Hello all-
I am trying to compare two neurodegenerative patient populations atrophy patterns using VBM. Sample sizes are 18 and 20.
I have run the VBM pipeline up until Displaying Cluster-Based Thresholding Results
I would like to initially threshold for clusters at P<.05, then correct at P<.01.
Would I just manipulate the default value of -c 2.3 -> to whatever my degrees of freedoms and desired alpha.?
Then modify the following line:
fslmaths fslvbm_clustere_corrp_tstat1 -thr 0.95 -bin mask_pcorrected
to
fslmaths fslvbm_clustere_corrp_tstat1 -thr 0.99 -bin mask_pcorrected ?
Or can I just modify the corrected p-value line, without modifying the initial t-stat line?
Also, is there anything else I should be doing following this?
Thanks
Karim
|