Dear FSL usergroup,
I am running TBSS for analysis of DTI data between two groups (n=12 per
group), and am looking, in particular, for differences in gray matter
regions (Heschl's Gyrus and Planum Temporalus). However, FA values in these
regions are below the threshold commonly used in TBSS analysis (<0.2) and
these areas are not mapped on the skeleton. I ran randomize on all my
non-skeletonized data (all_FA) with only a global brain mask and found
statistically significant differences (p<0.001) between my two groups in my
two areas of interest in the un-corrected data. Are these results valid?
i.e. is it reasonable to run this unpaired analysis on non-skeletonized data??
I have searched the forum, but found no discussion of TBSS analysis in GM
regions. Given that the non-linear warping is applied to the whole image, is
it reasonable to assume that GM regions are relatively well aligned ? (these
data sets have been aligned previously to MNI space prior to TBSS)
If the data are valid, what sort of correction should I take into account
for multiple comparisons?
Any input would be greatly appreciated.
Thank you very much.
Adrian Koziak, M.Sc.
Research Assistant
Brain-Body Institute,
Imaging Research Centre
St. Joseph's Healthcare
Hamilton, Ontario, CANADA.
|