Hi Iselin,

This can happen and suggests that the significant results found in the whole-brain analysis were "enhanced" by areas of sub-threshold signal that were connected to them in the TFCE's support region. A smaller ROI cuts this support region, then the findings may no longer be significant.

Hope this helps.

All the best,


On 17 August 2017 at 05:22, Iselin Meluken <[log in to unmask]> wrote:


Dear experts


In an experiment testing activity to emotional faces over baseline I am concerned that significant clusters observed at whole brain level does not reach significance in ROI analysis, although the ROI mask includes the specific area (insula and amygdala) using PALM and the TFCE option. Trying to understand why, I tried to rerun the same analysis with C=3.1 instead of TFCE, and then these areas do reach cluster-wise significance within the ROI mask as in whole brain analysis.


Could you help me understand why this discrepancy happens? Sorry if this is a naive question. My theory is that this somehow happens because the clusters disappearing are part of a larger cluster significant at whole brain level, where part of this larger area is NOT included in the ROI mask (e.g. putamen), and that this is somehow treated differently in the TFCE approach compared with the traditional cluster threshold defining method.  


All the best



Denne e-mail indeholder fortrolig information. Hvis du ikke er den rette modtager af denne e-mail eller hvis du modtager den ved en fejltagelse, beder vi dig venligst informere afsender om fejlen ved at bruge svarfunktionen. Samtidig bedes du slette e-mailen med det samme uden at videresende eller kopiere den.