Hi,
I am having difficulties when I rerun first-level post-stats. This
happens when I add a couple additional contrasts. Essentially, the log
file states (1) that I have an 'ERROR:: Empty mask image' or (2) I get
drastically different stats values. I have replicated the same error
across platforms, all running fsl 4.0. The error goes away when I run
the entire analysis over, but my design is quite long and I would prefer
not having to do this!
Here's an example problem 1:
/Applications/fsl/bin/fslstats thresh_zstat4 -l 0.0001 -R
ERROR:: Empty mask image
ERROR:: Empty mask image
ERROR:: Empty mask image
ERROR:: Empty mask image
ERROR:: Empty mask image
ERROR:: Empty mask image
0.000000 0.000000
This is what the log looked like when the contrast was first run:
/Applications/fsl/bin/fslstats thresh_zstat4 -l 0.0001 -R
1.644862 3.349561
I believe the problem arises when the contrast manager processes the new
design.con file. For some reason, the zstat4 file is processed as an
'empty' file when valid stats do exist.
I get this:
echo 0 > thresh_zstat4.vol
zstat4: DLH=0.0101663 VOLUME=0 RESELS=454.153
rather than this (the original):
echo 42375 > thresh_zstat4.vol
zstat4: DLH=0.0101663 VOLUME=42375 RESELS=454.153
This problem has been noted previously, yet I don't think a resolution
was reached. See:
http://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0801&L=FSL&P=R49090
Here's problem 2:
Original:
/Applications/fsl/bin/fslstats thresh_zstat25 -l 0.0001 -R
1.645139 39.740047
Error:
/Applications/fsl/bin/fslstats thresh_zstat25 -l 0.0001 -R
1.645410 18.897028
Cheers
Nick
|