Freesurfer August 2016

freesurfer@nmr.mgh.harvard.edu

127 participants
198 discussions

TRACULA: narrow probability distribution
by Newbold, Dillan 23 Aug '16

23 Aug '16

Dear Anastasia, I’ve been looking at a lot of Tracula path.pd files and I’ve found that some probability distributions are only a single voxel wide, similar to the path.map file. The few none-zero voxels in these path.pd files have very high probability values. When an isosurface is generated for these tracts, it looks like a short thin blob somewhere in the usual tract distribution. I’ve seen descriptions in the archives of similar “short thin tracts,” but, from what I have seen, no one has offered a satisfying explanation for why these occur. What I think is happening in these tracts is that a maximum-probability (or local maximum) path is found during a burn-in iteration and all following perturbations of that path are rejected. Since the probability value in the path.pd is equal to the number of sample paths intersecting that voxel, finding a local maximum early on results in a small number of very high-probability voxels. Consistent with this explanation, I’ve found that this issue occurs more frequently when nburnin is set to 1000 (default = 200). A similar issue can occur if a local maximum is found early during the sample iterations, and this results in a path.pd file containing a small number of voxels with very high values surrounded by a larger area of low-value voxels. When a 20% threshold is applied, the result is the same as when a local maximum occurs during a burn-in iteration. Does my understanding of this issue seem correct? None of this would be a problem if my only aim were to find the single path with the maximum a posteriori probability, but I’m concerned that the average and weighted_average sats for these tracts will be less accurate. Since these distributions include small fractions of the number of voxels included in most tract distributions, is it likely that the average and weighted_average stats from these narrow distributions are less representative of the whole tract and more subject to random noise? Given these concerns, what type of overall path statistics do you think is most descriptive of a tract? Also, do you feel that higher nburnin and nsample values should lead to superior results? I would have thought this to be the case, but now it seems to me that setting either of these values too high will result in narrow probability distributions and bad statistics. Thank you, Dillan

1 0

Tracula ROI analysis CVS vs MNI reg
by Chung, Yoonho 23 Aug '16

23 Aug '16

Hi Anastasia, Is there a preferable method between CVS vs MNI registration when extracting diffusion mesures from ROIs (white matter, ASEG etc)? Best, Yoon

2 1

Re: [Freesurfer] TRACULA - difference in tracts using MNI vs CVS
by Elijah Mak 23 Aug '16

23 Aug '16

Hi Anastasia, The error message occurs for other control points too, but it does not kill the -prior stage. I have attached the full trac-all.log file. This error message is also found in other subjects where I am struggling to get a good initialisation. The ERROR: fio_pushd:subject/dmri/xfms/cvs/final_CVSmorph_tocvs_avg35/mri Thank you! Best Wishes, Elijah -- Elijah Mak PhD Candidate *|* Psychiatry University of Cambridge Trinity College, Cambridge, CB2 1TQ

2 1

permutation/monte carlo
by maaike rive 23 Aug '16

23 Aug '16

Dear Douglas, Sorry to bother you with this, but after reading the thread about the Eklund papers I'm getting confused about what I actually did when correcting for multiple comparisons and I'd also like to know whether I'm on the safe side. I used pre-cashed Monte Carlo simulation with a cluster finding threshold of 0.05 (at least so I thought): mri_glmfit -sim - -glmdir - -cache 1.3 abs - - cache-dir - - cwpvalthresh 0.01 Am I correct? And am I also correct that the FDR may be about 13% instead of 5% when doing this? (I smoothed thickness and surface area data with FWMH 10mm, but not the lGI data, because they were already very smooth and pre-cashed simulation didn't work after additional smoothing). Thanks, Maaike

2 1

Recon-all Correcting Defect 0
by Heidi Foo 23 Aug '16

23 Aug '16

Dear FreeSurfer team, I apologize for sending this email again. I am currently running recon-all for my subjects but for 2 of them, it stops at XL correcting defect. Do you have any ideas how I can deal with this problem? Attached the log file for your reference. Thanks. Regards, Heidi

2 3

qcache for qdec longitudinal analysis
by Nick Corriveau Lecavalier 23 Aug '16

23 Aug '16

Hi Freesurfer team, I am planning on following this page for my longitudinal group analysis using QDEC: http://www.freesurfer.net/fswiki/FsTutorial/QdecGroupAnalysis_freeview. However, it tells me to run -qcache for all of my subjects. Since I already ran the -long stream for all my subjects, do I need to rerun them all adding the -qcache command? Also, I have also run the long_mris_slopes on all of my subjects, which produced smoothed thickness files for all of my subjects (which are all in the ''base'' files). Does these files are okay for my QDEC analysis, or I really need to run all of my subjects all over again adding -qcache? Thank you very much, Nick Corriveau-Lecavalier, B.Sc. (Hons.) Étudiant au Ph.D. recherche/intervention, option neuropsychologie clinique Coordonateur à la recherche, Association étudiante des cycles supérieurs en psychologie de l'Université de Montréal (AÉCSPUM) Université de Montréal Centre de recherche de l'Institut universitaire de gériatrie de Montréal, bureau M7819

2 2

Re: [Freesurfer] [Martinos-faculty] Open position as MRI physicist at Copenhagen University Hospital, Denmark (fwd)
by Bruce Fischl 23 Aug '16

23 Aug '16

FYI The Neurobiology Research Unit (NRU) at Copenhagen University Hospital, Rigshospitalet has an opening for a MRI physicist position. The position entails design, development, implementation and evaluation of novel multimodal MRI acquisition protocols and image analysis techniques in relation to our many exciting brain research projects which employ Siemens 3 Tesla MRI. Also, ample opportunities will be given to conduct own research projects. This is a fully funded 2-year full time position with potential for on-going renewal. Employment will start October 1st, 2016, or as soon as possible thereafter. Salary will be according to the current agreement and hence an annual gross salary of minimum DDK 460,000 (~USD 70,000) can be expected. A special reduced tax scheme is offered for researchers recruited abroad, see https://www.workindenmark.dk/Working-in-DK/Tax. Primary work place will be Neurobiology Research Unit, Rigshospitalet. Close collaboration will take place with our national collaborators at the Department of Clinical Physiology and the Department of Diagnostic Radiology at Rigshospitalet as well as with our international collaborators. For further information and application, see: https://candidate.hr-manager.net/ApplicationInit.aspx?cid=342&ProjectId=201… _______________________________________________ Martinos-faculty mailing list Martinos-faculty(a)nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/martinos-faculty

1 0

Visualising path.mean.txt in MNI space after trac-all -stat
by Elijah Mak 23 Aug '16

23 Aug '16

Hi Anastasia and others, I am trying to visualise the *.path.mean.txt on MNI template after running trac-all -stat. I used this command: freeview -v $FSLDIR/data/standard/MNI152_T1_1mm_brain.nii.gz -w stats/*.path.mean.txt However, it looks like something has gone wrong with the labelling and position of the tracts (note the CST highlighted in green). Everything was fine and correctly labelled when I checked the merged tracks. Thanks again for your help! Best Wishes, Elijah P.S. On another note, the reinit=1 trick works very well and I was able to reconstruct the tracts in almost all cases. Thanks! -- Elijah Mak PhD Candidate *|* Psychiatry University of Cambridge Trinity College, Cambridge, CB2 1TQ

1 0

voxel values in destrieux file
by pierre deman 23 Aug '16

23 Aug '16

Hello, I am looking at the destrieux atlas generated by recon-all (I use the freesurfer dev version), the file aparc.a2009+aseg.mgz), and sometimes the value of a voxel doesn't exist in the FreesurferColorLut. I have quite a lot of values between 2036 and 2100 (2038, 2074, 2075 for example) and I don't find these values in the FreeSurferColorLut. Is that normal ? What do they correspond at ? Regards, Pierre -- DEMAN Pierre Mobile : +33 7 82 57 80 94

2 3

Support for surface data on NeuroVault.org
by Chris Filo Gorgolewski 22 Aug '16

22 Aug '16

Dear FreeSurfer community, I am one of the developers of NeuroVault.org - an open repository for vizualizing, analyzing, and sharing of statistical and parcellation maps of the human brain. So far NeuroVault was limited only to volumetric data in the MNI space since this is the most common way of reporting MRI results. Now I would like to start working on adding support for surface data beginning with outputs of FreeSurfer. Since I'm not a FreeSurfer expert I would like to ask the community a few questions: 1. In NeuroVault users are asked to upload their statistical maps in form of files. I'm planning to target outputs of mri_glmfit: preferably Z.mgh but also sig.mgh. Is this the most sensible thing to do? 2. I want to be able to tell if a given map is in fsaverage space. Is there any place in the .mgh header that specifies which space the map is in? 3. fsaverage has different subsampling flavours - how often are they used? Is support for them an important feature? 4. Results come in two files - one per each hemisphere. Is there any way (from the headers) to check if two files provided by a user match? 5. As with the volumetric data I want to provide a 3D interactive viewer. I had some initial success with Brain Browser. Are you aware of any other javascript based viewers that work well with FreeSurfer data? 6. Which surface should be the default view: inflated, pial, or white? 7. Finally NeuroVault gives users ability to compare maps and even search through the whole repository to find statistical maps with similar patterns. This is currently implemented for volumetric data, but could be extended to surface data. The problem arises when one wants to compare MNI volumetric data with fsaverage surface data. So far I have been using mri_vol2surf with target subject being fsaverage (with a supplied affine correction matrix), but this solution is far from optimal since fsaverage folding patterns are quite different from MNI. I was thinking that maybe a more accurate way of dealing with this would be to reconstruct MNI152 cortical surface model, do surface registration of it to fsaverage, and use that to sample statistical maps in MNI space. What do you think? Thank you for any help feedback you can provide! Best, Chris Gorgolewski

3 5

← Newer
1
...
5
6
7
8
9
10
11
...
20
Older →

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 August 2016