Hi there,

I want to use 3dClustSim to FWE correct our group data. I estimated data smoothness in standard space using 3dFWHMx on the residuals, and calculated the cluster stats with 3dClustSim using a standard space mask, see example below. The cluster sizes are huge compared to what people usually report, which is in the range of 20-30 voxels. It appears that they calculated this in subject space (see for instance Rameson et al 2012, or Tremblay et al 2011), which seems incorrect as the stats presented in form of data tables and images are given in MNI space, thus using a cluster threshold of 20 determined from a matrix of 64 x 64 x 64 or something cannot be applied to an standard space imaging matrix of 91 x 109 x 91.

Do I miss something here?

Thanks very much for your advice.

Best,

Torsten


(-fwhmxyz 11.1911 11.0398 11.0225 -NN 3)
# Grid: 91x109x91 2.00x2.00x2.00 mm^3 (180576 voxels in mask)
#
# CLUSTER SIZE THRESHOLD(pthr,alpha) in Voxels
# -NN 3 | alpha = Prob(Cluster >= given size)
# pthr | 0.100 0.050 0.020 0.010
# ------ | ------ ------ ------ ------
0.020000 689.5 807.9 976.0 1103.0
0.010000 408.4 477.7 571.7 646.0
0.005000 261.6 307.2 368.0 409.6
0.002000 156.1 187.6 227.3 262.0
0.001000 107.6 133.3 165.6 192.0
0.000500 74.3 95.5 121.7 140.4
0.000200 44.8 60.2 80.1 96.0
0.000100 29.3 41.6 57.9 69.8