Dear Qian,
this is a great question. I don't know of a quick way to calculate cluster thresholds for the conjunction map, but it is questionable whether this even makes sense. The idea behind the minimum-statistics approach is rather the other way around: you threshold your 2 maps *before* running the conjunction. Then, the minimum conjunction is voxel-wise: Only voxels that are significant in both individual maps will be significant in the conjunction map; voxels that are significant in only 1 map or neither map are non-significant in the conjunction map. This is why some authors argue for using voxel-wise multiple comparisons correction (e.g. voxel-wise FWE or FDR corr.) when running a conjunction analysis like this (e.g. see https://www.sciencedirect.com/science/article/abs/pii/S1053811914000020).
All the best,
Philipp
|