Hello!

I am trying to get SVC values for a particularly small mask (left amygdala mask with 48 voxels). I tried to follow the advice I found in other threads and ran the following command:

3dClustSim -OKsmallmask -mask Amygdala_Left_Mask_resam+tlrc -fwhmxyz 6.2547 6.2360 5.5507 -pthr 0.02 0.01 0.005 0.001 -athr 0.05 -prefix CSAmyg


But I got the following error:

*+ WARNING: Simulation not effective for these cases:

1-sided NN=1 pthr= 0.001000 alpha= 0.050 [max simulated alpha= 0.022]
1-sided NN=2 pthr= 0.001000 alpha= 0.050 [max simulated alpha= 0.022]
1-sided NN=3 pthr= 0.001000 alpha= 0.050 [max simulated alpha= 0.022]

*+ This means that not enough clusters, of any size, +*
of voxels at or below each pthr threshold, were +*
found to estimate at each alpha level. +*
*+ In other words, the probability that noise-only +*
data (of the given smoothness) will cause +*
above-threshold (at the given pthr) clusters is +*
smaller than the desired alpha levels. +*
*+ This problem can arise when the masked region +*
being simulated is small and at the same time +*
the smoothness (FWHM) is large. +*
*+ Read the 'CAUTION and CAVEAT' section at the end +*
of the '-help' output for a longer explanation. +*

I have tried running it with a combined left and right amygdala mask as well just to see if that would be enough voxels and it also failed. Is it just extremely unadvisable to run 3dclustsim with an amygdala mask? I have strong a priori interest in the amygdala and want to follow up on results from a 3x6 ANCOVA with 2 covariates and can't run that model with an amygdala ROI using the amygdala as a mask. Whole brain I get up to 27 voxels of amygdala surviving at p uncorrected .005 for one of my general linear t tests however since it is whole brain the q in afni does not come close to .05. I used 3dclustsim with my group mask for cortical regions but given the size of the amygdala - none of my clusters within the amygdala survive with that threshold. Any advice on how to get 3dclustsim to work for the amygdala or any other advice on SVC using afni for the amygdala?

Thank you!

-Elise