Freesurfer

freesurfer@nmr.mgh.harvard.edu

11 participants
30080 discussions

Trouble making ROI labels from .mgz file that has been converted from .nii.gz
by JD Knotts 27 Mar '19

27 Mar '19

External Email - Use Caution Hello FreeSurfer Developers, I am having difficulty trying to convert mrVista-defined ROIs in matlab into FreeSurfer label files. I'm starting with an atlas.mgz file (containing subject-specific visual ROIs based on the Kastner FreeSurfer atlas), and I should note that I can successfully make ROI .label files in FreeSurfer from this .mgz file using successive calls to mri_vol2surf and mri_cor2label. For reference, when I do this for the first Kastner ROI, the resulting label correctly has 1591 label voxels. However, part of my pipeline involves intersecting these ROIs with a localizer ROI that was defined in mrVista. To do this, I first convert the atlas.mgz file to .nii.gz using mri_convert as follows: mri_convert lh.atlas.mgz lh.atlas.nii.gz --reslice_like subject_t1.nii.gz --resample_type nearest --out_data_type int --no_scale 1 After intersecting with the localizer in matlab, when I save a new .nii.gz file, convert back to .mgz, and run mri_vol2surf and mri_cor2label, I get the following error in mri_cor2label: ERROR: found no voxels matching id 1. To try to keep this short... as a sanity check I tried simply converting the lh.atlas.nii.gz file that pops out of the mri_convert step above directly back into .mgz (without modifying the .nii.gz file at all): mri_convert lh.atlas.nii.gz lh.atlas_reconverted.mgz --reslice_like subject_t1.nii.gz --resample_type nearest --out_data_type int --no_scale 1 When I run mri_vol2surf and mri_cor2label on the resulting lh.atlas_reconverted.mgz file, I don't get an error, but the resulting label file for the first Kastner ROI incorrectly finds 111 label voxels, none of which overlap with the correct 1591 label voxels above. So it appears that I'm making at least one mistake in the mri_convert step from nii.gz to .mgz above, but I'm struggling to figure out what it is. Any guidance on this would be greatly appreciated! Thank you for your time, JD Knotts 1) FreeSurfer version: freesurfer-Darwin-lion-stable-pub-v5.3.0 2) Platform: Mac OSX 10.11.6 3) Darwin Jeffreys-MacBook-Pro.local 15.6.0 Darwin Kernel Version 15.6.0: Thu Jun 23 18:25:34 PDT 2016; root:xnu-3248.60.10~1/RELEASE_X86_64 x86_64 4) recon-all.log: see attached for example subject

1 0

Conversion of thalamic nuclei labels
by Valentina Mancini 27 Mar '19

27 Mar '19

External Email - Use Caution Dear all, I ran the thalamic segmentation implemented in FreeSurfer and, according to visual inspection, I got nice results with a good delineation of all the nuclei. Then, I wanted to use the label of the thalamic nuclei to select an ROI in the MGN and the MD to perform functional connectivity with cortical regions in SPM. Therefore, I converted the outputs of the segmentation (ThalamicNuclei.v10.T1.mgz and ThalamicNuclei.v10.FSvoxelSpace.mgz) into nifti files by using the command "mri_convert .mgz .nii". However, the resulting labels of the thalamus have a different resolution and I cannot visualize anymore the different thalamic subnuclei but only the entire thalamus. Do you have any suggestion to convert thalamic labels into nifti without losing the different subnuclei? Do you think that the thalamic segmentation is accurate enough to be used to perform functional connectivity? Thank you and best regards, Valentina Mancini

4 9

Re: [Freesurfer] PetSurfer
by Shane Schofield 27 Mar '19

27 Mar '19

External Email - Use Caution Hi Doug, thanks for the explanation. My question is more related to how we might go about knowing what are the high binding regions for any tracer? If I am going to do an FDG-PET study, will the HB region parameters be different from a [11C]-PK11195 study for example? And secondly, might be these be different across subjects? Thanks for your help. Previous email :You should read up on the MRTM2. Basically, when you run MRTM1, it computes a separate k2 for each voxel eventhough the k2 should be the same across the entire brain (since it actually refers to the reference region). With MRTM2, you estimate the single k2 value from high binding regions.

2 1

Thalamic nuclei segmentation
by 郑凤莲 27 Mar '19

27 Mar '19

External Email - Use Caution Hi experts, I have segmented the thalamic nuclei into 25 nuclei using the new FS version. However, when I input "freeview -v nu.mgz -v ThalamicNuclei.v10.T1.mgz:colormap=lut ", there is no 3D atlas in Freeview, just nu.mgz. What's the matter? Please help me. I am looking forward to your reply. Thanks very much. Sincerely, Zheng

2 1

Re: [Freesurfer] Longitudinal Two Stage Model Question
by Arsenije Subotic 26 Mar '19

26 Mar '19

External Email - Use Caution Hi Doug, Thanks for getting back to me so quickly. Just to confirm, I can proceed to use the mri_glmfit analysis as shown on the group level tutorial? I'm not sure I can do the paired analysis since my subjects in one group are not matched with those from the other and one group has more participants than the other. I will however use the the link you gave me to compute the difference within one of my groups. Best, Arsenije --- Arsenije Subotic | MSc Student Department of Clinical Neurosciences Hotchkiss Brain Institute, University of Calgary | T2N 4N1, Canada Tel: (403) 918-6970 arsenije.subotic1(a)ucalgary.ca

1 0

Left and right pial surfaces crossing in medial cortex
by Matthieu VANHOUTTE 26 Mar '19

26 Mar '19

External Email - Use Caution Dear FS' experts, Using both T1 and FLAIR in recon-all to improve pial surfaces, I often have the case where pial surfaces from the left and right hemisphere are crossing in medial cortex (see attached images). Is it anormal and bad ? If so, How could I fix it ? Best, Matthieu

2 1

Re: [Freesurfer] Longitudinal Two Stage Model Question
by Arsenije Subotic 26 Mar '19

26 Mar '19

External Email - Use Caution Sorry, I was actually able to compile a FSGD file that has a single line for each subject (containing the fsid base) along with gender and age at baseline. I ran this through the mris_preproc command (https://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/GroupAnalysis) in order to stack the spc files which were previously cached. I guess my question really is can I now simply continue on and use the mri_glmfit instructions on the group analysis tutorial page or is there a special procedure because I am dealing with longitudinal data? I was planning on simply using contrasts i had designed before for my cross-sectional analysis, controlling for age at baseline and gender. Thanks and my apologies for all the questions! Arsenije --- Arsenije Subotic | MSc Student Department of Clinical Neurosciences Hotchkiss Brain Institute, University of Calgary | T2N 4N1, Canada Tel: (403) 918-6970 arsenije.subotic1(a)ucalgary.ca

2 1

PetSurfer
by Shane Schofield 26 Mar '19

26 Mar '19

External Email - Use Caution Hello PetSurfer Developers On the tutorial page, I do not understand this part: --km-hb 11 12 13 50 51 52 specifies the ROIs to use as the high-binding region if using MRTM2. This creates km.hb.tac.nii.gz with the value for the high-binding region for each time point. What exactly is a high-binding region and how do these inputs affect the rest of the estimation? If I do not know what the high binding regions for a tracer are, may I leave this blank? Thank you.

2 1

How to improve recon all on subjets with hight atrophy level like Alzheymer patients?
by Barbara Micaela Duarte Abritta 26 Mar '19

26 Mar '19

External Email - Use Caution Hello everyone, recently I have started to analyze a new sample of subjets which is made up of people with Alzheimer's, due to this pathology several of my subjects have a high level of atrophy in subcortical and temporal areas of the brain, which causes several errors in the pial surface as a wm surface. So to make both surfaces more accurate I tried to put some control points but in many places this solution was not enough due the level of inaccuracy. Is there any solution for this problem? I will appreciate if you could offer any suggestions to solve this problem. Regards Bárbara

2 1

mkanalysis-sess with a customized HRF
by Francesca Strappini 26 Mar '19

26 Mar '19

External Email - Use Caution Dear Freesurfer's experts, I would like to use a customized HRF with mkanalysis-sess. Is there a way to do it? Thanks! Best Francesca -- Francesca Strappini, Ph.D. Neurobiology Department Weizmann Institute of Science 234 Herzl Street, Rehovot 7610001 Israel Tel.: +972 58 444 2584 E-mail: francesca.strappini(a)weizmann.ac.il

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