Hi Anderson

Thank you very much for the explanation. However, I still have several questions that I don't understand.

(1) If the voxels outside the mask are already excluded, why does the result still showed these external voxels are included in the cluster? 
The cluster_mask_zstat*.nii.gz with accumbens mask used definitely showed that cluster 2 with masking contained much more voxels and spreading outside the bilateral nucleus accumbens mask by inspection (in fact, cluster 2 with masking was exactly the same as cluster 9 without masking.)

(2) I looked again in both gfeat folders, the thresh_pstat*.nii.gz with or without masking are exactly the same over every voxels in the brain. I suppose these thresh_pstat are the uncorrected p-value from each voxel's TFCE null-distribution, and therefore I thought the TFCE for each voxel are calculated the same way with or without masking?

(3) The thresh_zstat images did show the effect of masking, as values outside the mask did become zero, but the corresponding voxels within the mask in both the with and without masking options had exactly the same value. Does this also points to that the TFCE voxel value are calculated the same way with or without masking?

(4) The p-value in the cluster report with the masking did change the p-value and also the peak voxel coordinate, but this seemed to be more like a post-hoc masking procedure that the masking did change the peak coordinate and the null-distribution for FWE correction, therefore, the same Z value (that was supposedly transformed from a t-value and ultimately from a TFCE value?) corresponded to a different p-value. But the cluster size in the report which was exactly the same number of voxels and obviously exceeds much more than the mask should not be just a cooincidence?  

Thank you very much!

Chen-Chia

########################################################################

To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1