Dear FSLers,

in the past, I have succesfully used the cluster command to output the clusters from randomise-generated images. Using randomise with a mask usually produced very few, distinguishable clusters, whereas using wholebrain randomise usully did not yield anything. So far, so good.

Now I have encountered something that I cannot quite make any sense of. Using the cluster command on one of my whole brain randomise images, I get a huge cluster (see cluster output below). When I look at the actual image, there are many blobs distributed all over the brain. I checked the dlh and the volume and they are both fine.

Any idea as to why this time this is being regarded as one big cluster instead of many individual ones?
Is there a way how I can split the clearly anatomically very different blobs and obtain information about them (as the usual -thr, -uthr does not work here, obviously)?

Your help is greatly appreciated!

Thanks,

Hannah

Cluster Index    Voxels    P    -log10(P)    MAX    MAX X (mm)    MAX Y (mm)    MAX Z (mm)    COG X (mm)    COG Y (mm)    COG Z (mm)
1    219719    nan    nan    0.959    42    -55    -39    4.44    -29    26.7