I have a question about TBSS sensitivity. I ran TBSS on FA data in a group of 50 healthy subjects accordingly to http://fsl.fmrib.ox.ac.uk/fsl/tbss/
The quality of my DTI data really seems good (b=1000/ 41 directions/ 2mm isotropic)
For statistics, I created a design.mat & design.con with age (demeaned) as the variable of interest, and a -1 correlation with age for the contrast design (and a second contrast with a +1 correlation)
I then ran:
randomise -i all_FA_skeletonised -o tbss -m mean_FA_skeleton_mask -d design.mat -t design.con -n 500 --T2 -V -D
and to view the results:
tbss_fill tbss_tfce_corrp_tstat1 0.95 mean_FA tbss_fill
fslview mean_FA -b 0,0.6 mean_FA_skeleton -l Green -b 0.2,0.7 tbss_fill -l Red-Yellow
idem with the opposite contrast (positive correlation with age)
Nevertheless, this analysis yielded no significant results. I ran the same analysis in an another independant sample (N=45), without any more significant results. It seems quite odd to me, as the relationship of FA with age has been described several times.
Furthermore, people have found significant TFCE differences in FA using TBSS in small groups where the decrease of FA should be quite small and hardly significant, e.g. in 13 unaffected siblings of patients with autism vs 11 controls: http://archpsyc.ama-assn.org/cgi/content/full/67/10/1052
(and many several other similar papers)
Do you have any idea why I can't manage to get a significant negative correlation of FA with age ?