Dear Aditya,
Yes, there are 21 clusters and cluster number 10 would contain 14 voxels which
must be connected together to form a single cluster (i.e. contiguous). The same
z-max value in different clusters has no particular meaning and is certainly possible
as each cluster is formed from voxels that are over the cluster-forming threshold.
It does not matter how far the voxel values are over that threshold and so the
maximum value does not enter into the calculation of the cluster or the ranking.
It is reported in the table as it is often of interest, but it does not define the clusters
in any way.
All the best,
Mark
On 22 Jul 2011, at 18:24, Aditya Kumar Kasinadhuni wrote:
> Thank you very much Mark. Yes the input to the cluster command was a filename_tfce_corrp_tstat1 image generated by randomise.
>
> Just to clarify, what the table indicates, is that all in all in my image there are 21 clusters right?
>
> And for example, in my data for the cluster number 10, the table indicates that there are 14 contiguous voxels with a max corrp value of 0.957 occurring at the co-ordinates z-max x, y ,z respectively or is it just 14 voxels in the image as a whole and need not be contiguous?
>
> What would the recurrence of the same z-max value for different clusters signify? Just that there are both two different clusters (with different #voxels) with the same max z value?
>
> Thank you
>
|