Hi @Sarbani,
The same issue was reported and discussed here. I might need to reduce the implied severity of the reporting of that issue so that people don’t panic about it; as I said in the other thread, having one solitary voxel where the tissues don’t perfectly sum to 1 is unlikely to have any effect whatsoever.
Nevertheless, it would be handy to know exactly what’s happening in that voxel through the script that’s resulting in a residual error:
5ttgen fsl T1w_acpc_dc_restore_brain.nii.gz 5TT.mif -premasked -nocleanup
5ttcheck 5TT.mif -masks 5TTerr_
maskdump 5TTerr_5TT.mif # Print out the voxel location
# The following access image intensities just in that one voxel
for i in `seq 0 2`: do
mrdump 5ttgen-tmp-*/T1_masked_pve_${i}.nii* -mask 5TTerr_5TT.mif
done
mrdump 5ttgen-tmp-*/all_sgms.mif -mask 5TTerr_5TT.mif
(or something like that)
I have performed this step on this data before and executed without error but now its giving this error.
If it was prior to version 3.0_RC1, this check was simply not performed. Alternatively, one of the commands used by the script may not be perfectly deterministic.
Rob