] tesselations of head tissue types
> Hi.
>
> On Thu, 28 Feb 2002, Darren Weber wrote:
>
> > can any FSL tools provide tesselated surfaces of any of these: white
> > matter, grey matter, CSF, skull (inner & outer) and scalp?
>
> BET can output the tesselation of the brain surface (ie only a fairly
> smooth, non-detailed, surface) with the -x option. eg
>
> bet steve_struc_334 steve_struc_334_brain -x
>
> then the steve_struc_334.coo file contains the co-ordinates of the
> vertices, and the .dat file contains the details of the links between
> vertices (each line corresponds to each vertex in the .coo file, and the
> 1's correspond to which other vertices they are connected to)
>
> > I note with interest that BET can find not only the cortical or dura
> > surface, but also the outer skull surface. Can you output the vertex
> > points of this routine in ascii?
>
> not for the skull - BET doesn't offer the option of doing the same for the
> skull (it doesn't even model the skull as a surface, just as a sparse set
> of voxel coordinates).
>
> note - for much better surface tesselation stuff, look at FreeSurfer
> http://surfer.nmr.mgh.harvard.edu/ which is explicitly for this kind of
> thing.
>
> Thanks, Steve.
>
> Stephen M. Smith
> Head of Image Analysis, FMRIB
>
> Oxford University Centre for Functional MRI of the Brain
> John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
> +44 (0) 1865 222726 (fax 222717)
>
> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve
|