Freesurfer

freesurfer@nmr.mgh.harvard.edu

5 participants
30076 discussions

Running selxavg with many (~5000) different conditions
by Anna Mynick 27 Nov '18

27 Nov '18

External Email - Use Caution Hi, I'm currently running an analysis that requires defining each of 5000 stimuli as a condition during processing, and for each subject, I have about 140 runs. As a result, selxavg is taking a very long time and huge amounts of memory to run. For instance, it has gotten about 5 runs into "GLS Beta Pass" after 60 hours of running with 200G. Is there a way to (a) predict based on runs and conditions how long and how much compute power selxavg will need to run? (b) speed up running selxavg (maybe there is a lighter-weight version of selxavg that requires less compute power? Thanks very much for any input, Anna

1 0

segmentHA_T2.sh: amygdala recon-all command stoped.
by qimo601 27 Nov '18

27 Nov '18

External Email - Use Caution hi,FreeSurfer experts. I am trying to use segmentHA_T2.sh command to segment the T2 FLAIR Sequence MRI Data. Also I am now only having T2 Data. But there was a mistake in the first step. Sincerely ask for help. (1)First I use recon-all command to generate the subject file from my nii file.But it stop in Correct Defect step. I have run three times after 48 hours. It also stop there. I dont't know why. Can you help me? my command: recon-all -all -i nii-004/ADNI_004_3D_FLAIR.nii.gz -s Subj004 -itkthreads 4 The output of the terminal is as follows： CORRECTING DEFECT 45 (vertices=24, convex hull=53, v0=135114) After retessellation of defect 45 (v0=135114), euler #=96 (83072,238273,155297) : difference with theory (-19) = -115 CORRECTING DEFECT 46 (vertices=97, convex hull=30, v0=136402) After retessellation of defect 46 (v0=136402), euler #=97 (83078,238301,155320) : difference with theory (-18) = -115 CORRECTING DEFECT 47 (vertices=74, convex hull=24, v0=138982) After retessellation of defect 47 (v0=138982), euler #=98 (83088,238338,155348) : difference with theory (-17) = -115 CORRECTING DEFECT 48 (vertices=83, convex hull=55, v0=139717) After retessellation of defect 48 (v0=139717), euler #=99 (83098,238391,155392) : difference with theory (-16) = -115 CORRECTING DEFECT 49 (vertices=141431, convex hull=17606, v0=149662) (2)Because of the above mistakes, I haven't tried the next step yet. segmentHA_T2.sh subject004 ADNI_004_3D_FLAIR.nii.gz T2 0 (3)Attached is the log(recon-all.log). (4)My envientment as follows: Ubuntu 16.04 amd_x64 freesurfer-linux-centos6_x86_64-dev.tar.gz(21-Nov-2018) My data type: 3D-FLAIR (Pixel Spacing X=1.0 mm; Pixel Spacing Y=1.0 mm; Pulse Sequence=SE/IR; Slice Thickness=1.2000000476837158 mm; TE=441.0 ms; TI=1650.0 ms; TR=4800.0 ms; Weighting=T2;) Thanks, Qimo

2 3

Combining voxel-wise and vertex-wise data
by Carola Mayer 26 Nov '18

26 Nov '18

External Email - Use Caution Hello Freesurfer Developers, I have a general question about how to analyse a certain combination of data. I have an independent variable that contains voxel-wise data and I have a dependent variable that contains vertex-wise cortical thickness values. I would like to do a between-subjects analysis correlating the voxel-wise and vertex-wise data. What is a valid approach to do this? Should I first register the voxel-wise data onto the vertex-wise data and then register them into an average space to do group analysis? Or is a different method more valid? Thanks in advance for your time and help, Carola

2 1

Gray Matter Manual Edit Reverting
by David S Lee 26 Nov '18

26 Nov '18

External Email - Use Caution Hello Experts, I am currently using FreeSurfer version 6.0 to make manual edits to T1w images. In an earlier version 5.3, when GM has been erroneously removed, I used the "clone tool" to bring back voxels based on the original T1w. However, this tool seems to be not working in 6.0. Would you be able to help me troubleshoot this? If there is a better alternative to revert GM removal, please let me know! Thank you for your help. -- David S. Lee Research Specialist Center for Healthy Minds University of Wisconsin - Madison (608) 890-1115

2 3

Re: [Freesurfer] [EXTERNAL] Re: extracting FA from FA_skeletonised with freesurfer parcellation
by Yendiki, Anastasia 26 Nov '18

26 Nov '18

In that case, it seems that the low values are introduced when transforming the skeleton from template space to individual space. If this is so, it has nothing to do with using the freesurfer segmentation per se. I would ask on the FSL list. On Nov 26, 2018 3:41 PM, "Winkelbeiner, Stephanie A" <Swinkelbei(a)northwell.edu> wrote: External Email - Use Caution So, I initially did that with the all_FA_skeletonised and got values around 0.6 for the Cc as expected. Interestingly enough though, when I do the same with the individual FA_skeletons that I obtained with the tbss_deproject command the values are around 0.1… Could you think of a reason why that is? From: freesurfer-bounces(a)nmr.mgh.harvard.edu [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] On Behalf Of Yendiki, Anastasia Sent: Monday, November 26, 2018 1:58 PM To: Freesurfer support list <freesurfer(a)nmr.mgh.harvard.edu> Subject: Re: [Freesurfer] [EXTERNAL] Re: extracting FA from FA_skeletonised with freesurfer parcellation External Email. Use Caution. You don't want to be averaging zero voxels, so masking makes sense. When you click on the voxels inside one of those segmentation regions (e.g. one of the CC regions) in freeview, are the FA values that you see high? On Nov 25, 2018 8:42 AM, "Winkelbeiner, Stephanie A" <Swinkelbei(a)northwell.edu<mailto:Swinkelbei@northwell.edu>> wrote: External Email - Use Caution Hi Anastasia, I have checked and they do align. However the freesurfer parcels naturally cover more of the brain than the FA skeletons. I was wondering if this was the problem? Therefore I added a binary mask of the FA skeletons. I also tried subtracting the parcels from the FA skeletons with fslmaths to make sure that only the area of the skeleton is read out. But that doesn't give convincing results either. Anyone having other suggestions? Thanks, Steph Sent from my Samsung Galaxy smartphone. -------- Original message -------- From: "Yendiki, Anastasia" <AYENDIKI(a)mgh.harvard.edu<mailto:AYENDIKI@mgh.harvard.edu>> Date: 23/11/18 17:50 (GMT-05:00) To: Freesurfer support list <freesurfer(a)nmr.mgh.harvard.edu<mailto:freesurfer@nmr.mgh.harvard.edu>> Subject: [EXTERNAL] Re: [Freesurfer] extracting FA from FA_skeletonised with freesurfer parcellation Have you checked if the registered segmentation and the FA map align correctly in freeview? On Nov 23, 2018 1:27 PM, "Winkelbeiner, Stephanie A" <Swinkelbei(a)northwell.edu<mailto:Swinkelbei@northwell.edu>> wrote: Hi, I try to extract FA values from the FSL FA skeleton using the freesurfer aparc.a2009s+aseg+2mm parcellations. I did the following: 1. Recon-all to obtain à aparc.a2009s+aseg+2mm 2. TBSS in FSL to obtain à all_FA_skeletonised 3. Tbss_deproject to split up all_FA_skeletonised into individual skeletons (native space) (*_FA_FA_all_FA_skeletonised.nii.gz) 4. Non-linear transformation of aparc.a2009s+aseg+2mm to subjects’ native space (parc2dwi.nii.gz) 5. Creating binary mask for each subjects FA skeleton to exclude 0 voxels (*_FA_skeleton_mask.nii.gz) 6. Extracting FA values using mri_segstats with the following command: cd subj1/ mri_segstats \ --seg mri/parc2dwi.nii.gz \ --ctab $FREESURFER_HOME/FreeSurferColorLUT.txt \ --i skeletons/SPN01_CMH_0001_01_FA_FA_all_FA_skeletonised.nii.gz \ --mask skeletons/SPN01_CMH_0001_01_FA_skeleton_mask.nii.gz \ --excludeid 0 \ --sum skeletons_FA.stats The FA values I obtain are around 0.1 which is highly unlikely, especially for regions such as the corpus callosum. Is there some obvious error in my approach? Or can someone recommend a different approach? I would be very grateful for some help! Thanks, Steph Stephanie Winkelbeiner, MSc Research Coordinator Psychiatry Research Zucker Hillside Hospital 75-59 263 Street Glen Oaks, NY 11004 Tel: (718) 470-4588 Email: swinkelbei(a)northwell.edu<mailto:swinkelbei@northwell.edu> Northwell Health Visit us at Northwell.edu<http://northwell.edu/> The information contained in this electronic e-mail transmission and any attachments are intended only for the use of the individual or entity to whom or to which it is addressed, and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If the reader of this communication is not the intended recipient, or the employee or agent responsible for delivering this communication to the intended recipient, you are hereby notified that any dissemination, distribution, copying or disclosure of this communication and any attachment is strictly prohibited. If you have received this transmission in error, please notify the sender immediately by telephone and electronic mail, and delete the original communication and any attachment from any computer, server or other electronic recording or storage device or medium. Receipt by anyone other than the intended recipient is not a waiver of any attorney-client, physician-patient or other privilege. The information contained in this electronic e-mail transmission and any attachments are intended only for the use of the individual or entity to whom or to which it is addressed, and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If the reader of this communication is not the intended recipient, or the employee or agent responsible for delivering this communication to the intended recipient, you are hereby notified that any dissemination, distribution, copying or disclosure of this communication and any attachment is strictly prohibited. If you have received this transmission in error, please notify the sender immediately by telephone and electronic mail, and delete the original communication and any attachment from any computer, server or other electronic recording or storage device or medium. Receipt by anyone other than the intended recipient is not a waiver of any attorney-client, physician-patient or other privilege. The information contained in this electronic e-mail transmission and any attachments are intended only for the use of the individual or entity to whom or to which it is addressed, and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If the reader of this communication is not the intended recipient, or the employee or agent responsible for delivering this communication to the intended recipient, you are hereby notified that any dissemination, distribution, copying or disclosure of this communication and any attachment is strictly prohibited. If you have received this transmission in error, please notify the sender immediately by telephone and electronic mail, and delete the original communication and any attachment from any computer, server or other electronic recording or storage device or medium. Receipt by anyone other than the intended recipient is not a waiver of any attorney-client, physician-patient or other privilege.

2 1

Re: [Freesurfer] [EXTERNAL] Re: extracting FA from FA_skeletonised with freesurfer parcellation
by Winkelbeiner, Stephanie A 26 Nov '18

26 Nov '18

External Email - Use Caution Hi Anastasia, I have checked and they do align. However the freesurfer parcels naturally cover more of the brain than the FA skeletons. I was wondering if this was the problem? Therefore I added a binary mask of the FA skeletons. I also tried subtracting the parcels from the FA skeletons with fslmaths to make sure that only the area of the skeleton is read out. But that doesn't give convincing results either. Anyone having other suggestions? Thanks, Steph Sent from my Samsung Galaxy smartphone. -------- Original message -------- From: "Yendiki, Anastasia" <AYENDIKI(a)mgh.harvard.edu> Date: 23/11/18 17:50 (GMT-05:00) To: Freesurfer support list <freesurfer(a)nmr.mgh.harvard.edu> Subject: [EXTERNAL] Re: [Freesurfer] extracting FA from FA_skeletonised with freesurfer parcellation Have you checked if the registered segmentation and the FA map align correctly in freeview? On Nov 23, 2018 1:27 PM, "Winkelbeiner, Stephanie A" <Swinkelbei(a)northwell.edu> wrote: Hi, I try to extract FA values from the FSL FA skeleton using the freesurfer aparc.a2009s+aseg+2mm parcellations. I did the following: 1. Recon-all to obtain --> aparc.a2009s+aseg+2mm 2. TBSS in FSL to obtain --> all_FA_skeletonised 3. Tbss_deproject to split up all_FA_skeletonised into individual skeletons (native space) (*_FA_FA_all_FA_skeletonised.nii.gz) 4. Non-linear transformation of aparc.a2009s+aseg+2mm to subjects’ native space (parc2dwi.nii.gz) 5. Creating binary mask for each subjects FA skeleton to exclude 0 voxels (*_FA_skeleton_mask.nii.gz) 6. Extracting FA values using mri_segstats with the following command: cd subj1/ mri_segstats \ --seg mri/parc2dwi.nii.gz \ --ctab $FREESURFER_HOME/FreeSurferColorLUT.txt \ --i skeletons/SPN01_CMH_0001_01_FA_FA_all_FA_skeletonised.nii.gz \ --mask skeletons/SPN01_CMH_0001_01_FA_skeleton_mask.nii.gz \ --excludeid 0 \ --sum skeletons_FA.stats The FA values I obtain are around 0.1 which is highly unlikely, especially for regions such as the corpus callosum. Is there some obvious error in my approach? Or can someone recommend a different approach? I would be very grateful for some help! Thanks, Steph Stephanie Winkelbeiner, MSc Research Coordinator Psychiatry Research Zucker Hillside Hospital 75-59 263 Street Glen Oaks, NY 11004 Tel: (718) 470-4588 Email: swinkelbei(a)northwell.edu<mailto:swinkelbei@northwell.edu> Northwell Health Visit us at Northwell.edu<http://northwell.edu/> The information contained in this electronic e-mail transmission and any attachments are intended only for the use of the individual or entity to whom or to which it is addressed, and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If the reader of this communication is not the intended recipient, or the employee or agent responsible for delivering this communication to the intended recipient, you are hereby notified that any dissemination, distribution, copying or disclosure of this communication and any attachment is strictly prohibited. If you have received this transmission in error, please notify the sender immediately by telephone and electronic mail, and delete the original communication and any attachment from any computer, server or other electronic recording or storage device or medium. Receipt by anyone other than the intended recipient is not a waiver of any attorney-client, physician-patient or other privilege. The information contained in this electronic e-mail transmission and any attachments are intended only for the use of the individual or entity to whom or to which it is addressed, and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If the reader of this communication is not the intended recipient, or the employee or agent responsible for delivering this communication to the intended recipient, you are hereby notified that any dissemination, distribution, copying or disclosure of this communication and any attachment is strictly prohibited. If you have received this transmission in error, please notify the sender immediately by telephone and electronic mail, and delete the original communication and any attachment from any computer, server or other electronic recording or storage device or medium. Receipt by anyone other than the intended recipient is not a waiver of any attorney-client, physician-patient or other privilege.

3 3

Bar/scatter plots for Interaction effects
by Azeez, Azeezat 26 Nov '18

26 Nov '18

External Email - Use Caution Hello, I used FSGD option to run statistical test on 3 fixed factors each with 2 contrast levels. (DEVELOPMENT : old vs young, SEX: male vs female, DIAGNOSE: Control Vs disease ) I report significant interaction effects of DEVxSEXxDIA at a spefic cluster. How can I plot groups to observe trends (ie: young male control). Which output file can I potentially export into Excel/SPSS to make this figure? Thank You, -- Azeezat Azeez

2 4

aseg.stats calculations
by Song, Da-Yea 26 Nov '18

26 Nov '18

External Email - Use Caution Hello, I had a question regarding the aseg.stats and was hoping to get some help. 1. How is the TotalGray calculated? On the wiki page it mentioned that it is the sum of lhCortex + rhCortex + SubCortGray + CerebellumGM. I tried calculating this based on the formula but got different results. a. TotalGray (from the aseg.stats) = 751765.612392 b. 298534.487540 (lhCortex) + 287682.124852 (rhCortex) + 58860.000000 (SubCortGray) + 49.3682 (L_cerebellum-cortex) + 50.1232 (R_cerebellum-cortex) = 645176.103792 2. We have subjects that were processed in v4.0.4 and v6.0.0. Since the TotalGray measure did not exist in the v.4.0.4, what could we use to calculate the volume? Thank you so much in advance for your help! Best, Da-Yea ---------------------------------------------------------------------- Disclaimer: The materials in this e-mail are private and may contain Protected Information. Please note that e-mail communication is not encrypted by default. You have the right to request further emails be encrypted by notifying the sender. Your continued use of e-mail constitutes your acknowledgment of these confidentiality and security limitations. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying, distribution, or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this e-mail in error, please immediately notify the sender via telephone or return e-mail.

2 3

myelin analysis with cortical thickness pvr
by Damien MARIE 26 Nov '18

26 Nov '18

External Email - Use Caution Hi, I would like to replicate an analysis I did with qdec but regressing out thickness at the vertex level. It’s a myelin group difference with age and left average myelin as nuisance factors. I used the mri_glmfit command line generated during that qdec analysis and added the pvr flag but I got that error message: ERROR: dimension mismatch between X and contrast /home/damien/bin/freesurfer6/subjects/qdec/10nov/contrasts/lh-Avg-Intercept-R1projfrac0.5.mat X has 7 cols, C has 6 cols I understood I had to change the contrast.mat and added a 0 at the end for the pvr: +1.00000 -1.00000 +0.00000 +0.00000 +0.00000 +0.00000 +0.00000 Can you please tell me if it is correct??? I would like to be sure. I see more or less the same results than before except that some clusters are a bit degraded let say, as if I was working with a lower level of smoothing. Damien

2 1

FSGD-File
by Larissa Bechtle 26 Nov '18

26 Nov '18

2 1

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Freesurfer