peak outside of cluster?

[okepinska]

Hello, thanks so much for the great toolbox!
I'm using it to label FSL randomize results of 1-p tfce images.
Weirdly enough the peaks of two of the clusters are listed as having a value of 0 (so the location of the peak is outside of the cluster).

Here's the command I used:
atlasreader dr_stage3_alice_tfce_corrp_tstat1.nii.gz -t 0.95 5 -a 'all'

and the results look like this:

Any idea of what might be going wrong here?

Many thanks,
Olga