Dear Antonios,
> Thank you for your answers. I would like to ask one more question :
> In cases
> when you cope with such non concavity clusters that COG is located
> outside
> the cluster's mass what kind of approach do you use to characterize
> the
> cluster's activation center? Is there an fsl's utility that can be
> used for
> that?
you can always use the coordinates of the maximum voxel.
I would also suggest that in general when you have a large irregular
cluster where the COG doesn't make sense, that cluster is likely to be
hard to interpret/characterize anyway. If you increase your initial
cluster forming thresholds there is a chance that your cluster will
become easier to interpret. There is of course also a risk it will go
away.
Good Luck Jesper
|