Erik,

I prefer a "highlight but not hide" approach to reporting results because the cluster-based adjustment tends to waste a lot of information:

[afni.nimh.nih.gov]

If you want to adopt the cluster approach, use 3dClustSim.

Gang