Dear David,
> ---------- Forwarded message ----------
> From: David McGonigle <[log in to unmask]>
> Date: 2009/11/16
>
> it looks like the only mesh that can be used for linear
> inverse solutions is one derived from the Colin Brain. Is this right -
> and if so, will we be able to make our own meshes for future use and
> import them in? I do still rather like looking at un-normalised
> data...!
>
That's not exactly the case. What SPM can do is use the individual
structural image from your subject to compute the deformation field
for normalizing your subject's head (including the brain) to the
template Colin head and then apply the inverse of this deformation to
the canonical meshes derived from the Colin head which effectively
generates individual meshes for your subject. This has the advantage
of being a much simpler and faster process and does not rely on the
structural images being very high quality. In addition you can still
relate every vertex and every triangle in the deformed mesh to the
original mesh with a one-to-one mapping. So when it comes to writing
out the solution SPM uses the canonical mesh rather than the deformed
mesh. So the solutions are computed using individual meshes but are
written out in canonical space which is quite useful for doing group
statistics.
Best,
Vladimir
|