Hello Rosalia,
cluster is a general tool, so an appropriate threshold is dependent on the input data. 2.5 is an example value for t-statistics ( e.g. as output by FEAT ). For FWE-corrected p-value images, where the range by definition cannot be outside of 0 to 1, 0.95 would be a sensible value to use.
Hope this helps,
Kind Regards
Matthew
> Dear FSL experts,
>
> Sorry for that simple question, but I expected to see in the cluster command line, in the fslwiki. -t 0.95 (as the threshold, instead of 2.5...that it has nonsense for me when applying to tfce_corrp_tstat images....
>
> I would be very happy if someone of you could explain me this.
>
> Thank you a lot for your help.
>
> Rosalia.
|