Freesurfer

freesurfer@nmr.mgh.harvard.edu

6 participants
30055 discussions

Re: [Freesurfer] Help with samseg
by Cacciaguerra, Laura, M.D., Ph.D. 30 Aug '24

30 Aug '24

External Email - Use Caution Hello everyone, I am trying to segment MS lesions with "samseg" (FreeSurfer version 7.3.2), but it ends immediately with the following error: "zsh: illegal hardware instruction run_samseg --input T1.nii --input T2.nii --lesion" Have tried to run the command without the option --lesion and it works, so I guess the error is somehow linked to that option? Thank you in advance for your help, All the best, Laura

1 0

Skullstripping Issues / Estimated ICV Concern
by Carly McIntyre-Wood 29 Aug '24

29 Aug '24

1 0

Longitudinal base processing error
by Maximo, Jose Omar 28 Aug '24

28 Aug '24

External Email - Use Caution Hello all, I am using the longitudinal pipeline and keep getting the following error when generating the base for one of our subjects. I have 3 timepoints I am including. I have checked the nu.mgz and orig.mgz files for both the base and each time point. Everything looks fine with the exception of the nu.mgz for timepoint 1 (which is a solid white and looks like a mask). Could this be the problem? I have re-run recon-all for that cross sectional subject, but the image remains the same. Attached the full recon-all.log from the base output. 3d normalization pass 1 of 2 error: MRInormFindControlPoints: could not find enough control points error: MRInormFindControlPoints failed @#@FSTIME 2024:08:16:18:05:30 mri_normalize N 10 e 10.45 S 0.04 U 10.40 P 99% M 434992 F 0 R 10859 W 0 c 4 w 11 I 0 O 0 L 1.14 0.90 0.57 @#@FSLOADPOST 2024:08:16:18:05:40 mri_normalize N 10 1.19 0.92 0.58 Linux c0235 3.10.0-1160.24.1.el7.x86_64 #1 SMP Thu Mar 25 21:21:56 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux Best, Jose O. Maximo, Ph.D. | Scientist II Department of Psychiatry and Behavioral Neurobiology Heersink School of Medicine UAB | The University of Alabama at Birmingham Sparks Center 1059 | 1720 7th Ave S | Birmingham, AL 35233 P: 205.934.4245 | jomaximo(a)uabmc.edu<mailto:jomaximo@uabmc.edu><mailto:jomaximo@uabmc.edu> <mailto:jomaximo@uabmc.edu> <jomaximo(a)uabmc.edu>

1 0

Question on hippocampus segmentation algorithm
by Savoia, Alice (Stud. FPN) 27 Aug '24

27 Aug '24

External Email - Use Caution Hi everyone! I have segmented the hippocampus subfields using the segmentation tool provided by the latest version of FreeSurfer. I am now looking at the pictures and I have a few questions: * In some cases, the hp tail takes a bit of the thalamus, does the size of the tail or its borders affect the subfields. This to know if we should exclude the participant or if we can just exclude the tail from t he analysis * How the algorithm works? I also have provided a picture as an example. Thank you for the help.

2 2

FSGD Error
by Recht, Grace Owen 26 Aug '24

26 Aug '24

External Email - Use Caution Hello there, I am attempting to run the runGLMs.sh script and I am receiving the error message: ERROR: gdfRead: reading <path to FSGD file> is not formatted properly. The first string is ‘GroupDescriptorFile’, should be ‘GroupDescriptorFile’ It looks like the .fsgd file is correct – I can attach if necessary! I have searched the archive and it looks like one other person posted about it a while back with no conclusion that I can see. I am using FreeSurfer 7.4.1 using Linux i42.red.uits.iu.edu 4.18.0-553.8.1.el8_10.x86_64 #1 SMP Fri Jun 14 03:19:37 EDT 2024 x86_64 x86_64 x86_64 GNU/Linux Thank you for your time and help! Sincerely, Grace Grace Recht, MA PhD Student Clinical Neurotrauma Lab Department of Kinesiology | School of Public Health Indiana University – Bloomington gorecht(a)iu.edu

1 0

petsurfer questions - No PVC stats file in the similar format as gtm.stats.dat
by Sneha Pandya 26 Aug '24

26 Aug '24

External Email - Use Caution Dear team, A] I am running mri_gtmpvc on our data and need to extract stats data that is in the similar format as gtm.stats.dat file from non-pvc corrected PET data, in the sense getting mean regional uptake from non-pvc PET data alongside region names, volumes, var, etc information using ROIs from gtmseg. mri_gtmpvc command (turned off PVC): mri_gtmpvc --i PET.nii.gz --reg PET_to_nu.reg.lta --seg ${ FreeSurferDir }/mri/gtmseg.mgz --default-seg-merge --auto-mask ${FWHM} .01 --psf 0 --no-tfe --no-rescale --save-input --o nopvcoutput I tried this with following approaches as suggested in the forum: 1] mri_convert nopvc.nii.gz --ascii nopvc_1.stats.dat 2] mri_segstats --i input.nii.gz --seg aux/seg.nii.gz --ctab aux/seg.ctab --excludeid 0 --sum nopvc_2.stats.dat On comparing mean uptake between nopvc_1.stats.dat and nopvc_2.stats.dat, I see that values are similar but not same. Is this expected? Head of nopvc_1.stats.dat: 54832.960938 65543.554688 59080.023438 75662.140625 47840.519531 61120.222656 66590.773438 69356.742188 58847.007812 65715.578125 Head of nopvc_2.stats.dat: Index SegId NVoxels Volume_mm3 StructName Mean StdDev Min Max Range 1 2 106905 223886.7 Left-Cerebral-White-Matter 54832.95 8839.049 15160.92 96317.77 81156.85 2 7 7723 16174 Left-Cerebellum-White-Matter 65543.57 10289.78 26951.39 90987.36 64035.97 3 8 24806 51950.2 Left-Cerebellum-Cortex 59079.91 14349.17 13388.75 107889.7 94500.95 4 10 3427 7177 Left-Thalamus 75662.06 13054.39 33599.36 105854.3 72254.89 5 11 1813 3796.9 Left-Caudate 47840.52 10624.56 19515.38 88224.69 68709.31 6 12 2280 4774.9 Left-Putamen 61120.21 8453.317 43591.13 83701.97 40110.84 7 13 976 2044 Left-Pallidum 66590.77 5816.431 50056.58 83431.62 33375.04 8 16 6686 14002.2 Brain-Stem 69356.77 16951.96 20199.62 115301.7 95102.11 B] Also, when we turn off PVC with "--psf 0 --no-tfe" flags, SGTM is still performed right whose regional data is saved in gtm.stats.dat file? If we perform RBV PVC does it in addition affect in anyway SGTM outputs (i.e gtm.stats.dat)? C] Is there a way to erode only reference region in the mri_gtmpvc command both when PVC is performed (SGTM or RBV) and with no PVC? If not, is there a way to do so after running mri_gtmpvc to rescale PVC'ed and non-PVC'ed uptake with eroded cerebellum derived from gtmseg? I have extracted eroded GM cerebellum from gtmseg using mri_binarize and then planning to use that eroded GM cerebellum mask to rescale PVC'ed/non-PVC'ed regional uptake, not sure if it is the right approach. mri_binarize command: mri_binarize --i ${FreeSurferDir}/mri/gtmseg.mgz --o CerebellumGrayMatterMaskEroded_gtmseg.mgz --match 8 47 --erode 2 Thank you, Sneha

2 2

Combining masks mri_binarize
by Sorby-Adams, Annabel 26 Aug '24

26 Aug '24

Hi Freesurfers! I am trying to merge a L and R hippocampal mask derived from fsl FIRST. I have binarized the masks (using the --match flag) but when I try to merge them (mri_binarize --i <L hippocampal mask> --merge <R hippocampal mask> --min --max --o <merged mask>) but the output merged mask is blank. Should I be specifying the min/max differently or is there another way to merge the two masks more accurately? Cheers, Annabel

1 0

Extracting cortical thickness from a sphere with different coordinates for different participants.
by Michel, Léa 26 Aug '24

26 Aug '24

External Email - Use Caution Hi everyone, I'm working on a dataset where T1 scans were taken before the patients went under surgery. During the procedure, a small, non-pathological brain tissue sample was resected from the temporal cortex. The sample, extracted 4cm from the anterior pole within the middle temporal gyrus (Brodmann area 21), measured 1-1.5 cm in diameter. My goal is to extract the MRI-derived cortical thickness from that tissue on the T1 scan. My first step was creating a sphere as my ROI with the coordinates that I decided by participants (through looking at fsleyes). The problem now is that I'm unsure on how to extract the cortical thickness from that ROI for each participant. Do I need to register my ROI into the inflated template? If you have any ideas or codes to solve that issue, please let me know. Here the script that I used to get the sphere: # choose radius radius=20 # in mm # replace with image of your subject subj_image='/Project_2.5/MRI/Subject_103/Scan103.nii.gz' # NB, these are not MNI coordinates, but coordinates in the so called voxel space # change to the coordinates you are interested in coordinate_x=70 coordinate_y=53 coordinate_z=32 ## do not change anything below fslmaths ${subj_image} -mul 0 -add 1 -roi ${coordinate_x} 1 ${coordinate_y} 1 ${coordinate_z} 1 0 1 _tmp_point -odt float fslmaths _tmp_point -kernel sphere ${radius} -fmean spherical_roi_r${radius} -odt float rm -f _tmp_point.nii.gz fsleyes ${subj_image} spherical_roi_r${radius}.nii.gz & Thank you for your help, Lea __________________________________________________________ Léa C. Michel PhD Student Lifespan Cognitive Dynamics Lab - www.rogierkievit.com<http://www.rogierkievit.com/> Affiliation: Radboud university medical center, Donders Institute for Brain, Cognition and Behaviour, Cognitive Neuroscience Department, Nijmegen, The Netherlands Email: Lea.Michel(a)radboudumc.nl Website: https://leacmichel.github.io/ Twitter: @LeaCMichel<https://twitter.com/LeaCMichel> De informatie in dit bericht is uitsluitend bestemd voor de geadresseerde. Aan dit bericht en de bijlagen kunnen geen rechten worden ontleend. Heeft u deze e-mail onbedoeld ontvangen? Dan verzoeken wij u het te vernietigen en de afzender te informeren. Openbaar maken, kopiëren en verspreiden van deze e-mail of informatie uit deze e-mail is alleen toegestaan met voorafgaande schriftelijke toestemming van de afzender. Het Radboudumc staat geregistreerd bij de Kamer van Koophandel in het handelsregister onder nummer 80262783. The content of this message is intended solely for the addressee. No rights can be derived from this message or its attachments. If you are not the intended recipient, we kindly request you to delete the message and inform the sender. It is strictly prohibited to disclose, copy or distribute this email or the information inside it, without a written consent from the sender. Radboud university medical center is registered with the Dutch Chamber of Commerce trade register with number 80262783.

1 0

Freesurfer Longitudinal analysis
by De Giorgi, Sara 23 Aug '24

23 Aug '24

Hi all, Could it be a problem in FreeSurfer longitudinal analysis if I use two different sequences (T1 MP2RAGE and T1 MPRAGE) for different time points/follow-ups? Thanks for the support, Sara

1 0

Longitudinal base processing error
by Nelson, Eric (Campus) 23 Aug '24

23 Aug '24

External Email - Use Caution Hello all, I am using the longitudinal pipeline and keep getting the following error when generating the base for one of our subjects. I have 3 timepoints I am including. I have checked the nu.mgz and orig.mgz files for both the base and each time point. Everything looks fine with the exception of the nu.mgz for timepoint 1 (which is a solid white and looks like a mask). Could this be the problem? I have re-run recon-all for that cross sectional subject, but the image remains the same. Attached the full recon-all.log from the base output. 3d normalization pass 1 of 2 error: MRInormFindControlPoints: could not find enough control points error: MRInormFindControlPoints failed @#@FSTIME 2024:08:16:18:05:30 mri_normalize N 10 e 10.45 S 0.04 U 10.40 P 99% M 434992 F 0 R 10859 W 0 c 4 w 11 I 0 O 0 L 1.14 0.90 0.57 @#@FSLOADPOST 2024:08:16:18:05:40 mri_normalize N 10 1.19 0.92 0.58 Linux c0235 3.10.0-1160.24.1.el7.x86_64 #1 SMP Thu Mar 25 21:21:56 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux Best, Eric Nelson, Ph.D. | Postdoctoral Fellow Department of Psychiatry and Behavioral Neurobiology School of Medicine UAB | The University of Alabama at Birmingham

2 2

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