Dear FSL experts,
I am running FLAME 1+2 in higher-level FEAT analyses (on lower-level COPEs). For some of the analyses, the zstat-images become "stretched" in the Z-dimension or there are several "volumes" on top of each other. Furthermore, it doesn't produce any thresh-images. All the lower level images are in 2mm MNI space (91x109x91), but the higher-level ones that went bad are 91x109x364 or 224 etc. I don't have this problem when running the same analyses with FLAME 1 only.
Also, I am running these analyses on a cluster network, where I have assigned a "scratch working space" of 10 GB for each analysis, which should be enough for these analyses?
Many thanks,
Dan
|