Freesurfer

freesurfer@nmr.mgh.harvard.edu

17 participants
30086 discussions

gaussian and mean curvature error
by Ana Catarino 23 Mar '10

23 Mar '10

Hello, I'm trying to do a gaussian curvature and mean curvature analysis using qdec but I get the following error message: Error in Analyze: Couldn't open file $SUBJECTS_DIR/AS1/surf/lh.white.K.fwhm10.avgsubject.mgh Can anyone help? Thank you, Ana Catarino

1 0

Retinotopy with FsFast
by p.kok＠fcdonders.ru.nl 23 Mar '10

23 Mar '10

Hello everyone, I used FsFast to do retinotopy analysis as described at https://surfer.nmr.mgh.harvard.edu/fswiki/FsFastIndividualRetinotopyAnalysis I have data for both a clockwise and a counter-clockwise rotating wedge, and when I do retinotopy on these separately, it seems to work fine, but when I put both runs in one retinotopy analysis (as 001 and 002), they seem to cancel each other out and the results looks rubbish; sliceview-sess shows very little activation and the surf-sess polar maps don't make sense. Do you have any idea what might be the problem here? Of course I defined opposite directions for both runs in the rtopy.par files. Some 'problems' with my data which might give you a clue as to what's going on: - I don't have eccentricity runs, so I copied polar data in a 'fake' eccentricity run. Doesn't seem to be the problem, since retinotopy works fine for the clockwise and counterclockwise runs separately. - My wedge starts at the top right of the screen, so I need to state an angle offset when watching the results in tksurfer (-0.125) in order for the colour coding to make sense (blue = horizontal meridian, right?) - Although the movement of the wedge fits perfectly with the scans in the clockwise run (each wedge lasts one TR and starts exactly at the start of a scan), there is a slight offset for the counterclockwise run; wedges still last exactly one TR but the first wedge doesn't start exactly at the start of a scan. This doesn't seem to cause problems when I analyse the runs separately, but could it explain why the activations seem to cancel each other out when I combine them in one analysis? I would appreciate any suggestions. Thanks, Peter Kok

1 0

DTI group analysis
by Ryan 23 Mar '10

23 Mar '10

Hello all, In trying to perform a group analysis on volumetric and DTI data (with test4d.nii.gz file created by fslmerge), I encounter an error. Any help is greatly appreciated. Also, can anyone tell me what programs they prefer to use for DTI group analysis? It seems QDEC isn't helpful if you need volumetric analyses. bash-3.2$ mri_glmfit --y test4d.nii.gz --fsgd /Applications/freesurfer2/subjects/glm/gender_age.fsgd dods --C /Applications/freesurfer2/subjects/glm/lh-Avg-thickness-age-Cor.mtx --glmdir lh.gender_age.glmdir gdfReadHeader: reading /Applications/freesurfer2/subjects/glm/gender_age.fsgd INFO: DeMeanFlag keyword not found, DeMeaning will NOT be done. Continuous Variable Means (all subjects) 0 Age 64.2 Class Means of each Continuous Variable 1 gender_male 52.5000 2 gender_female 72.0000 INFO: gd2mtx_method is dods $Id: mri_glmfit.c,v 1.138.2.3 2008/08/27 13:33:05 greve Exp $ cwd /Applications/freesurfer2/subjects/4964-59 cmdline mri_glmfit --y test4d.nii.gz --fsgd /Applications/freesurfer2/subjects/glm/gender_age.fsgd dods --C /Applications/freesurfer2/subjects/glm/lh-Avg-thickness-age-Cor.mtx --glmdir lh.gender_age.glmdir sysname Darwin hostname ryan-mcgarrys-mac-pro.local machine i386 user ryanmcgarry FixVertexAreaFlag = 1 UseMaskWithSmoothing 1 OneSampleGroupMean 0 y /Applications/freesurfer2/subjects/4964-59/test4d.nii.gz logyflag 0 usedti 0 FSGD /Applications/freesurfer2/subjects/glm/gender_age.fsgd glmdir lh.gender_age.glmdir DoFFx 0 Creating output directory lh.gender_age.glmdir Loading y from /Applications/freesurfer2/subjects/4964-59/test4d.nii.gz Matrix condition is 1765.88 Pruning voxels by frame. Found 279419 voxels in mask Saving mask to lh.gender_age.glmdir/mask.mgh search space = 2.23535e+06 ERROR: dimension mismatch between y and X. y has 2 inputs, X has 5 rows. I also ran bugr and the following output was produced: FREESURFER_HOME: /Applications/freesurfer3 Build stamp: freesurfer-Darwin-leopard-i686-stable-pub-v4.5.0 Kernel info: Darwin 9.8.0 i386

1 0

DTI group analysis
by Ryan 23 Mar '10

23 Mar '10

1 0

Re: [Freesurfer] Freesurfer Digest, Vol 73, Issue 56
by Sebastian Moeller 22 Mar '10

22 Mar '10

Hi Janani, I seem to recall that newer distributions (and judging from your kernel it is a rather new distribution) come with libtiff.so.4 instead of libtiff.so.3. My solution to this problem so far just was to symlink libtiff.so.4 to libtiff.so.3. (Just go in the directory where the library lives (might be /usr/lib or /usr/lib64) and issued the following command: "ln -s ./libtiff.so.4 ./libbtiff.so.3", you might have to get root permissions first (or ask the system administrator for help)). While there might be differences between the two versions I did not run into problems with freesurfer yet... best Sebastian On Mar 22, 2010, at 9:00 AM, freesurfer-request(a)nmr.mgh.harvard.edu wrote: > Message: 2 > Date: Mon, 22 Mar 2010 09:18:11 -0400 (EDT) > From: Bruce Fischl <fischl(a)nmr.mgh.harvard.edu> > Subject: Re: [Freesurfer] Recon runs longer than the previous time and > exits this time with errors > To: Janani Dhinakaran <janinthesky(a)gmail.com> > Cc: freesurfer(a)nmr.mgh.harvard.edu > Message-ID: > <Pine.LNX.4.62.1003220917180.30973(a)gate.nmr.mgh.harvard.edu> > Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed > > Hi Janani, > > do you not have libtiff installed on your system? Maybe your > LD_LIBRARY_PATH is incorrect? You should talk to a local sysadmin about it. > > cheers > Bruce > > > > > On Mon, 22 Mar > 2010, Janani Dhinakaran wrote: > >> I was unable to extract parts of the cortex with the commands you had told >> me to try last time. I realised then that it was because recon hadn't >> finished running and I was using partially processed data. So I tried >> running it again. Here below is what I got. Since I'm using a partitioned >> machine space may be a problem but it doesn't seem to say so. >> Thanks in advance for the advice. >> >> #-------------------------------------------- >> #@# Cortical ribbon mask Sat Mar 20 07:17:29 GMT 2010 >> /home/s0964091/Desktop/jan/mri >> >> mris_volmask --label_left_white 2 --label_left_ribbon 3 --label_right_white >> 41 --label_right_ribbon 42 --save_ribbon --save_distance jan >> >> mris_volmask.bin: error while loading shared libraries: libtiff.so.3: cannot >> open shared object file: No such file or directory >> Linux CRIC-0012 2.6.31-20-generic #57-Ubuntu SMP Mon Feb 8 09:02:26 UTC 2010 >> x86_64 GNU/Linux >> >> recon-all exited with ERRORS at Sat Mar 20 07:17:29 GMT 2010 >> >> >> >

1 0

Recon runs longer than the previous time and exits this time with errors
by Janani Dhinakaran 22 Mar '10

22 Mar '10

I was unable to extract parts of the cortex with the commands you had told me to try last time. I realised then that it was because recon hadn't finished running and I was using partially processed data. So I tried running it again. Here below is what I got. Since I'm using a partitioned machine space may be a problem but it doesn't seem to say so. Thanks in advance for the advice. #-------------------------------------------- #@# Cortical ribbon mask Sat Mar 20 07:17:29 GMT 2010 /home/s0964091/Desktop/jan/mri mris_volmask --label_left_white 2 --label_left_ribbon 3 --label_right_white 41 --label_right_ribbon 42 --save_ribbon --save_distance jan mris_volmask.bin: error while loading shared libraries: libtiff.so.3: cannot open shared object file: No such file or directory Linux CRIC-0012 2.6.31-20-generic #57-Ubuntu SMP Mon Feb 8 09:02:26 UTC 2010 x86_64 GNU/Linux recon-all exited with ERRORS at Sat Mar 20 07:17:29 GMT 2010 -- I am the Sun, I know no darkness.

2 1

volume segmentation corrections
by Roser Sala 22 Mar '10

22 Mar '10

Hello, We are running FreeSurfer in order to obtain Callosal volumes from pathologic brain. Segmentation volumes (aseg.mgz) seem 'not bad', but specially in the Callossum there are some misclassifications (attached image with an example). Which would be the best way to correct this? Do you recommend us to add some control points? we are not sure if adding control points will affect only the surface reconstruction, or will help aslo to correct segmentation, which is our purpose. Many thanks in advance - Roser -

2 1

White matter volumes of different lobes?
by Gergely Orsi 22 Mar '10

22 Mar '10

Hi all. Is there any fast method to get the total (R+L) white matter volumes of different lobes? I need the: average ratio of temporal lobe wm to total white matter average ratio of parietal lobe wm to total white matter average ratio of frontal lobe wm to total white matter average ratio of occipital lobe wm to total white matter in our 100 normal healthy population. Autorecon 1-2-3 (and qcache) is already done. Anyone can help in this? Thank you. Gergely Orsi

1 0

Re: [Freesurfer] Freesurfer Digest, Vol 73, Issue 52
by Sebastian Moeller 19 Mar '10

19 Mar '10

Hi Paige, I ran into a similar problem a few months ago. Freesurfer 3.0.5 could not read data from a TIM trio with numaris B17. I solved this by transplanting (copying) a number of files from freesurfer 4.4.0 (the recent one at the time) into 3.0.5. unpacsdcmdir, mri_convert and a few more, that worked okay, but I since switched to freesurfer 4.5.0 for good. If you have to stick to 3.0.3 I can try to look up which files I had to replace in 3.0.5. Or alternatively you can install v4.5.0 parallel to 3.0.3 and only use it to import the DICOM data... best regards Sebastian On Mar 19, 2010, at 12:37 PM, freesurfer-request(a)nmr.mgh.harvard.edu wrote: > Message: 7 > Date: Fri, 19 Mar 2010 14:37:24 -0500 > From: Paige Scalf <pscalf(a)gmail.com> > Subject: [Freesurfer] trio dicom files and mri_convert > To: freesurfer(a)nmr.mgh.harvard.edu > Message-ID: > <1113e70c1003191237y6f5723aeha927124dd30b593e(a)mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > Hello, > > I've been trying to use mri_convert > (freesurfer-Darwin-tiger-stable-pub-v3.0.3) to reconstruct images from our > Siemens Magnetom Trio. When I type: > > mri_convert -it siemens_dicom -ot analyze4d DB_A0A1_318.MR.BECK-PAIGE_ > A0A1_INHABITION.0003.0001.2010.03.18.16.34.50.500000.21615932.IMA Block1.img > > > DB*.IMA and Block1.img are the input and output files respectively, I get > the output detailed below (the salient error seems to be that mri_convert > can't find a useful transformation matrix). > > Can you tell me what I can do to correct this problem? > Many thanks in advance, > > Paige Scalf > > > mri_convert -it siemens_dicom -ot analyze4d > DB_A0A1_318.MR.BECK-PAIGE_A0A1_INHABITION.0003.0001.2010.03.18.16.34.50.500000.21615932.IMA > Block1.img > reading from > DB_A0A1_318.MR.BECK-PAIGE_A0A1_INHABITION.0003.0001.2010.03.18.16.34.50.500000.21615932.IMA... > Getting Series No > Scanning Directory > INFO: Found 111 files in /Volumes/dubois/A0A1inhibition/DB_A0A1_318/Block1 > INFO: Scanning for Series Number 3 > INFO: found 107 files in series > INFO: loading series header info. > > INFO: sorting. > RunNo = 2 > INFO: (120 110 50), nframes = 107, ismosaic=1 > ERROR: incorreclty formatted version string syngo MR B17 > found in dicom tag 18,1020 (len = 3 != 6) > Numaris Version: syngo MR B17 Maj = 0, Min=-1073775504, MinMin = 0 > mojo:/Volumes/dubois/A0A1inhibition/DB_A0A1_318/Block1 paige$ mri_convert > -it dicom -ot analyze4d > DB_A0A1_318.MR.BECK-PAIGE_A0A1_INHABITION.0003.0001.2010.03.18.16.34.50.500000.21615932.IMA > Block1.img > mri_convert -it dicom -ot analyze4d > DB_A0A1_318.MR.BECK-PAIGE_A0A1_INHABITION.0003.0001.2010.03.18.16.34.50.500000.21615932.IMA > Block1.img > reading from > DB_A0A1_318.MR.BECK-PAIGE_A0A1_INHABITION.0003.0001.2010.03.18.16.34.50.500000.21615932.IMA... > Getting Series No > Scanning Directory > INFO: Found 111 files in /Volumes/dubois/A0A1inhibition/DB_A0A1_318/Block1 > INFO: Scanning for Series Number 3 > INFO: found 107 files in series > INFO: loading series header info. > > INFO: sorting. > RunNo = 2 > INFO: (120 110 50), nframes = 107, ismosaic=1 > ERROR: incorreclty formatted version string syngo MR B17 > found in dicom tag 18,1020 (len = 3 != 6) > Numaris Version: syngo MR B17 Maj = -1073770984, Min=22021328, MinMin = > -1610546680 > 107 DICOM 3.0 files in list > Found 107 DICOM Files > WARNING: NumberOfFrames 0 != Found Count of slices 107. > reading DICOM image... > ------------------------------------------------- > DICOM meta-header > > file name > /Volumes/dubois/A0A1inhibition/DB_A0A1_318/Block1/DB_A0A1_318.MR.BECK-PAIGE_A0A1_INHABITION.0003.0001.2010.03.18.16.34.50.500000.21615932.IMA > Date and time > study date 20100318 > study time 160300.312000 > series time 162802.453000 > acquisition time 162802.453000 > Identification > patient name DB_A0A1_318 > manufacturer SIEMENS > Dimensions > number of rows 880 > number of columns 960 > number of frames 107 > pixel width 2.13333 > pixel height 2.13333 > slice thickness 2.1 > field of view 1877.33 > image number 1 (might be not reliable) > transfer syntax UID 1.2.840.10008.1.2.1 > Acquisition parameters > echo time 25 > repetition time 3000 > inversion time not found > echo number 1 > flip angle 1.5708 > bits allocated 16 > Spatial information > first image position -1027.46 -12.1727 952.833 > last image position -1027.46 -12.1727 952.833 > image orientation 1 0 0 0 -0.00174528 -0.999998 > ------------------------------------------------- > > TR=3000.00, TE=25.00, TI=0.00, flip angle=90.00 > i_ras = (-1, -0, 0) > j_ras = (-0, 0.00174528, -0.999999) > k_ras = (nan, nan, nan) > Reslicing using trilinear interpolation > MRIresample(): error inverting matrix; determinant is nan, matrix is: > -2.133 -0.000 nan nan; > -0.000 0.004 nan nan; > 0.000 -2.133 nan nan; > 0.000 0.000 0.000 1.000; >

3 3

Re: [Freesurfer] merging rh and lh surfaces
by Raghav Puranmalka 19 Mar '10

19 Mar '10

Hi all - I was trying to merge rh and lh surfaces. I succeeded after reading the mailing list archives from March 2009 found at http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg09665.html. I following this guide: you can get something like what you want with cd $SUBJECTS_DIR/subject mri_binarize --i mri/ribbon.mgz --min .0001 --o mri/tmp.mgh mri_tessellate mri/tmp.mgh 1 surf/lh.both mris_smooth -nw surf/lh.both surf/lh.smboth tksurfer subject lh smboth Note that eventhough you call it "lh" it has both hemis. "lh" is just to fool freesurfer into thinking it is a hemi. The issue I run into is that I try to get matlab to read the surface through read_surf, I find that the magic number is not recognized. Has anyone experienced this before? Any thoughts about navigating through this? Best, Raghav

2 3

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