Freesurfer

freesurfer@nmr.mgh.harvard.edu

30132 discussions

Start a nNew thread

FreeSurfer6 on Ubuntu
by Fereshte 18 Sep '17

18 Sep '17

Dear FreeSurfer Developers Is FreeSurfer 6 available on Ubunto? If so, does it work properly? Thank you!

2 1

FS-FAST longitudinal analyis
by stdp82＠virgilio.it 17 Sep '17

17 Sep '17

Hi list, I'm not sure that the files reported below are correct to assess the longitudinal difference between T1 and T2.Please, could you check it? The result which I have obtained bu running mri_glmfit is too strange.ThanksBest regardsStefano 1) PAIRWISE.fsgd file-------------------GroupDescriptorFile 1Class subj1Class subj2...Class subjNVariables TP1-vs-TP2Input subj1_tp1 subj1 1Input subj1_tp2 subj1 -1Input subj2_tp1 subj2 1Input subj2_tp2 subj2 -1...Input subjN_tp1 subjN 1Input subjN_tp2 subjN -1---------------------------- 2) mean.mtx1 1 ... 1 0------------------------------3) tp-effect.mtx0 0 ... 0 1------------------------------4) mri_glmfit --glmdir PAIRWISE --y ces.nii.gz --wls cesvar.nii.gz --surface fsaverage rh --nii.gz --fsgd PAIRWISE.fsgd doss --C mean.mtx --C tp-effect.mtx

2 1

Multiple comparison, color maps and contrast
by miracle ozzoude 17 Sep '17

17 Sep '17

2 2

postdoc ad
by Lilla Zollei 16 Sep '17

16 Sep '17

For info see below or in the attached file. >>>>>>>>>>>>>>>>>>>>> Postdoctoral Research Fellow Position in Infant and Child Brain MRI analysis at the A.A. Martinos Center, MGH and Harvard Medical School, and Harvard University under Dr. Lilla Zöllei (MGH) and Leah Somerville (Harvard) A postdoctoral fellow is recruited for two NIH funded projects focusing on developing Freesurfer-compatible brain image analysis tools for children. One project aims to extend the tools for brain imaging data for children between 2-6 years of age. The second project aims to extend the tools for longitudinal analysis of developmental data from older children and adolescents as part of the Human Connectome Project in Development. Code development will focus on pair- and group-wise registration tools, segmentation methods and surface-based image analysis algorithms. The successful applicant should have a doctoral degree in a field related to medical imaging (e.g., computer science, electrical engineering, biomedical engineering, applied mathematics, computational neuroscience). She/he is expected to have a strong research background in method development and brain image analysis and should be able to work independently. Strong programming skills (for example, C/C++, MATLAB, and/or Python) are a must and prior experience with brain image-analysis packages (such as SPM, FSL, FreeSurfer, AFNI) is desirable. In addition, the candidate must also possess excellent English verbal and written communication skills. This position will be hosted by the Laboratory of Computational Neuroimaging, MGH and it will offer great opportunities to collaborate with a large group of scientists from diverse research disciplines. Approximate start date is fall 2017. Successful applicants will be appointed 50% effort at Massachusetts General Hospital, Boston and Harvard Medical School and 50% at Harvard University. For consideration please send a statement of interest, a CV and a list of three potential references via email to Lilla Zöllei, PhD (lzollei(a)nmr.mgh.harvard.edu) and Leah Somerville (somerville(a)fas.harvard.edu) The search will continue until the position is filled.

1 0

Re: [Freesurfer] correcting defect
by limengsecret＠126.com 16 Sep '17

16 Sep '17

Thank you, Dr Bruce. The initial surfaces indeed had large defects (figure 1) based on version 6.0. figure1 The following figure was obtained using version 5.3 for the same subject. (figure 2) It seems to be significant different between these two versions. I am sure that I have multiple threads available. When I add the -openmp 20 flag, it showed 2000% CPU utilization. Today when I processed the data using version 5.3 on another workstation, it showed errors as follows. ------------------------------ mri_nu_correct.mni --n 1 --proto-iters 1000 --distance 50 --no-rescale --i orig.mgz --o orig_nu.mgz Linux yxk-ubuntu 4.8.0-36-generic #36~16.04.1-Ubuntu SMP Sun Feb 5 09:39:57 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux recon-all -s try exited with ERRORS at ...... ----------------------- However, I cannot find the reason for this error. Any help would be greatly appreciated. Meng Date: Thu, 14 Sep 2017 11:42:02 -0400 (EDT) From: Bruce Fischl <fischl(a)nmr.mgh.harvard.edu> Subject: Re: [Freesurfer] correcting defect To: Freesurfer support list <freesurfer(a)nmr.mgh.harvard.edu> Message-ID: <alpine.LRH.2.20.1709141140430.19340(a)gate.nmr.mgh.harvard.edu> Content-Type: text/plain; charset="utf-8" Hi Meng we can't diagnose either of these issues without more information and possibly some screen shots. The topology fixer will take a very long time if the initial surfaces have large defects. You can view these by visualizing the ?h.inflated.nofix and ?h.orig.nofix. The -openmp flag works for us. Are you sure that you have mutliple threads available? cheers Bruce On Thu, 14 Sep 2017, Meng Li wrote: > Hi all, > > I tried to use FreeSurfer version 6.0.0 to process fifteen subjects, but half of the subjects was > stuck at the ?Correcting Defect? step. However, when I changed to FreeSurfer version 5.3.0, the > ?Correcting defect? step finished successfully. Is it possibility that the unsuccessful ?Correcting > Defect? step is caused by the different versions of Freesurefer? > > In addition, when I added the flag -parallel -openmp 30 based on the version 6.0.0, the recon-all > ?all runtime still need 20+ hours. It seems that the parallel processing did not reduce the > processing time. I wonder why the parallel processing did?t save time? > > Any help would be greatly appreciated. > > Meng > > > > > ? > > > ------------------------------ _______________________________________________ Freesurfer mailing list Freesurfer(a)nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer End of Freesurfer Digest, Vol 163, Issue 22 *******************************************

2 1

mixed-design with fs-fast
by stdp82＠virgilio.it 15 Sep '17

15 Sep '17

Hi list, I have 3 groups and two measure of rest-state at different time (baseline and t1).Using FS-FAST, I would look at:1-crossectional differences between groups (at baseline and T1)2- within group, the longitudinal differences between the conditions.Is possible to use a mixed-design with fast?Please could give me any suggestion on the statistical processing.ThanksRegards Stefano

1 0

how to get the volume of the complete ventricle system ?
by Benjamin Luerweg 15 Sep '17

15 Sep '17

dear all, Ive tried to get the volume of the ventricle system by using two different calculations. first of all i tried to get the volume by: brainsegvol - brainsegvolnotvent than i tried to simply add all the seprate parts of the ventricle system (rh lat ventricle + lh lat ventricle + 3rd, 4th and 5th ventricle) after doing this i realized that the two values were differing like -> 217820,00 (first calculation) and 20115,899999999998 (second calculation) any ideas why the difference is that big? kind regards

1 0

Running statistical analysis on surface data in R
by Mehta, Chintan 15 Sep '17

15 Sep '17

Dear FreeSurfer users, I am using R to apply a statistical method on surface data registered to fsaverage (for example, left hemisphere area smoothed at bandwidth=10mm). This method requires data be in a matrix Y with NS rows and V columns, where NS = #subjects and V=#vertices. The output produces a statistic Sj > 0 at each vertex j that, without loss of generality, rejects the null hypothesis for higher values with significance derived through permutation tests. Let's assume that Sj is a F-statistic with J & NS-NB-J degrees of freedom in a GLM with design matrix X (NS rows and NB columns) and contrast matrix CON (J rows and NB columns). While mri_glmfit can do the same for F-tests, my workflow uses matrix operations to produce the same F-statistics as mri_glmfit but considerably faster. My purpose is to apply a statistics method not supported by FreeSurfer and also use GPUs for permutation tests (since these methods rely on matrix operations). I am asking for the best way to manage the data (observed and permutation null distributions) so I can perform spatial cluster corrections using tools from FreeSurfer. My output will be a vector S = (S_1,...,S_V) containing observed statistics at the V vertices indexed in the same order as the input data matrix Y containing surface data for each subject. Performing Q permutation tests will produce matrix R with V rows that contain values of the statistic R_{j1}, ...,R_{jQ} at vertex j (j=1,...V) under the null distribution. P-value at a given vertex could be: P_j := (1/(Q+1)) * \sum_{q=1}^Q I( S_j < R_{jq}) I can do all the computations described above (and any conversions, like p-values to -log10 scale). Does FreeSurfer offer an efficient way of taking the vector S of observed statistics and matrix R of permutation distributions (or P-values on any scale) to identify significant clusters on the surface at some threshold? Thresholds could be applied equivalently to either P-values (P_j < P* := 0.001) or statistic value (S_j > S*). Significance of observed clusters would be evaluated under null distribution of the largest cluster from the permutation tests (at a given cluster defining threshold). Please forgive my unfamiliarity with how spatial information is dealt with in mgh files. I am also trying to make a cluster summary report (for example, the file cache.th30.abs.sig.cluster.summary when using cached simulations for a contrast in a GLM) and and also visualize significant clusters. I can export or analyze this data using any matlab function in the ~/freesurfer/matlab folder ("save_mgh" for example). For exporting, I'm unsure about how to preserve or define appropriate header information with save_mgh. To clarify, my workflow contains following steps: 1. mri_preproc: Produce DATA.mgh containing concatenated lh-area-fsaverage-fwhm10 for subjects 1,...,NS 2. Use "load_mgh" in Matlab to load DATA.mgh into matrix Y0 of dimension [V, 1, 1, NS], where V=#vertices. 3. Define matrix Y of dimension [NS, V], which is transpose of squeeze(Y0(:,1,1,:)). 4. Apply algorithm/test to matrix Y to obtain vector S of observed statistics at vertices and matrix R rows containing null distributions respective vertices. Indexing of vertices for entries in S and columns of R are in the same order as first dimension of matrix Y0 loaded using "load_mgh". Thank you for your attention to this very long question! Best, Chintan

1 0

Freeview aborts under centos6
by Markus Gschwind 15 Sep '17

15 Sep '17

Dear experts, I am installing freeusrfer on our cluster which is running Centos6. [gschwind@login1 ~]$ uname -a Linux login1.cluster 2.6.32-696.1.1.el6.x86_64 #1 SMP Tue Apr 11 17:13:24 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux Freeview aborts without any other complaint. [gschwind@login1 ~]$ freeview Abort When running freeview.bin it complains about a libraries error: [gschwind@login1 ~]$ freeview.bin freeview.bin: error while loading shared libraries: libvtkverdict.so.5.6: cannot open shared object file: No such file or directory I searched the archives and found this info to be given, [gschwind@login1 ~]$ ls -l $FREESURFER_HOME/lib/vtk/lib/ vtk-5.6/*libvtkverdict* -rwxr-xr-x 1 sagon unige 189017 Aug 17 2010 /opt/ebsofts/Core/FreeSurfer/ 5.3.0-centos6_x86_64/lib/vtk/lib/vtk-5.6/libvtkverdict.so -rwxr-xr-x 1 sagon unige 189017 Aug 17 2010 /opt/ebsofts/Core/FreeSurfer/ 5.3.0-centos6_x86_64/lib/vtk/lib/vtk-5.6/libvtkverdict.so.5.6 -rwxr-xr-x 1 sagon unige 189017 Aug 17 2010 /opt/ebsofts/Core/FreeSurfer/ 5.3.0-centos6_x86_64/lib/vtk/lib/vtk-5.6/libvtkverdict.so.5.6.0 I then integrated this into bashrc export LD_LIBRARY_PATH=$FREESURFER_HOME/lib/vtk/lib/vtk-5.6 however, the problem persists. Any help is highly appreciated! Thanks you! Markus

1 0

Repronim Training Workshop on Reproducible Research Nov 10-11 before SFN
by Satrajit Ghosh 14 Sep '17

14 Sep '17

*ReproNim Training Workshop: Nov 10 full day - Nov 11 morning - George Washington** University* *Register and information at:* https://tinyurl.com/repronim-sfn17 *Purpose:* The issue of lack of reproducibility has been described in several scientific domains for several years, raising concerns specifically in the life science community. ReproNim has developed a curriculum ( http://www.reproducibleimaging.org/#trai... <http://www.reproducibleimaging.org/#training>) that will give the students the information, tools and practices to perform repeatable and efficient research. This training workshop will introduce material on the critical aspects of reproducible brain imaging and will orient attendees using a hands on and practical experience to conduct neuroimaging analyses with the next generation of tools. By the end of this course, the student will be aware of training materials and concepts necessary to perform reproducible research in neuroimaging. The student will be able to reuse these materials to conduct local workshops and training and be able to customize the training for their specific scenario. *Prerequisites:* If you are a student, postdoc or researcher in life science who directly works with neuroimaging data - or wish to work with these data, and you have some basic computational background, this training workshop is for you. For instance, you should have already done either some R, or Python, or Matlab or Shell scripting, or have used standard neuroimaging tools (SPM, FSL, Afni, FreeSurfer, etc) and be engaged in neuroimaging research projects. You should have already completed a neuroimaging analysis or know how to do one. *Logistics:* *Location:* George Washington University, Marvin Center, Room 402-404 https://events-venues.gwu.edu/meeting-ro... <https://events-venues.gwu.edu/meeting-rooms> *Dates:* November 10-11, 2017. *Costs:* Free - but space is limited - please apply for approval. *Schedule:* Friday November 10th: 8:30-9am: Introduction to the course and participants "setup" 9am-10:45: Reproducibility Basics (Module 0) 10:45-11am : Coffee break 11am-12:45 : FAIR data (Module 1) 12:45-2pm : Lunch+coffee 2pm-3:45: Data Processing (Module 2) 3:45-4pm: coffee break 4pm-5:45pm: Statistics for reproducible analyses (Module 3) 5:45-6:15: Questions and answers and feedback session Saturday November 11th: 9am-12pm: The Re-executable Micro Publication Challenge During this time, we will propose a small challenge around producing an entirely re-executable neuroimaging analysis from fetching data to producing statistical results. This will also be a time with close interactions with neuroimaging experts in data handling and analysis. 12pm-12:30: Closing session: feedback and future: "become a trainer". Weekly online office hours will be held prior to the workshop. Registered attendees will receive information via email. *Modules:* *Module 0 - Reproducibility Basics:* Friday Nov. 10. 9am-10:45am. This module guides through three somewhat independent topics, which are in the heart of establishing and efficiently using common generic resources: command line shell, version control systems (for code and data), and distribution package managers. Gaining additional skills in any of those topics could help you to not only become more efficient in your day-to-day research activities, but also would lay foundation in establishing habits to make your work actually more reproducible. *Module 1 - FAIR Data:* Friday Nov. 10. 11am-12:45. This module provides an overview of strategies for making research outputs available through the web, with an emphasis on data. It introduces concepts such persistent identifiers, linked data, the semantic web and the FAIR principles. It is designed for those with little to no familiarity with these concepts. More technical discussions can be found in the reference materials. *Module 2 - Data Processing:* Friday Nov. 10. 2pm-3:45pm. This module teaches you to perform reproducible analysis, how to preserve the information, and how to share data and code with others. We will show an example framework for reproducible analysis, how to annotate, harmonize, clean, and version brain imaging data, how to create and maintain reproducible computational environments for analysis and use dataflow tools to capture provenance and perform efficient analyses (docker) and other tools. *Module 3 - Statistics:* Friday 4am-5:45 The goal of this module is to teach brain imagers about the statistical aspects of reproducibility. This module should give you a critical eye on most of the current literature and the knowledge to do solid work, understand exactly what is a p-value and its limitation to represent evidence for results, practical notion of power and associated tools, etc. *Instructors:* J. Bates, S. Ghosh, J. Grethe, Y. Halchenko, C. Haselgrove, S. Hodge, D. Jarecka, D. Keator, D. Kennedy, M. Martone, N. Nichols, S. Padhy, JB Poline, N. Preuss, M. Travers *This workshop is brought to you by ReproNim: A center for Reproducible Neuroimaging **Computation NIH-NIBIB P41 EB019936*

1 0

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