Hello Stefan & list,
I followed the discussion you had on the analysis thresholding for VBM data.
I am not intending to produce more confusion but would like to take the question
on thresholding one step further:
UN-modulated maps are probability maps for which absolute thresholding is logical.
In modulated maps however - when the probabilty has been translated into an absolute volume -
a voxel value which is identical between two subjectes before modulation may be different after modulation.
Would you recommend to use an absolute or a proportional analysis threshold for modulated data?
Thanks for any hints on this,
Philipp
|