Hi,
If you do your vertex analysis in MNI space then brain size is already accounted for. If you want to explicitly account for intracranial volume, and have a measurement of this, then that can easily be entered into the GLM as a confound.
If you are doing volume correlations and not a vertex analysis then you will need to divide by a measure of intracranial volume (or brain volume if that is correlated well enough for your population) prior to doing your analyses. As the segmentations themselves are in the native space.
All the best,
Mark
> On 2 Feb 2018, at 15:10, Tamir Eisenstein <[log in to unmask]> wrote:
>
> Hello FSL experts,
>
> I want to examine a correlation between hippocampal volume and some physiological measurements.
> Does the FIRST segmentation generates a normalized segmentation? or do I a need to account intracranial volume as a covariate in the design?
>
> Thank you all,
> Tamir
|