Freesurfer

freesurfer@nmr.mgh.harvard.edu

5 participants
30076 discussions

License on fsaverage datasets
by Marmaduke Woodman 10 Nov '16

10 Nov '16

hi As part of developing some tools [1] partly on the FreeSurfer ecosystem, we'd like to use the fsaverage5 dataset as a test data. In our workflow that would mean committing those files to our Git repo, pushing to GitHub, testing on Travis CI etc which makes them public in our repo. We have, however, no intention of otherwise redistributing the data. If we place the FreeSurfer license file next to the fsaverage5 folder in our repo, is this OK? or would you have a strong preference against this? Thanks, Marmaduke [1] The project in question is here https://github.com/ins-amu/bnm-recon-tools though not well developed at its current stage.

2 1

FreeSurfer Dev Releases
by Elia S. 10 Nov '16

10 Nov '16

Dear FreeSurfer Developers, We have been trying out, for the past few months, the freesurfer dev releases, and in particular this one: freesurfer-X86_64-unknown-linux-gnu-dev5-20150924 My question is, does the above version classify as a nightly version of FreeSurfer 5, or as an early beta of the 6? Or......do the dev versions fall in some kind of grey area that are neither 5 or 6? Thank you in advance, Elia.

2 2

Manual pial and white matter corrections cause irregularities
by Cleo Vandervost 10 Nov '16

10 Nov '16

Hi Bruce, As you asked, I uploaded the recon-all log-file for that subject, edited and unedited. We're using version 5.3. How can you check for sure that the WM significantly gets >110? I also uploaded the pictures once again. Thanks, Cleo ________________________________________ Van: freesurfer-bounces(a)nmr.mgh.harvard.edu <freesurfer-bounces(a)nmr.mgh.harvard.edu> namens Bruce Fischl <fischl(a)nmr.mgh.harvard.edu> Verzonden: vrijdag 21 oktober 2016 02:14 Aan: Freesurfer support list Onderwerp: Re: [Freesurfer] Manual pial and white matter corrections cause irregularities Hi Cleo can you upload a subject before and after editing so we can replicate what you are seeing? And what version of FreeSurfer are you using? Does the WM after editing get signifcantly >110? If so, you may have accidentally placed a control point in a gray matter voxel cheers Bruce On Thu, 20 Oct 2016, Cleo Vandervost wrote: > > Hello Freesurfer team, > > > For my thesis I'm researching the differences in grey and white matter and cortical thickness between > children with and without dyslexia. My supervisor advised me to manually edit the Freesurfer-generated > data. I've been editing my data as follows: > > - adding control points to (larger) WM-areas that FS did not include > > - manually editing (smaller) WM-areas through recon-editing > > - correcting the pial surface through recon-editing > > > After finishing and saving my edits I reran recon-all with this command: recon-all -autorecon2-cp > -autorecon -subjid (subject number). When I opened my data again in Freesurfer to check my edits, I saw > that there were two big problems: > > - 'distortions' in the pial surface > > - sometimes large areas of WM suddenly were excluded, even when I didn't make any edits in that area > > It makes the edited data look much worse than the original data, which is weird I suppose. > > > I've included some pictures so you can see what I mean. > > - Picture (a) is the original unedited version. > > - Picture (b) is the edited version: control points and WM edits. You can see that there's a large area > in the left hemisfere excluded from the original WM area > > - Picture (c) is the re-edited version of (b), control points added in that large area. This time it's > included but the pial surface shows a lot of irregularities. > > > Me and my supervisor tried a couple of times with other data, but we still had the same results. We also > tried by only adding control points and editing manually, without the pial-corrections. > > > Do you have any idea of what's going wrong in the editing proces? > > > Thank you very much, > > > Kind regards > > > Cleo Vandervost > > 1st master speech and language therapy > > Katholieke Universiteit Leuven > > > > >

2 3

Extracting structural measures from trend-level PET clusters
by Elijah Mak 09 Nov '16

09 Nov '16

Hi Freesurfer Team, 1) We've found some subtle differences in tau accumulation between two groups. Unfortunately, it does not survive mri_glmfit-sim (voxel threshold 0.001, cwp < 0.05). Still, I would like to extract the structural measures from the uncorrected tau clusters (p<0.001). What is the most convenient approach for this? Could the following URL (http://www.mail-archive.com/f reesurfer(a)nmr.mgh.harvard.edu/msg16268.html ) be a solution? Can mri_label2label work for sig.mgh with multiple clusters? 2) A slightly unrelated question: how can I go about using this particular cluster as a seed for functional or tractography analysis in other softwares (i.e. FSL or MRTRIX) or even within Freesurfer if possible? Thanks again. Really appreciate your help! Best Wishes, Elijah

2 1

Re: [Freesurfer] Assessing local correlations across 2 modalities
by Elijah Mak 09 Nov '16

09 Nov '16

Hello, Thanks Doug for your help on this previously. I am quite puzzled by the vertex-wise local correlations that I've found (i.e more tau thicker cortex)... and before I go down the rabbit hole of trying to dissect what could be happening, I'd like to check if one of the steps in my workflow has gone wrong? 1) My FSGD is a simple one group design with no other covariates. 2) mri_concat --f stacks_cth_rh15 --o concat_cth_rh15_.mgh **stack of thickness maps after qcache 3) C_pvr_cth.mtx = 0 1 4) mri_glmfit --y concat_tau_lh_sm8.mgh --pvr concat_cth_lh15.mgh --fsgd FSGD_tau dods --C C_pvr_cth.mtx --surf fsaverage lh --cortex --glmdir pvr_lh.glmdir Thanks a lot! :) Best Wishes, Elijah Writing results C_pvr_cth maxvox sig=4.6783 F=34.7936 at index 31050 0 0 seed=1479612145 mri_glmfit done Thanks a lot for your help. Best Wishes, Elijah

2 1

mri_convert slope scale
by Rockers, Elijah D. 09 Nov '16

09 Nov '16

Hello, We have been using mri_convert to convert our dicoms into .nii niftis. There is an entry in the .nii header for the slope scale, but mri_convert appears to discard this value instead of copying it into the header (whereas mricron retains this value) We prefer to use mri_convert. Is there a way to have mri_convert carry this value over? Why is it being discarded? Thanks, Eli Houston Methodist. Leading Medicine. U.S. News & World Report has recognized Houston Methodist as one of the nation’s top 20 hospitals, placing it for the second time on the magazine’s prestigious Honor Roll. It is also designated as a Magnet hospital for excellence in nursing. Visit us at houstonmethodist.org. Follow us at twitter.com/MethodistHosp and www.facebook.com/HoustonMethodist. ***CONFIDENTIALITY NOTICE*** This e-mail is the property of Houston Methodist Hospital and/or its relevant affiliates and may contain restricted and privileged material for the sole use of the intended recipient(s). Any review, use, distribution or disclosure by others is strictly prohibited. If you are not the intended recipient (or authorized to receive for the recipient), please contact the sender and delete all copies of the message. Thank you.

2 1

GM = SurfArea x Thick
by Daya P 09 Nov '16

09 Nov '16

Hi all, I grey matter volume supposed to be equal to surface area x average thickness in each ROI when looking at mris_anatomical_stats for individual subjects? Thank you, Daiana

2 1

GM ROI list?
by P Taylor 09 Nov '16

09 Nov '16

Is there an available list of GM regions within the list of regions in the FreeSurferColorLUT.txt? (In particular, I am interested in a list of GM regions output in a default run of recon-all, which seems to have a maximum regionnumber of 2035. But a full list is great, too-- I'm not sure which might be already available or easier to generate.) I had thought of using the segmentation volumes to help me differentiate ROIs. For example, I could get rid of regions that overlap with the WM volume, but I think that the segmentation might come from a slightly different bit of information than the sum of segmented regions. As shown in the attached slice image, if I underlay the WM segmentation volume (color = white), and overlay a "recon-all" parcellation+segmentation map (color = red for GM on right, yellow for GM on left, and green for non-GM) to look for overlap, I see that some of the WM seg volume overlaps with GM ROIs (overlaps are isolated pink-ish voxels and light yellow voxels; one highlighted in cross hairs). I guess this difference occurs because the segmentation volume comes from the surfaces themselves? Anyways, this makes me think that I can't automate finding the GM, and I would rather go from LUT numbers, and hence the above question. Thanks, pt

3 16

latest version of mri_glmfit-sim?
by Elijah Mak 09 Nov '16

09 Nov '16

H <ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_glmfit-si…>ello Freesurfer Team, Where may I download the latest version of mri_glmfit-sim and QDEC to run the mri_surfcluster? I am having errors on mri_glmfit-sim and monte-carlo simulations in QDEC (dev version that was released quite awhile back). A snippet of the error below after MC correction in QDEC Searching for Clusters ... thmin=2.000000 (2.000000), thmax=-1.000000 (-1), thsignid=0, minarea=0.000000 Found 36 clusters Max cluster size 2817.874756 INFO: fixing MNI talairach coordinates Pruning by CW P-Value 0.05 ERROR! Error in Monte Carlo simulation: Error running mri_surfcluster! Many thanks for your help. Best Wishes, Elijah

2 1

mri_convert scaling
by Funk, Quentin 09 Nov '16

09 Nov '16

All, I have some DICOM files from a pet scan and an MRI. I viewed the files in three different ways: 1. Opened the DICOM's directly in a viewer (Pmod) 2. Used mri_convert to change the DICOMs to a .nii, and opened them. 3. Used dcm2nii to change the DICOMS to a .nii and opened them. It's my intuition that the raw intensity values are displayed when I directly open the DICOM images. However, both step (2) and (3) give images with intensity values scaled--(2) scales by different factors depending on the modality, anywhere from 16 to 50; (3) scales by a fixed amount (two orders of magnitude, possibly understandable). The images are indistinguishable when the levels are set properly, so this scaling would appear to be a flat linear multiplier. I'd like to know where it comes from--is there some information in the DICOM headers that mri_convert is using to scale things? I checked the archives of this mailing list and found the tag -ns 1, but this didn't change the output file at all. Any help is appreciated. -- Quentin Funk, PhD Houston Methodist Research Institute 713-363-9003 Houston Methodist. Leading Medicine. U.S. News & World Report has recognized Houston Methodist as one of the nation’s top 20 hospitals, placing it for the second time on the magazine’s prestigious Honor Roll. It is also designated as a Magnet hospital for excellence in nursing. Visit us at houstonmethodist.org. Follow us at twitter.com/MethodistHosp and facebook.com/HoustonMethodist. ***CONFIDENTIALITY NOTICE*** This e-mail is the property of Houston Methodist and/or its relevant affiliates and may contain restricted and privileged material for the sole use of the intended recipient(s). Any review, use, distribution or disclosure by others is strictly prohibited. If you are not the intended recipient (or authorized to receive for the recipient), please contact the sender and delete all copies of the message. Thank you.

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