Hi Tyler - In general this means that for some reason the pathway doesn't move at all from the location where it's initialized when the algorithm is run. There are a few different reasons why that'd happen, the mask being only one of them.

If you could find the dmri_paths command line in trac-all.log, add "--debug" to it and run it again and send the log file, I can try to track down what's going on. The command line will probably contain all the tracts, but you can run it only for one that fails.

Thanks, a.y

On Wed, 30 Nov 2011, Tyler Blazey wrote:

Hi Anastasia,

Thanks so much for your quick reply. As far as I can tell the masks seem fine. There was a small amount of erosion around the edge of the brain when the recon mask was used. However, the end points of the tract were still within the mask. Just in case I reran trac-all with the 'usemaskanat = 0' setting and a more generous bet threshold. Even with these changes I still saw the problem with the bbregister version of a tract. Is there anything else I should check/change?

Thanks again,

-Tyler

On Nov 30, 2011, at 4:00 PM, Anastasia Yendiki wrote:

...

Hi Tyler - Sometimes this happens if the brain mask (which will come either from the FS recon or from the DWIs themselves, depending on your settings) is missing part of that tract (usually one of the end region). Have you checked your brain masks?

a.y

On Wed, 30 Nov 2011, Tyler Blazey wrote:

...

Hi List,

I recently ran a group of around 70 subjects through trac-all. I have started to check the results, and have noticed that Tracula will often produce a tract with a really small volume and a posterior distribution that consists of just 0s and 1000s. When this occurs, the end point images usually contain only one voxel greater than 0. I am not sure what, if anything, is going wrong. The original recons are all of good quality, and the registrations to MNI space look good (the end point images look fine in MNI space as well). I also noticed this same issue in the right SLFP (bbregister version) of the Tracula tutorial subject.

Is this a fixable error, or is it actually expected behavior? Thanks for any help.

• Tyler

---

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.

Thanks, Tyler. It looks like the initial pathway is problematic in that it goes off the white matter (look at the "Reject due to" lines in the log for exact voxel coords where this happens).

This could mean that the subject is not well aligned to the atlas in that area, or that you just got unlucky. If it's the latter, you can try "set reinit = 1" in your dmrirc and rerun that pathway and see if you get a better result with the new initialiation.

Let me know if this makes sense.

a.y

On Thu, 1 Dec 2011, blazey@artsci.wustl.edu wrote:

Hi Anastasia,

Attached is the log.txt produced by dmri_paths for the failed tract. I also have included the output of the terminal command in case that helps. Thanks again for all of your help.

• Tyler

...

Hi Tyler - In general this means that for some reason the pathway doesn't move at all from the location where it's initialized when the algorithm is run. There are a few different reasons why that'd happen, the mask being only one of them.

If you could find the dmri_paths command line in trac-all.log, add "--debug" to it and run it again and send the log file, I can try to track down what's going on. The command line will probably contain all the tracts, but you can run it only for one that fails.

Thanks, a.y

On Wed, 30 Nov 2011, Tyler Blazey wrote:

...

Hi Anastasia,

Thanks so much for your quick reply. As far as I can tell the masks seem fine. There was a small amount of erosion around the edge of the brain when the recon mask was used. However, the end points of the tract were still within the mask. Just in case I reran trac-all with the 'usemaskanat = 0' setting and a more generous bet threshold. Even with these changes I still saw the problem with the bbregister version of a tract. Is there anything else I should check/change?

Thanks again,

-Tyler

On Nov 30, 2011, at 4:00 PM, Anastasia Yendiki wrote:

...

Hi Tyler - Sometimes this happens if the brain mask (which will come either from the FS recon or from the DWIs themselves, depending on your settings) is missing part of that tract (usually one of the end region). Have you checked your brain masks?

a.y

On Wed, 30 Nov 2011, Tyler Blazey wrote:

...

Hi List,

I recently ran a group of around 70 subjects through trac-all. I have started to check the results, and have noticed that Tracula will often produce a tract with a really small volume and a posterior distribution that consists of just 0s and 1000s. When this occurs, the end point images usually contain only one voxel greater than 0. I am not sure what, if anything, is going wrong. The original recons are all of good quality, and the registrations to MNI space look good (the end point images look fine in MNI space as well). I also noticed this same issue in the right SLFP (bbregister version) of the Tracula tutorial subject.

Is this a fixable error, or is it actually expected behavior? Thanks for any help.

• Tyler

---

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.

Great! You just need to rerun cases where the original initialization was bad.

On Thu, 1 Dec 2011, blazey@artsci.wustl.edu wrote:

Anastasia,

Rerunning with the reinit = 1 option worked great. One last question. If I want to compare tracts across subjects, should I rerun every subject with the reinitialization, or is it ok to only rerun the subjects/tracts that need it? Thanks for all you help.

-Tyler

...

Thanks, Tyler. It looks like the initial pathway is problematic in that it goes off the white matter (look at the "Reject due to" lines in the log for exact voxel coords where this happens).

This could mean that the subject is not well aligned to the atlas in that area, or that you just got unlucky. If it's the latter, you can try "set reinit = 1" in your dmrirc and rerun that pathway and see if you get a better result with the new initialiation.

Let me know if this makes sense.

a.y

On Thu, 1 Dec 2011, blazey@artsci.wustl.edu wrote:

...

Hi Anastasia,

Attached is the log.txt produced by dmri_paths for the failed tract. I also have included the output of the terminal command in case that helps. Thanks again for all of your help.

• Tyler

...

Hi Tyler - In general this means that for some reason the pathway doesn't move at all from the location where it's initialized when the algorithm is run. There are a few different reasons why that'd happen, the mask being only one of them.

If you could find the dmri_paths command line in trac-all.log, add "--debug" to it and run it again and send the log file, I can try to track down what's going on. The command line will probably contain all the tracts, but you can run it only for one that fails.

Thanks, a.y

On Wed, 30 Nov 2011, Tyler Blazey wrote:

...

Hi Anastasia,

Thanks so much for your quick reply. As far as I can tell the masks seem fine. There was a small amount of erosion around the edge of the brain when the recon mask was used. However, the end points of the tract were still within the mask. Just in case I reran trac-all with the 'usemaskanat = 0' setting and a more generous bet threshold. Even with these changes I still saw the problem with the bbregister version of a tract. Is there anything else I should check/change?

Thanks again,

-Tyler

On Nov 30, 2011, at 4:00 PM, Anastasia Yendiki wrote:

...

Hi Tyler - Sometimes this happens if the brain mask (which will come either from the FS recon or from the DWIs themselves, depending on your settings) is missing part of that tract (usually one of the end region). Have you checked your brain masks?

a.y

On Wed, 30 Nov 2011, Tyler Blazey wrote:

> Hi List, > > I recently ran a group of around 70 subjects through trac-all. I > have > started to check the results, and have noticed that Tracula will > often > produce a tract with a really small volume and a posterior > distribution that consists of just 0s and 1000s. When this occurs, > the > end point images usually contain only one voxel greater than 0. I am > not sure what, if anything, is going wrong. The original recons are > all of good quality, and the registrations to MNI space look good > (the > end point images look fine in MNI space as well). I also noticed > this > same issue in the right SLFP (bbregister version) of the Tracula > tutorial subject. > > Is this a fixable error, or is it actually expected behavior? Thanks > for any help. > > - Tyler > > > > _______________________________________________ > 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.