Thanks Andrew!  I’ll look into these options.

 

Take care,

Laura

 

 

Hi Laura,

 

The error in the last few lines tells you what you need to know. Your install is missing a freesurfer license, which you must copy into $FREESURFER_HOME. I assume you already have one since you’ve run recons before, but if you need to re-register, you can do this at http://surfer.nmr.mgh.harvard.edu/registration.html

 

And when you combine results from different operating systems, your analysis is no longer consistent across subjects. While we try to minimize cross-platform discrepancies as much as we can, there is a possibility that low-level differences will skew (and invalidate) the results of your study. If you’re stuck running things on a macbook for now, I recommend installing a VM or docker container that matches the OS and version previously used in your analysis, and running your recons on that.

 

best

Andrew

 

        External Email - Use Caution        

Hi Andrew,

 

Here is the entirety of the terminal after the recon-all -wm command (with personal info removed).  Can you tell from this what is causing it to die?  I’m still getting everything set up so I may have an incomplete setup/download.

 

----

INFO: FreeSurfer build stamps do not match

Subject Stamp: freesurfer-Linux-centos6_x86_64-stable-pub-v6.0.0-2beb96c

Current Stamp: freesurfer-Darwin-OSX-stable-pub-v6.0.0-2beb96c

INFO: SUBJECTS_DIR is /Volumes/###

Actual FREESURFER_HOME /Applications/freesurfer

-rwx------+ 1 ###  ###\Domain Users  1689971 Mar 20 15:50 /Volumes/###/scripts/recon-all.log

Darwin ### 18.7.0 Darwin Kernel Version 18.7.0: Thu Jan 23 06:52:12 PST 2020; root:xnu-4903.278.25~1/RELEASE_X86_64 x86_64

#--------------------------------------------

#@# Mask BFS Mon Mar 23 11:55:31 EDT 2020

/Volumes/###/mri

\n mri_mask -T 5 brain.mgz brainmask.mgz brain.finalsurfs.mgz \n

threshold mask volume at 5

--------------------------------------------------------------------------

ERROR: FreeSurfer license file /Applications/freesurfer/license.txt not found.

  If you are outside the NMR-Martinos Center,

  go to http://surfer.nmr.mgh.harvard.edu/registration.html to 

  get a valid license file (it's free).

  If you are inside the NMR-Martinos Center,

  make sure to source the standard environment.

--------------------------------------------------------------------------

Darwin ### 18.7.0 Darwin Kernel Version 18.7.0: Thu Jan 23 06:52:12 PST 2020; root:xnu-4903.278.25~1/RELEASE_X86_64 x86_64

 

recon-all -s ### exited with ERRORS at Mon Mar 23 11:55:35 EDT 2020

----

 

Unfortunately, I may not have a choice about using 2 different OS to do the analysis, due to the total shutdown of our offices and additional time constraints.  Is there any other way to work around this other than starting from scratch?  If I combine the data later for analyses, what are the potential problems I could face?

 

Thanks for your help!

 

Laura

 

 

Hi Laura,

 

For any study, it is essential that you process all of your subjects on the same operating system. Even though the same FS version is used, there will be underlying system-level differences that may lead to small inconsistencies across your linux and mac results.

 

Also, are your recons dying after printing “build stamps do not match”? That message should just be a warning and not force a recon-all exit unless REQUIRE_FS_MATCH is enabled (it’s not by default).

 

Best

Andrew

 

 

        External Email - Use Caution        

Hi,

 

I have been processing datasets on my Linux on v6.0 but have recently been made to work from home where I only have access to a Mac.  I’m getting the following error message when continuing the editing on datasets that were already started on my Linux:

 

INFO: FreeSurfer build stamps do not match

Subject Stamp: freesurfer-Linux-centos6_x86_64-stable-pub-v6.0.0-2beb96c

Current Stamp: freesurfer-Darwin-OSX-stable-pub-v6.0.0-2beb96c

 

A few questions:

 

Is it feasible to override this error message for subjects that I have already started processing since they are both using v6.0?

 

I have about half of my data set already fully processed using a Linux – if I start processing the second half from scratch using my Mac, can I eventually combine the 2 parts of the dataset for comparison?

 

If I use a mix of Freesurfer builds on different data sets, can I still process data sets from the same subject using the longitudinal stream?

 

Thanks,

Laura Danielian