[Freesurfer] Initializing FreeSurfer's recon-all with rigid registration

[Yeun Kim]

External Email - Use Caution

Hi,

How do I concatenate the resulting recon-all transform with another rigid
transform to produce a single transform?

I have the aseg.nii.gz (generated from the T1w image in T1w space) output
label in the atlas space and I would like to transform this label into the
same subject's T2w space. I am aware there is a transform that moves the
image from atlas to T1w native space (although I'm not quite sure of the
specific file names of the transforms, only the command mri_label2vol).

I have the transformation from the rigid registration between T1w and T2w
(T1w moving to T2w space) and would like to concatenate the atlas-to-T1
transform with T1-to-T2 transform to reduce the amount of error caused by
resampling (since the labels are in discrete integers).

Thank you in advance,
Yeun
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Re: [Freesurfer] aparc a2009s stats ROI names don't match ColorLUT for FS 6.0

[Hoopes, Andrew]

Hi Don,

Thanks for pointing this out. The stats labels were being taken from an older 
lookup table embedded in the aparc atlases. Putting together a fix for this in 
the development version now. If it helps, I've attached updated LUTs as well as 
a simple 5.3-to-6.0 label name translation table.

best
Andrew


From: freesurfer-boun...@nmr.mgh.harvard.edu 
 on behalf of Tim Schäfer 

Sent: Tuesday, March 10, 2020 11:29 AM
To: Freesurfer support list
Subject: Re: [Freesurfer] aparc a2009s stats ROI names don't match ColorLUT for 
FS 6.0

External Email - Use Caution

Hi,

I also have code that parses the stats files and noticed this a while ago, a 
short thread is here:

https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg64090.html

Basically just the names changed, the data is identical, so I just adapted my 
parsing code to check for both naming conventions.

Best,

Tim

> On March 6, 2020 at 1:49 AM Don Hagler  wrote:
>
>
> External Email - Use Caution
>
> I have recently started using FreeSurfer 6.0 after previously working with 
> 5.3 for several years. I found that the lh.aparc.a2009s.stats files use 
> different ROI names (e.g., G_frontomargin) than they did in FS5.3 
> (G_and_S_frontomargin). However, the FreeSurferColorLUT.txt file still has 
> the old names.
>
> What caused the ROI names to change for the a2009s.stats files? Was the 
> ColorLUT supposed to have been updated but wasn’t? Where do those shorter 
> names come from? Is there already a mapping file for the different versions 
> of the names?
>
> Because some of the parcel names in the stats files no longer match the LUT, 
> my existing code for compiling FreeSurfer results fails to load the results 
> for those ROIs with new names.
>
> I think I can find a work-around for this, such as using a mapping file or a 
> renaming rule applied to the a2009s stats ROI names, before matching to the 
> LUT.
>
> But I thought I would email this list to raise the issue. I searched the 
> archives and didn’t see something similar, so maybe no one else tries to 
> match those names to the LUT.
>
> ___
> Freesurfer mailing list
> Freesurfer@nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

--
Tim Schäfer

___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
#$Id: FreeSurferColorLUT.txt,v 1.105 2015/07/29 18:23:03 greve Exp $

#No. Label Name:R   G   B   A

0   Unknown 0   0   0   0
1   Left-Cerebral-Exterior  70  130 180 0
2   Left-Cerebral-White-Matter  245 245 245 0
3   Left-Cerebral-Cortex205 62  78  0
4   Left-Lateral-Ventricle  120 18  134 0
5   Left-Inf-Lat-Vent   196 58  250 0
6   Left-Cerebellum-Exterior0   148 0   0
7   Left-Cerebellum-White-Matter220 248 164 0
8   Left-Cerebellum-Cortex  230 148 34  0
9   Left-Thalamus   0   118 14  0
10  Left-Thalamus-Proper0   118 14  0
11  Left-Caudate122 186 220 0
12  Left-Putamen236 13  176 0
13  Left-Pallidum   12  48  255 0
14  3rd-Ventricle   204 182 142 0
15  4th-Ventricle   42  204 164 0
16  Brain-Stem  119 159 176 0
17  Left-Hippocampus220 216 20  0
18  Left-Amygdala   103 255 255 0
19  Left-Insula 80  196 98  0
20  Left-Operculum  60  58  210 0
21  Line-1  60  58  210 0
22  Line-2  60  58  210 0
23  Line-3  60  58  210 0
24  CSF 60  60  60  0
25  Left-Lesion 255 165 0   0
26  Left-Accumbens-area 255 165 0   0
27  Left-Substancia-Nigra   0   255 127 0
28  Left-VentralDC  165 42  42  0
29  Left-undetermined   135 206 235 0
30  Left-vessel 160 32  240 0
31  Left-choroid-plexus 0   200 200 0
32  Left-F3orb  100 50  100 0
33  Left-lOg135 50  74  0
34  Left-aOg122 135 50  0
35  Left-mOg51  50  135 0
36  Left-pOg74  155 60  0
37  Left-Stellate   120 62  43  0
38  Left-Porg   74  155 60  0
39  Left-Aorg   122 135 50  0
40  Right-Cerebral-Exterior  

[Freesurfer] Neuroimaging programmer job opening at Mayo Clinic

[Senjem, Matthew L., M.S.]

External Email - Use Caution

The Mayo Clinic Aging and Dementia Imaging Research Lab is an international 
leader in studies of Alzheimer's Disease and Related Dementias.  IT staff from 
the  Research Application Solutions (RAS) Unit provide data management and 
processing expertise in support of that role.  RAS has an opening for a highly 
capable person to further advance image processing within the lab.  This 
includes developing new, or improving existing, methods to extract quantitative 
biomarkers from medical images and the integration of those methods into robust 
automated pipelines.   Development is done in a collaborative environment that 
includes physicians,  PhD scientists, biostatisticians and medical physicists.

Candidates interested in this position must have a background in image 
processing and mathematical computation.   Clear communication skills, both 
oral and written, as well as demonstrated desire for continuous learning are 
essential. Ideally, candidates will have experience developing automated 
medical image processing pipelines and a fundamental understanding of medical 
imaging modalities including MRI and PET.   Experience with both 
cross-sectional and longitudinal analyses is desired.  Initial areas of work 
for the successful applicant will, depending on existing skills and interests, 
be improved functional MRI analysis or multi-modal PET/MRI fusion.

Experience coding in Python and/or Matlab is required. Fluency in the DICOM 
imaging standard is desired. Applicants should have experience developing 
version-controlled software in a Linux-based environment. Experience with 
distributed computation using resource management systems such as slurm or open 
grid engine will be helpful.  Linux system administration experience is 
considered a plus.

For more information and to apply, please visit the following link:
https://apply.mayoclinic.org/TGnewUI/Search/Home/Home?partnerid=25309=5277#jobDetails=2539715_5277





___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

[Freesurfer] Petsurfer: Gtmstats < 1, voxel wise mean is fine

[Eckbo, Ryan]

Hello experts,

I've recently run into a case the gtmstats uptake (reference region: 
cerebellum) for the pericarlcarine ROI's are as low as 0.40,
however computing the voxel wise mean using input.rescaled.nii.gz and 
mgx.ctxgm.nii.gx and aux/seg.nii.gz, the values are closer to 1,
as expected:

| vol   | psf | roi  | gtmstats_uptake | 
mean_uptake |
| - | --- |  | --- | 
--- |
| input.rescaled.nii.gz |   0 | ctx-rh-pericalcarine |   0.807 |  
1.226… |
| input.rescaled.nii.gz |   0 | ctx-lh-pericalcarine |   0.758 |  
1.083… |
| input.rescaled.nii.gz |   6 | ctx-rh-pericalcarine |   0.409 |  
1.197… |
| input.rescaled.nii.gz |   6 | ctx-lh-pericalcarine |   0.470 |  
1.058… |
| mgx.ctxgm.nii.gz  |   0 | ctx-rh-pericalcarine |   0.807 |  
1.007… |
| mgx.ctxgm.nii.gz  |   0 | ctx-lh-pericalcarine |   0.758 |  
0.888… |
| mgx.ctxgm.nii.gz  |   6 | ctx-rh-pericalcarine |   0.409 |  
0.915… |
| mgx.ctxgm.nii.gz  |   6 | ctx-lh-pericalcarine |   0.470 |  
0.809… |

This is using Freesurfer 6.0, and the PET to anatomical registration looks fine.

Any ideas why we get such bad values from the GTM method?

Thanks for any help,
Ryan

vol,psf,roi,gtmstats_uptake,mean_uptake
input.rescaled.nii.gz,0,ctx-rh-pericalcarine,0.807,1.2259972095489502
input.rescaled.nii.gz,0,ctx-lh-pericalcarine,0.758,1.083396077156067
input.rescaled.nii.gz,6,ctx-rh-pericalcarine,0.409,1.1970176696777344
input.rescaled.nii.gz,6,ctx-lh-pericalcarine,0.47,1.057787299156189
mgx.ctxgm.nii.gz,0,ctx-rh-pericalcarine,0.807,1.006852149963379
mgx.ctxgm.nii.gz,0,ctx-lh-pericalcarine,0.758,0.8876161575317383
mgx.ctxgm.nii.gz,6,ctx-rh-pericalcarine,0.409,0.9153279066085815
mgx.ctxgm.nii.gz,6,ctx-lh-pericalcarine,0.47,0.8090171217918396
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Re: [Freesurfer] overlay threshold in freeview

[Ruopeng Wang]

HI Rachele,

You can set overlay threshold from the command-line like this:

freeview -f lh.inflated:overlay=overlay_file:overlay_threshold=low,high

More information about threshold setting can be found by:

freeview -h

Best,
Ruopeng

On 3/10/20 11:29 AM, Rachele Sanfelici wrote:


External Email - Use Caution

Dear Freesurfer experts,

I created .mgh files from a dimensionality reduction analyses (similar 
to parcels from a known parcellation). These matrices have low values 
near to 0. When I overlay each of the .mgh file in freeview, it gets 
automatically thresholded with what seemed to me the median or mean of 
all the vertices' values (i.e., when looking at the "configure 
overlay" option)--however, this is not the case.
Thus, I wanted to ask how the threshold is precisely set, so that I 
can automate and incorporate it in other scripts for binarizing my 
matrices.


Thanks a lot in advance for your help,
Rachele

___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

[Freesurfer] overlay threshold in freeview

[Rachele Sanfelici]

External Email - Use Caution

Dear Freesurfer experts,

I created .mgh files from a dimensionality reduction analyses (similar to
parcels from a known parcellation). These matrices have low values near to
0. When I overlay each of the .mgh file in freeview, it gets automatically
thresholded with what seemed to me the median or mean of all the vertices'
values (i.e., when looking at the "configure overlay" option)--however,
this is not the case.
Thus, I wanted to ask how the threshold is precisely set, so that I can
automate and incorporate it in other scripts for binarizing my matrices.

Thanks a lot in advance for your help,
Rachele
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Re: [Freesurfer] aparc a2009s stats ROI names don't match ColorLUT for FS 6.0

[Tim Schäfer]

External Email - Use Caution

Hi,

I also have code that parses the stats files and noticed this a while ago, a 
short thread is here:

https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg64090.html

Basically just the names changed, the data is identical, so I just adapted my 
parsing code to check for both naming conventions.

Best,

Tim

> On March 6, 2020 at 1:49 AM Don Hagler  wrote:
> 
> 
> External Email - Use Caution
> 
> I have recently started using FreeSurfer 6.0 after previously working with 
> 5.3 for several years. I found that the lh.aparc.a2009s.stats files use 
> different ROI names (e.g., G_frontomargin) than they did in FS5.3 
> (G_and_S_frontomargin). However, the FreeSurferColorLUT.txt file still has 
> the old names.
> 
> What caused the ROI names to change for the a2009s.stats files? Was the 
> ColorLUT supposed to have been updated but wasn’t? Where do those shorter 
> names come from? Is there already a mapping file for the different versions 
> of the names?
> 
> Because some of the parcel names in the stats files no longer match the LUT, 
> my existing code for compiling FreeSurfer results fails to load the results 
> for those ROIs with new names.
> 
> I think I can find a work-around for this, such as using a mapping file or a 
> renaming rule applied to the a2009s stats ROI names, before matching to the 
> LUT.
> 
> But I thought I would email this list to raise the issue. I searched the 
> archives and didn’t see something similar, so maybe no one else tries to 
> match those names to the LUT.
> 
> ___
> Freesurfer mailing list
> Freesurfer@nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

-- 
Tim Schäfer

___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

[Freesurfer] Group analysis with qdec: qcache change this?

[Miguel Ángel Rivas Fernández]

External Email - Use Caution

Dear freesurfer experts,

Before running group analysis in qdec is necessary run the qcache command
over all subjects but, does this command change some segmentation file (eg.
rh/lh Pial or white surface) or other importants file?
We would also like to perform external analysis using the output metrics
obtained with the recon-all.  This is why we want to know if we can safely
use this command without affecting any segmentation measure.

Thanks in advance


Best regards,
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Re: [Freesurfer] freesurfer- recon-all "exited with ERRORS"

[Tim Schäfer]

External Email - Use Caution

Maybe Doug or Bruce should comment on that, but afaik that should not affect 
your analysis at all: recon-all will resample your data to 256x256x256 (as part 
of creating what FreeSurfer calls the 'conformed' version of the image) anyways.

The error you get, if I understand correctly from the logs, is that *after* the 
resampling, the FOV *still* is not 256.

Best,

Tim

> On March 9, 2020 at 7:34 PM Frank Chau  wrote:
> 
> 
> External Email - Use Caution
> 
> Dear Tim,
> 
> The FOV of my T1 series is 256 x 240. Is it the FOV of the T1 series will 
> affect the process Recon-all? If yes, how would it impact my output results?
> 
> Regards
> 
> Frank
> 
> From: freesurfer-boun...@nmr.mgh.harvard.edu 
>  on behalf of Frank Chau 
> 
> Sent: 06 March 2020 22:17
> To: Freesurfer support list
> Subject: Re: [Freesurfer] freesurfer- recon-all "exited with ERRORS"
> 
> External Email - Use Caution
> 
> Dear Tim,
> 
> The FOV of my T1 series is 256 x 240. Is it the FOV of the T1 series will 
> affect the process Recon-all? If yes, how would it impact my output results?
> 
> Regards
> 
> Frank
> 
> 
> 
> From: freesurfer-boun...@nmr.mgh.harvard.edu 
>  on behalf of Tim Schäfer 
> 
> Sent: 05 March 2020 2:25
> To: Freesurfer support list
> Subject: Re: [Freesurfer] freesurfer- recon-all "exited with ERRORS"
> 
> External Email - Use Caution
> 
> Thanks for providing the image. Looks fine to me as well.
> 
> What is a bit strange is that when running `mri_info` on the file you 
> provided, it shows a FOV of 256 (not 284!) for me, so I have no idea why this 
> error is triggered in recon-all.
> 
> Maybe somebody else has an idea...
> 
> Tim
> 
> > On March 4, 2020 at 4:47 PM Frank Chau  wrote:
> >
> >
> > External Email - Use Caution
> >
> > Dear Tim,
> >
> > I have checked the orig.mgz file and nothing usual was found (please see 
> > the attachment image)
> >
> > For downloading my orig.mgz: 
> > https://www.dropbox.com/s/crkndenx8ev957v/orig.mgz?dl=0
> >
> > Regards
> >
> > Frank
> >
> > 
> > From: freesurfer-boun...@nmr.mgh.harvard.edu 
> >  on behalf of Tim Schäfer 
> > 
> > Sent: 04 March 2020 23:27
> > To: Freesurfer support list
> > Subject: Re: [Freesurfer] freesurfer- recon-all "exited with ERRORS"
> >
> > External Email - Use Caution
> >
> > Dear Frank,
> >
> > if you search the log in your Dropbox for 'error', this comes up:
> >
> > 
> > ERROR! FOV=284.000 > 256
> > Include the flag -cw256 with recon-all!
> > Inspect orig.mgz to ensure the head is fully visible.
> > 
> >
> > Did you have a look at the orig.mgz?
> >
> > Best,
> >
> > Tim
> >
> >
> > > On March 4, 2020 at 2:59 PM Frank Chau  wrote:
> > >
> > >
> > > External Email - Use Caution
> > >
> > > Hello FreeSurfer Developers,
> > >
> > > I ran a recon-all on a recent batch of T1 Nifti files and all ended with 
> > > "exited with ERRORS". For example: "recon-all -s Chan_Pui_Kim_42M_T1 
> > > exited with ERRORS at Sat Jan 11 01:32:55 HKT 2020".
> > >
> > > I've searched the email archive list and I found other researcher 
> > > encountered same problem before. However, I am quite sure their root 
> > > causes are not same as my case. 3 possible causes were found in the email 
> > > archive which about "exited with ERRORS":
> > >
> > >   1.  The version of freesurfer (I have already used the latest 6.0.0 
> > > version)
> > >   2.  Hard disk space (My computer have enough hard disk space)
> > >   3.  sharing folder between virtual machine and original platform. (I 
> > > use freesurfer in Mac OS which do not have this problem)
> > >
> > > I uploaded the recon-all.log to my dropbox in case it's of any use.
> > >
> > >
> > > 1) FreeSurfer version: Darwin-OSX-stable-pub-v6.0.0-2beb96c
> > >
> > > 2) Platform: MAC OS X 10.11.6
> > >
> > > 3) uname -a: Darwin PFCs-MacBook-Pro.local 15.6.0 Darwin Kernel Version 
> > > 15.6.0: Thu Jun 21 20:07:40 PDT 2018; 
> > > root:xnu-3248.73.11~1/RELEASE_X86_64 x86_64
> > >
> > > 4) recon-all.log, please download at:
> > >
> > > https://www.dropbox.com/s/bxb2vyhz3koo83s/recon-all.log-%20Terminal%20Output%20log%2020%20Jan%202020%20.txt?dl=0
> > >
> > >
> > >
> > > Regards
> > >
> > > Frank
> > >
> > > ---
> > >
> > > FRANK  C Y CHAU
> > >
> > >
> > >
> > > Research assistant
> > >
> > > Division of Neurology
> > >
> > > Department of Medicine
> > >
> > >
> > >
> > > Tel: +852 6340 2553
> > >
> > > Fax: 2872 5828
> > >
> > >
> > >
> > > E-mail : cyfr...@hku.hk
> > >
> > >
> > >
> > >
> > >
> > > Rm 304, 3/F, New Clinical building, Queen Mary Hospital, 102 Pokfulam 
> > > Road, Hong Kong
> > >
> > >
> > > ___
> > > Freesurfer mailing list
> > > Freesurfer@nmr.mgh.harvard.edu
> > >