Hi,
The volume correction is not done because the fundamental measure is TBSS is not related to volume. It is a microstructural measure (e.g. Fractional Anisotropy) which is averaged over a voxel volume, but this average value does not need to be increased or decreased based on the warping. For example, consider the case where FA was constant everywhere in one brain. If you warp this to another brain, then the right value to use everywhere in the warped brain is still the constant FA value, as you know that this is what was present in the microstructure at all points in the original brain. This argument obviously then extends to the non-constant case, and hence this is why no volume-based modulation is done for FA.
The modulation is needed in VBM because there you are working with a measure that is fundamentally linked to the volume (amount of GM in a region divided by the volume of the region).
I hope this helps clarify it for you.
All the best,
Mark
On 29 Aug 2012, at 09:56, SUBSCRIBE FSL Jay <[log in to unmask]> wrote:
> Thank you very much for your reply, Jesper.
>
> But why is volume correction NOT performed in TBSS? it is quite obvious that different subjects in the groups have different brain sizes and hence different concentration of white matter. Shouldn't these different sizes be volume corrected before continuing with skeletonisation and statistical analyses? Is there a particular reason as to why volume correction is done in fslvbm and not tbss?
>
> BW
> Jay
>
|