Hi Tracula team,
I have just successfully processed my first scan through all
the Tracula steps! Though, I have to say it could have gone much smoother /
quicker if the documentation on the wiki were a little more up-to-date. I
realize it's an ongoing struggle to keep documentation current, so I thought
I'd send you my observations as I worked through all the bumps in the road.
You are probably aware of most of these issues but here they are in one big
email for you to consider:
1. I'm unable to find anything among the Tracula setup instructions
where it says which version of FSL that Tracula requires, and this version
issue was a crucial piece of the puzzle. I had been failing at step 2 with
an older version of FSL (v4.1.4), and per your suggestion I updated to
v4.1.9. However, when called by Trac-all -bedp, bedpostx ALWAYS failed to
created the "merged*.nii.gz" files for me, even after modifying it and
updating FSL versions. I'm not sure why that is but it may be helpful to
users to know they can simply bypass that step by launching bedpostx
separately in the latest FSL, and pointing it at the "dmri" folder created
during the Trac-all -prep stage. Also, we had to modify "bedpostx", as well
as "fsl_sub", to get it to launch to our cluster without issues. Once all
the fixes were in place, this step took only an hour or two to run. I then
copied bedpostx's outputs into the requisite "dmri.bedpostX" directory to
launch the -path step, and it completed in about an hour. Launching
"bedpostx_datacheck <dmri folder>" beforehand was very helpful for
troubleshooting the prep stage.
2. There is a typo in the description of the configuration file setup:
"set dorotvecs" is listed in the detailed description under Step9, but this
is incorrect and will cause errors. It's a bit confusing because it is
correctly listed as "set dorotbvecs" in the complete config file example you
provide above that section. It's unclear which one is correct and forces
users to debug something which they probably have never run before.
3. Also in the config file is the variable $FSLDIR, which I discovered
we had to set after receiving errors during mni registration. I suppose the
setting of that variable should be obvious, but it wasn't clear based on the
example which folder, exactly, should be associated with that variable name.
We set our $FSLDIR to the /bin folder and all was well. Entirely possible
this is just a lab-specific issue on our end. Since this variable needs to
be set properly, just like the $SUBJECTS_DIR, perhaps this information could
be included towards the top of the page where the rest of the required
environmental variables are mentioned?
4. Finally, in the "Tractography Outputs" section, the command line
example given needs to be updated because it shows "lh.ilf_AS_avg32_mni_flt"
as example. The "32" should be "33". The same goes for
"merged_avg32_mni_flt.mgz", obviously.
Until very recently, I was experiencing all the same issues Scott Hayes has
been reporting lately. (The singular "Data_slice_0002" folder, with missing
files, and the over-and-over messages of "1 slices processed" in a terminal
that never throws an error message despite its failure. During the
subsequent -path step, without those "merged*.nii.gz" files, I would receive
the dreaded & perplexing "Word Too Long" error. A more specific error
message regarding the missing files might be more useful for troubleshooting
purposes.)
Many thanks for all your advice and help throughout the process! We can now
move forward and do some exploration... Very exciting!
Shannon T. Buckley
SRA III - Supervisor
CIND, UCSF
4150 Clement St. (114M)
San Francisco, CA 94121
Phone: (415) 221-4810 x4251
Fax: (415) 386-3954
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is
addressed. If you believe this e-mail was sent to you in error and the e-mail
contains patient information, please contact the Partners Compliance HelpLine at
http://www.partners.org/complianceline . If the e-mail was sent to you in error
but does not contain patient information, please contact the sender and properly
dispose of the e-mail.
