Daniel Glen Wrote:
-------------------------------------------------------
> Typically, a single table should suffice across
> all sub-bricks because they are all based on the
> same volume, error time series and smoothness
> estimates. Do you have another scenario?


Daniel Glen:

Thanks for your reply.
First, I computed the FWHMs for all sub-bricks
in the input dataset, each one separately by 3dFWHMx. So when I want to get the cluster table for each sub-brick, different FWHMs were used. As a result, each sub-brick got different cluster tables.
I will show you the commands so you will know the processing more precisely.



foreach type("Model 0" "F_Task 1" "F_Lang 2" "F_TaskByLang 3" "T_Task 4" "T_Lang 6" "TaskByLang 8" "Lis_Chn-Eng 10" "Rea_Chn-Eng 12" "Chn_Lis-Rea 14" "Eng_Lis-Rea 16" "Listen 18" "Reading 20" "Chinese 22" "English 24" "Lis_Chn 26" "Lis_Eng 28" "Rea_Chn 30" "Rea_Eng 32")
set tt=(`echo ${type} | awk '{print $1}'`)
set n=(`echo ${type} | awk '{print $2}'`)

set fn = ANCOVA.nii\["$n"]

set fx = (`3dFWHMx -dset "$fn" | awk '{print $1}' `)
set fy = (`3dFWHMx -dset "$fn" | awk '{print $2}' `)
set fz = (`3dFWHMx -dset "$fn" | awk '{print $3}' `)
echo "#================================================="
echo "#=== $tt "
echo "#================================================="
# check cluster size with uncrrected P thresh, -fwhmx,y,z from 3dFWHM:
3dClustSim -quiet -mask "$fn" -fwhmxyz $fx $fy $fz -iter 2000 -nodec -niml -prefix
echo " "
echo " "
echo " "
end


Any suggestion?

Mengxing