Freesurfer

freesurfer@nmr.mgh.harvard.edu

2 participants
30134 discussions

Support for fsaverage surface maps in NeuroVault
by Chris Filo Gorgolewski 11 Nov '16

11 Nov '16

Dear all, I'm very happy to announce a preliminary support for uploading, visualizing and sharing surface maps in fsaverage space in NeuroVault.org. The following features are currently supported: 1. Uploading .mgh, .curv, and .gii scalar maps: 2. Visualization of the surface map projected into the MNI volume (using MNI template surface reconstruction instead of fsaverage) 3. Downloading of the maps in GIFTI format (with metadata correction making viewing of the files in both Freeview and Connectome Workbench easy). Example map: http://neurovault.org/images/31996/ Thanks to the projection to the MNI volume space all of the extra analysis features of NeuroVault (cognitive decoding using neurosynth, gene expression decoding using Allen Brain Atlas data, searching for maps with similar activation patterns) will work seamlessly. TODO (aka help needed): 1. Improve the surface reconstruction of the MNI atlas (the cortex is too thin in some areas) 2. Provide 3D visualization using brainbrowser 3. Add Connectome workbench scene files and include fsaverage GIFTI surfaces in download packages. 4. Make volume projection asynchronous to improve response time (it takes several seconds to process one map right now leaving the user with no feedback). Please let me know what do you think! Don't forget to share your maps! Best, Chris

1 0

Re: [Freesurfer] Extracting structural measures from trend-level PET clusters
by Elijah Mak 11 Nov '16

11 Nov '16

Thanks, Doug. It works but I am not sure if the values I'm getting are correct. Is this supposed to be reporting the mean cortical thickness of each cluster? I have pasted the output below from the avgwf below. The names of the clusters correspond to the regions from the "summary" text, am I right? Group lh_fusiform lh_precentral lh_insular lh_precuneus lh_parsorbitalis 1 24.29454 29.53935 14.24736 28.97924 31.54577 1 25.68872 28.26156 17.47828 28.20604 30.38902 2 25.46281 30.75879 19.34943 29.59019 32.0404 2 25.55514 27.23871 17.27937 29.14274 29.75657 Another related question: Is it possible to simply take the mean cortical thickness across the voxels exceeding p<0.001 from the uncorrected sig.mgh? Thanks again! Best Wishes, Elijah

1 0

recon all
by John Anderson 11 Nov '16

11 Nov '16

3 7

How to view overlapping voxels across multiple sig.mgh files?
by Elijah Mak 11 Nov '16

11 Nov '16

Hi Freesurfer Team, I am just wondering if there is a neat way to derive .mgh files containing cluster information about spatially overlapping voxels across various sig.mgh? Thank you. Best Wishes, Elijah

1 0

mri_info --tkr2scanner result for run with multiple input scans
by Krieger, Donald N. 11 Nov '16

11 Nov '16

If I run recon-all with both a T1 and a T2 scan and switch t2pial, does mri_info -tkr2scanner give me the transform to the T1 scanner frame of reference? And if I run recon-all with many T1's , what does mri_info -tkr2scanner give me then? Thanks - Don

2 5

Mean intensity with mri_segstats : different results with .annot or .label
by Matthieu Vanhoutte 11 Nov '16

11 Nov '16

Dear experts, I am in trouble with two ways of computing mean intensity with mri_segstats. First I have used on .annot files with three different labels inside (SegId 1 to 3) : *mri_segstats --annot fsaverage lh cache.th23.pos.sig.ocn.annot --i lh.PET.fsaverage.sm10.mgh --sum lh.pet.sum* which give me those results: # ColHeaders Index SegId NVertices Area_mm2 StructName Mean StdDev Min Max Range 1 1 27805 14182.6 cluster-001 8.8194 1.4351 6.2105 14.2947 8.0842 2 2 2321 1047.3 cluster-002 10.3661 1.0848 8.8375 12.9317 4.0942 3 3 552 313.1 cluster3 10.3365 0.4711 9.3719 11.8694 2.4975 Second I have created on .label files containing the three segmented labels (SegId 1 to 3) and used this label with mri_segstats: *mri_segstats --slabel fsaverage lh lh.th23.cluster.thresh1.3.label --i lh.PET.fsaverage.sm10.mgh --excludeid 0 --sum lh.pet.label.sum * which give me those results: *# ColHeaders Index SegId NVertices Area_mm2 StructName Mean StdDev Min Max Range * * 1 1 30678 15542.9 Seg0001 8.9637 1.4701 6.2105 14.2947 8.0842 * I verified if I obtained the same results with the two ways (.annot or .label) and everything is fine comparing NVertices, Area_mm2 but Mean in first way gives me (8.8194+10.3661+10.3365)/3 = 9.84066 and with .label file Mean = 8.9637. Did I have miss anything and how is computed mean intensity ? Many thanks in advance. Best regards, Matthieu

2 8

CVS repository
by Melvin Robinson 11 Nov '16

11 Nov '16

I am trying to checkout the development version of Freesurfer from the CVS repository. I need to build against VTK 7.1 instead of 5.6 for freeview and am using these instructions: https://surfer.nmr.mgh.harvard.edu/fswiki/ReadOnlyCVS Using this command: cvs -d :pserver:anonymous@fsvm.nmr.mgh.harvard.edu:/usr/fscvsroot checkout -P dev Results in the following error: cvs server: cannot find module `dev' - ignored Can you tell me the correct module name?

2 1

AAL atlas
by stella maris sanchez 11 Nov '16

11 Nov '16

Hi all! I'm very new using FS and I have some questions. I have been doing parcellations with the FreeSurferLUT on a few subjects, with the comand recon-all and it worked OK. Now I'd want to try with the AAL atlas, I mean, I want my subject's brain parcellated in the AAL atlas. I have the template corresponding to AAL parcellation in nifti format, so how can I obtain this parcellation in my subject space? what (and how) comands I have to use? Thanks in advanced! Cheers.- Stella

2 3

FSL View in the virtual box version?
by Jeffrey Crawford 11 Nov '16

11 Nov '16

Hello there! I currently use the virtualbox version of freesurfer on my windows and the CentOS 4 (32b) version on a linux computer. I noticed today that the Virtual box version does not seem to have fsl view. I found this link to a blog where this person has a virtual box that contained fslview, but I think his methods and version are different from the typical installation procedures. http://kylemathewson.com/cnl-virtualbox-setup/ I was just wondering if it is possible to use fslview on the virtual box version or not and how I might go about obtaining that capability?

2 1

Large Ventricles
by Tamara Tavares 11 Nov '16

11 Nov '16

Hello, I have a participant who has very large ventricles and as a result, there are substantial errors in the aseg output. Looking at the mailing list, I see that the newest version of Freesurfer is suppose to be better at handling large ventricles; however, I want to be consistent with the Freesurfer version I have been using (version 5.1). Are there any other recommendations with dealing with large ventricles? Thank you in advance for your help, Tamara

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