Hello AFNI users,

I've run 3dclust on a dataset to detect clusters in the difference sub-brick between two groups.

3dclust -1thresh 2.581 -1dindex 4 -1tindex 5 6 422 dataset+orig

The cluster results look fine and correspond to the images as expected.

Next I decided to reduce the z-threshold to 1.965 (p>0.05) and increase the interclusterconnectivity distance (rmm) from 6 to 10 millimeters (voxel dimensions =3.75x3.75x6 ). This is so that rmm > 6*sqrt(2). Here is the command line -

3dclust -1thresh 1.965 -1dindex 4 -1tindex 5 10 422 dataset+orig

The first row of the output of 3dclust now shows a huge (compared to the rest) cluster (247388 microliters). Locating this on the images according to the coordinates listed takes me to an empty area without any activation.
This makes me wonder if there might be something fishy going on. The remaining cluster locations look fine on the images.

Is there something I'd be happy to upload a dataset if you'd like to take a look.
Your help is appreciated.

Thank you.
Mithun