Freesurfer March 2008

freesurfer@nmr.mgh.harvard.edu

55 participants
81 discussions

mri_ca_label segfaults
by Martin Kavec 11 Mar '08

11 Mar '08

Hi, I have an emergency here with recon-all -all on a VIP, which segfaults on mri_ca_label -align -nobigventricles ...I tried to run it from the command line and left out the -nobigventricles argument, since actually the subject has enlarged ventricles, but it didn't help. After running for about 45 minutes the last messages are: 471 hippocampal voxels changed. 1 amygdala voxel changed. pass 1: 96288 changed. image 11: -2.434, PF=1.000 pass 2: 14342 changed. image 11: -2.434, PF=1.000 pass3: 5141 changed. Bus error # The images were acquired with our standard protocol, which yields in 95% successful FS results. I run the analysis (exceptionally) on a MacBook Pro with the latest FS. Thanks a lot in advance, Martin

1 0

mri_glmfit problem
by Nicolas Cherbuin 11 Mar '08

11 Mar '08

Hi, I am analysing structural data with mri_glmfit. After pre-processing I have first used one group variable (gender) with one covariate age. I had no problem processing the data and visualising it in tksurfer. I then tried to add another covariate (measure of immediate recall). So I had in total one group variable and two covariates. I reanalysed the data but got the following error: FixVertexAreaFlag = 1 UseMaskWithSmoothing 1 OneSampleGroupMean 0 y /media/storage2/proc/freesurfer/60_w2/checked/analysis/lh.gender_age_immrec.thickness.10.mgh logyflag 0 usedti 0 FSGD gender_age_immrec_fsgd.txt glmdir lh.gender_age_immrec.glmdir DoFFx 0 Creating output directory lh.gender_age_immrec.glmdir Loading y from /media/storage2/proc/freesurfer/60_w2/checked/analysis/lh.gender_age_immrec.thickness.10.mgh Matrix condition is 12866.7 ERROR: matrix is ill-conditioned or badly scaled, condno = 12866.7 Possible problem with experimental design: Check for duplicate entries and/or lack of range of continuous variables within a class. 0.000 1.000 68.000 10.000; 1.000 0.000 65.000 5.000; [trunkated the fsgd file output but it is the same as the one used for the sucessful analysis except for one extra column] After this error I thoroughly checked all my files but could not find anything wrong. So to check I could do an analysis using the extra variable I did another analysis with gender and immrec. This worked without problem. I have attached all my fsgd files and mat files. They all use the same subjects and gender info and only differ by one column. Since permutations worked I cannot fathom why it does not work when all put together. The mri_glmfit syntax used for each analysis was: mri_glmfit --y lh.gender_age_immrec.thickness.10.mgh --fsgd gender_age_immrec_fsgd.txt doss --glmdir lh.gender_age_immrec.glmdir --surf average100 lh --C gender_age_immrec.mat mri_glmfit --y lh.gender_immrec.thickness.10.mgh --fsgd gender_immrec_fsgd.txt doss --glmdir lh.gender_immrec.glmdir --surf average100 lh --C gender_immrec.mat mri_glmfit --y lh.gender_age_immrec.thickness.10.mgh --fsgd gender_age_immrec_fsgd.txt doss --glmdir lh.gender_age_immrec.glmdir --surf average100 lh --C gender_age_immrec.mat I use 4.01 (but also checked with 4.02) under Ubuntu 64 bit on a Xeon machine. Thanks very much for your help Nic

2 3

more de-meaning
by Doug Greve 11 Mar '08

11 Mar '08

Hi Chistophe, I'm not surprised by your results as it is very likely that DIQ and F1 have non-zero means. If you were to demean them, then the no-variable analysis would be the same as the with-variable analysis. I actually just sent an email to the Freesurfer list on this topic of demeaning, and I'm copying the list on this response. The important idea here is the difference between computing/testing the mean response and computing the intercept. The intercept is only meaningful when a continuous variable is present. When you do a statistical test, the results will be different for two possible reasons: 1. The statistical efficiencies are different (due to correlations caused by the non-zero mean of the continuous variable), which will reduce significance. This is always the case. 2. The values being tested (ie, the mean vs the intercept) are possibly different. I say "possibly" here because, scientifically speaking, there may be no effect of your continuous variable, in which case it would not add to the overall mean. In my opinion, the most appropriate way to analyze the data is to leave the means in your continuous variables (ie, do not demean). Here's why. When you believe that a variable is important scientifically, you posit a model with a population effect (parameterized by an intercept) and the continuous variable (parameterized by a slope). BOTH OF THESE PARAMETERS ARE INDEPENDENT OF THE SAMPLE YOU HAVE CHOSEN. Therefore, when you perform statistical tests on these parameters, your results are independent of your sample -- very important when doing science! In contrast, the mean of your sample may be dependent on the sample you have chosen, and so statistical tests may only then apply to your sample. For example, if age adds to the hemodynamic response (HRF), older subjects will have a larger amplitude to their HRF. Let's say you perform an experiment on a group of subjects and find that their mean HRF is significantly different than 0. Someone else tries to replicate your experiment and can't. Upon further examination, your sample was somewhat older than the other which caused your sample to have a higher mean and so achieve significance. When both sets are reanalyzed using age as a nuisance variable, the results are the same. This does not necessarily mean you should not demean your variables, but you just have to be careful what conclusions you draw from it. Note: this applies to variables and regressors -- don't demean your observed raw data (unless you know what you are doing). doug Christophe Destrieux wrote: > Hi Doug > > We are using mri_glmfit to study the influence on 2 factors (DIQ anf > F1) on activations observed within a group of autistic patients. > > We used a fsgdf with only one class (autists), and 2 variables : DIQ > and F1, the matrix was computed using DOSS > > We used the following contrast : > > * "0 1 0", to study the influence of DIQ on activations > * "0 0 1", to study the influence of F1 on activations > > We thought that the contrast 1 0 0 would represent activations of the > whole group independently of F1 or DIQ ; > > For comparison purposes we also computed the same map with one class > but without any variable ; We assumed that the contrast that should > represent activation of the whole group should then be "1" > > We expected the same results in both analysis, but in fac maps are > very different : they look ok with the second analysis (1class, 0 > variable, contrast "1"), but were almost empty with the first one (1 > class, 2 variables, contrast "1 0 0") > > There is probably something wrong but we cant find out where > > thanks > > PS the message is CC to Frederic Andersson who joined the lab a few > weeks ago ; he is engineer and he will probably bug you from time to > time : ) > > cheers > -- Douglas N. Greve, Ph.D. MGH-NMR Center greve(a)nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422 In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting

1 0

Matlab and Freesurfer
by Robert Baldwin 09 Mar '08

09 Mar '08

Hello- I am new to Freesurfer and Matlab. I have Matlab 7.4.0 installed on an Intel Mac. I just installed Freesurfer. Now when I start Matlab I get the following warnings: Warning: Name is nonexistent or not a directory: /matlab. In path at 110 In startup at 4 In matlabrc at 255 Warning: Name is non existent or not a directory: /toolbox. In path at 110 In startup at 11 In Matlabrc at 255 I have tried adding the freesurfer folder to these files to the path for matlab, but the error still occurs. Any suggestions on how to get this error to quit? Thanks! -Robert

2 4

cerebellum
by Anthony Dick 07 Mar '08

07 Mar '08

Hello- Are there any plans to implement the cerebellum segmentation of Makris et al. (2005; Neuroimage, 25) as an automated segmentation into part of the freesurfer workflow? Anthony -- Anthony Steven Dick, Ph.D. Post-Doctoral Fellow Human Neuroscience Laboratory Department of Neurology The University of Chicago 5841 S. Maryland Ave. MC-2030 Chicago, IL 60637 Phone: (773)-834-7770 Email: adick(a)uchicago.edu Web: http://home.uchicago.edu/~adick/

2 1

Re: problems with retaining wm edits using -autorecon2-cp flag
by Teddy Youn 06 Mar '08

06 Mar '08

Hi all, another quick question. I don't know if anyone else has encountered this problem, but after making pial, wm and cp edits our group don't seem to retain all our edits. The OtherUsefulFlags wikipage suggests that FS should retain all the edits from a flag that is more upstream in the autorecon2's workflow. We've noticed that a lot of our manual wm edits for some of our brains are lost when we use -autorecon2-cp flag. Has anyone else encountered this problem? Is there a way to prevent this from happening? -Thanks in advance, Teddy Doris Duke Clinical Research Fellow '07-'08 North Texas TBI Research Center UT Southwestern Medical Center at Dallas Warren Alpert School of Medicine at Brown University, MD '09 Teddy.Youn(a)utsouthwestern.edu (401)-714-2847

2 1

optseq2 obscenely long null event
by Jasony Brozinsky 06 Mar '08

06 Mar '08

Hi, Optseq2 is returning designs where the last NULL even is upwards of 75 seconds (more than 4 times my psdMax!. Can I keep the rest of the order as is, and just truncate the last NULL event to be equal to my PsdMax value? Below is my command and a sample output file: optseq2 --ntp 130 --tr 2 --cost eff --tprescan 0 --psdwin 0 18 1 --ev EVT3 6 2 --ev EVT4 7 2 --ev EVT5 8 2 --ev EVS3 6 2 --ev EVS4 7 2 --ev EVS5 8 --ev Other 11 2 --tnullmin 2 --tnullmax 2 --nsearch 10000 --focb 1000 --nkeep 100 --o utput --log log --polyfit 2 0.0000 2 7.000 1.0000 EvT4 7.0000 0 6.000 1.0000 NULL 13.0000 7 11.000 1.0000 Other 24.0000 0 6.000 1.0000 NULL 30.0000 7 11.000 1.0000 Other 41.0000 0 4.000 1.0000 NULL 45.0000 3 8.000 1.0000 EvT5 53.0000 0 6.000 1.0000 NULL 59.0000 5 7.000 1.0000 EvS4 66.0000 0 8.000 1.0000 NULL 74.0000 5 7.000 1.0000 EvS4 81.0000 0 4.000 1.0000 NULL 85.0000 1 6.000 1.0000 EvT3 91.0000 0 5.000 1.0000 NULL 96.0000 4 6.000 1.0000 EvS3 102.0000 0 4.000 1.0000 NULL 106.0000 1 6.000 1.0000 EvT3 112.0000 0 6.000 1.0000 NULL 118.0000 6 8.000 1.0000 EvS5 126.0000 0 12.000 1.0000 NULL 138.0000 3 8.000 1.0000 EvT5 146.0000 0 9.000 1.0000 NULL 155.0000 2 7.000 1.0000 EvT4 162.0000 0 5.000 1.0000 NULL 167.0000 6 8.000 1.0000 EvS5 175.0000 0 2.000 1.0000 NULL 177.0000 4 6.000 1.0000 EvS3 183.0000 0 77.000 1.0000 NULL Craig ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs

2 1

2 Questions about Parcellation
by Raymond Hill 06 Mar '08

06 Mar '08

Hello, We are beginning to explore the applicability of FreeSurfer for our research. Having done some initial runs, we have the following questions that I was hoping you might be able to address. 1.) For stats output, it seems there are two atlases for parcellation, resulting in ?h.aparc.stats and ?h.aparc.a2005s.stats. I was wondering if you could tell me which atlases were used during the manual parcellation that ultimately went into creating these two systems? 2.) We were wondering what the entity "unknown" refers to in the ? h.aparc.stats output. What is this capturing? Thanks so much for any assistance you can provide and kudos on this amazing software! Sincerely, Ray

2 1

rawavg
by joel bruss 06 Mar '08

06 Mar '08

When FreeSurfer performs the average of 2 or more scans to create rawavg.mgz, is the 001.mgz used as the target to which all other scans are moved /averaged to or is there some between-space alignment amongst the scans? -Joel

2 1

CentOS 5.1 and freesurfer and tksurfer
by tulu＠ms36.hinet.net 06 Mar '08

06 Mar '08

Dear FS experts, I install CentOS 5.1 and latest version of ATI driver and FS4.0.2, when I start tkmedit, it looks fine. In tksurfer, it just appear a sliver of surface without error message. In console, it stock on reading vertex. Is there anyway to correct it ? Thanks replay Kl chen

5 4

← Newer
1
2
3
4
5
6
7
8
9
Older →

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Freesurfer March 2008