Odd segmentations

[Caitlin Taylor]

Hello all,

I ran a batch of hippocampal segmentations recently and most turned out fine, but a few had really funky results (e.g. segmentation was attempted in the skull/lateral temporal cortex instead of the hippocampus, screenshot attached among other examples) and others just wouldn't run (all stopped with the same error, below). I re-ran the scans with erroneous segmentations a second time and got the same results. I can't tell what would be different between these scans and the others that ran without error.

Any ideas how to troubleshoot this? I thought I would run it on the cloud to more easily report if the errors persisted, but my web token isn't working to log me into the ITK-Snap DSS.

Thank you!

Caitlin

Error message for incomplete segmentations:

**************** !! ERROR !! *******************
Validity check at end of stage 3 detected
missing files.
(multiatlas/fusion/lfseg_corr_nogray_left.nii.gz
 and 5 other files).
************************************************