Freesurfer

freesurfer@nmr.mgh.harvard.edu

18 participants
30131 discussions

FW: difference # vertices in sphere&
by Kirk Gregory R. 09 Jul '09

09 Jul '09

hi q. from a collaborator, someone got a comment for 1 subject the # of vertices of sphere file different from the length of thickness data for the same subject. thanks Greg. -----Original Message----- From: Deepti Pachauri [mailto:deepti.pachauri@gmail.com] Sent: Wed 6/24/2009 1:02 AM To: Kirk Gregory R. Subject: Re: Help Deepti with some questions Dear Greg I have been looking at FreeSurfer data provided by you for 21 subjects(AD/Control). For P1_AD subject I found the # of vertices of sphere file different from the length of thickness data for the same subject. # of vertices for P1.AD subject are 131364 and length of thickness data for the same subject is 131215. However for rest 20 subjects # of vertices and length of thickness data are same. I am using MATLAB to read thickness.asc data and last column is thickness (mm) at given x,y,z. For vertices and faces I am using the standard freesurfer_read_surf.m form http://eeg.sourceforge.net/doc_m2html/bioelectromagnetism/freesurfer_read_s… I am just wondering how to fix this discrepancy of freesurfer output? Also thickness data P10 is missing in asc folder. Sphere and pial files are present in surf folder for P10 subject. Thanks On Tue, Jun 23, 2009 at 3:55 PM, Vikas Singh <vsingh(a)biostat.wisc.edu>wrote: > Hi Greg, > > Deepti, a new CS grad student (copied on this email) is studying the > cortical surface characterization problem, and using the output you > generated. Is is possible that you can answer some of her questions > regarding FreeSurfer? I am copying her on this email. I am out of town > until July 16th, and we wanted to get on top of this over the summers. > Please let us know, possibly with a phone number she can reach you at. > > regards, > > Vikas. > -- Deepti Pachauri

2 1

segmentation
by simon＠cyceron.fr 09 Jul '09

09 Jul '09

Hi. I'm a new user of freesurfer. The default settings of recon-all performed a good white segmentation of my data but I find that grey matter (see attached picture) could be optimized when I look at aseg.mgz results. Indeed the surface cortical grey matter was not extend as expected. What were the different parameters that I can optimize to have a better grey matter segmentation without manual correction? Thank a lot for your answers. Gregory SIMON

2 1

contrast matrix
by Feng-Xian 09 Jul '09

09 Jul '09

Hi, I want to analysis three independence groups. And, I want to display a result of A group > B group > C group with positive(red) and A group < B group < C group with negative(blue). To show that, how do I write the number of columns in the contrast matrix? Thank you in advance! Feng-Xian ------------------------------------------------- Chang Gung University Mail Server Web Mail system By Chang-Gung University Information Center (http://www.cgu.edu.tw/ic/)

2 2

POSTDOC POSITION OPEN
by Juranek, Jenifer 08 Jul '09

08 Jul '09

Subject: POSTDOCTORAL POSITION - structure & function of fronto-cerebellar circuits in neurodevelopmental disorders Applications are invited for an immediate postdoctoral opening in the Department of Psychology at the University of Houston, Houston TX. The position is full-time for 1 year, with the possibility of annual extensions up to 2 years, depending on funding. The current opening entails analysis of already collected MRI datasets from children with and without a neurodevelopmental disorder and relating quantitative MRI findings to neuropsychological assessments. The successful candidate will be responsible for processing and analyzing structural data and writing up the results for publication in peer-reviewed journals. Essential skills include demonstrated knowledge of structural MRI analysis techniques, including Freesurfer and FSL. While this position requires some technical skill with image analysis, it is best suited for someone who is interested in understanding brain-behavior correlates of cognitive function in children with and without a developmental disorder rather than someone with a strict methods development orientation. Applications are welcomed from current Ph.D. students who will have submitted their dissertations in the near future. The position is available for a start date as soon as July 15th 2009. Salary is expected to be $40,000, depending on knowledge and experience, and will come with full health and other benefits at the University of Houston, Houston TX. Non US citizens are welcome to apply. Please send enquiries or applications (which must include a covering letter detailing professional objectives and interests, CV, and the names and email addresses of three references, all together in one pdf file) to Dr. Jack M Fletcher, jmfletch(a)central.uh.edu<mailto:jmfletch@central.uh.edu> , with the subject line "Postdoctoral imaging position". ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ Jenifer Juranek, PhD Assistant Professor Developmental Pediatrics University of Texas Health Science Center at Houston 713.500.3813 (voice) Jenifer.Juranek(a)uth.tmc.edu

1 0

RoI,Space
by Barnali Basu 08 Jul '09

08 Jul '09

Experts, I have some snapshots of ROIs made in BrainVoyager (in non-TAL space on inflated brain, as well as ones projected onto TAL space). I need to take the above as reference and make near-exact ones (ROIs) using Freesurfer on inflated brains. Which one of the above do I use as my reference Barnali

2 6

Fwd: [Freesurfer] mri_label2vol
by Natalia Sanchez 08 Jul '09

08 Jul '09

Hi FreeSurfer experts! I am kind of new to FreeSurfer and this is my first time using the software! I am using Freesurfer to create a corrtical mask for M1 for probabilistic Tracking in FSL. I am doing recon-all for segmentation, then in tksurfer I am selecting a label for the precentral gyrus for the left hemisphere for the subject of interest. The label is selected in the same subject in which I am planning on doing the mask, so I think it is not necessary to do label2label. Afterwards, I am doing vol2roi which gives me a .nii output. When plotting this in FSL, I have a complete misalignment of the ROI, which is not corrected by swapping the image. Obviously this means I am having problems with the registration. How can I improve this? I have been using the mni152register.dat for the registration, and I also tried doing fslregister before running vol2roi. I have also tried using label2vol but the output has given me the same problem. Which of the tools is better for this type of process label2vol or vol2roi? I attached a snapshot of the way the mask looks, (which is really bad) Am I following an adequate procedure, does this make any sense at all, is this a good approach to integrating FreeSurfer with FSL? Is there a more efficient way to do cortical masks? I would really appreciate any help! Thank you -- Natalia Sanchez -- Natalia Sanchez

2 1

recon-all qcache
by Victoria Williams 08 Jul '09

08 Jul '09

Hello, I was hoping to get some more information about the qcache option. Would running all my subjects with qcache save me any time if I plan on using mri_glmfit instead of qdec? The GLMs I am running have to be sent through mris_preproc and mri_surf2surf each time since some inputs are #'d out due to missing variables in the fsgd file and each GLM differs. I was wondering if somehow mri_glmfit could use the fsgd file inputs to concatenate the already smoothed data outputted by qcache. Thus, saving me having to concatenate and smooth over and over again due to slight differences. Also, where is the output directory for qcahce? Is it in the average subject's directory or individual subject directories?? Thanks! Tori

2 1

reconstructing .mnc file with freesurfer
by Jesse Bledsoe 08 Jul '09

08 Jul '09

Hi Everyone, What is the process for reconstructing a .mnc file with freesurfer? Do I just run the recon-all -s *.mncfile* -autorecon-all ? Thanks in advance, jesse ______________________________________ Note: Information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately and destroy all copies of this message and any attachments.

3 2

dimension mismatch error
by Markus Donix 08 Jul '09

08 Jul '09

Hi all, after running a group comparison using qdec I wanted to use mc-z instead of FDR for multiple comparison correction I tried to run mri_glmfit-sim-mc-z.abs.2 and got an error... ERROR: dimension mismatch between X and contrast /Users/..../subjects/ qdec/Untitled/tmp.mri_glmfit-sim-5637/lh-Avg-Intercept-area.pial.mtx X has 8 cols, C has 4 cols what does this mean and how to fix it? thanks, Mark

1 0

BUG in multiple comparisons correction for ABS
by Douglas N Greve 08 Jul '09

08 Jul '09

Nasim and Judith (and everyone else), I've found and fixed a bug that was causing the surfcluster results for the mc and perm simulations of absolute "abs" contrasts. To implement the fix, get Linux 64bit versions of these files and copy them to $FREESURFER_HOME/bin (make backups of the originals:). ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_surfclust… ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_glmfit You can then run something like: mri_glmfit-sim --glmdir yourglmdir --no-sim mc-z.abs The "--no-sim" option tells it NOT to do the full simulation again, just use the CSDs that are there to recompute the cluster results for each contrast, so it should be done in a matter of minutes. If you've used a permutation, then run mri_glmfit-sim --glmdir yourglmdir --no-sim perm.abs Please let me know if this fixes the problem. doug

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