There should be an ocn.dat file that corresponds to the CSD base name
that you passed to mri_glmfit-sim. The cache just uses a pre-computed
table. The computations are pretty simple. It synthesizes white gaussian
noise on the surface, smooths it by a certain amount, then finds
clusters at vario
Hi Zhang, there should be a file named something like
"cache.th13.abs.y.ocn.dat" that is output as part of the correction for
multiple comparisons. This will have a column for each significant
cluster in the order that the clusters are in. Each row is a subject. So
you can take the mean over th
