Freesurfer

freesurfer@nmr.mgh.harvard.edu

19 participants
30088 discussions

qdec contrast and data extraction
by Maximo, Jose Omar 19 Sep '19

19 Sep '19

External Email - Use Caution Hi, I plotted the data from the *.y.ocn.dat file and the graph still shows no apparent significant difference between the groups (see attached pic). Then I found a post with the following: the ocn.dat files have data that is uncorrected in that sense and might need to nuisance factors removed before plotting. There is a design matrix in there (Xg.dat). You can load that into matlab along with the ocn.dat, compute beta = inv(X'*X)*(X'*ocn) to get the betas. You can then compute yhat = X2*beta2 where X2 has nuisance columns removed and beta2 has the same nuisance coefficients removed, then treat yhat as your data to be plotted. I tried that since I have 2 nuisance factors (age and TICV) and want to plot my significant cluster w/o any nuisance effects. This is what I get 1.8973 1.8973 1.8973 … 3.4728 3.4728 3.4728 … Are these the mean averages for each group (cortical thickness). If so, is there a way to get a score for each subject? Best, Omar [cid:image001.png@01D568A6.6E149480] >Look in the *.y.ocn.dat file. > > From mri_glmfit-sim --help > >csdbase.y.ocn.dat - this is a summary of the input (y) over each >cluster. It has a column for each cluster. Each row is a subject. The >value is the average of the input (y) in each cluster. This is a >simple text file. > > > >On 9/3/2019 5:16 PM, Maximo, Jose Omar wrote: > External Email - Use Caution > > Hi, > > Which specific file should I load? I see cluster.mgh, cluster.summary, sign.ocn.annot, sig.ocn.mgh, sig.vertex.mgh, and pdf.dat. > > How can I extract the values from fsaverage space? > > Basically, what is the correct way to extract values from these significant clusters? > > Many thanks, > Omar > > Date: Tue, 3 Sep 2019 15:15:02 +0000 > From: "Greve, Douglas N.,Ph.D." <DGREVE at mgh.harvard.edu<https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer>> > Subject: Re: [Freesurfer] qdec contrast and data extraction > To: "freesurfer at nmr.mgh.harvard.edu<https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer>" <freesurfer at nmr.mgh.harvard.edu<https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer>> > Message-ID: <2286f9bf-37a4-1be9-2252-0bccc833a218 at mgh.harvard.edu<https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer>> > Content-Type: text/plain; charset="utf-8" > > You might have done something wrong along the way. When you run the montecarlo correction, it will create a file with the thickness values in it for each cluster. the first thing to do is to load that and see if you see the expected differences. The other thing is to not go back into native space to extract the numbers. There are several operations that happen as it moves into fsaverage space and in preparation for group analysis (interpolation, and smoothing); sometimes these make a big difference. if the ROI is small, it may not map accurately back into the native space (and you should not need to draw it in the first place) > > On 8/30/2019 11:36 AM, Maximo, Jose Omar wrote: > > External Email - Use Caution > Hi, > > I have a question: > > My design is 2 groups (HC and Patients in that respective order) and 2 nuisance factors (age and eTIV). When I look at the average volume difference between the 2 groups, I get blue and red clusters. I presume the color coding is where each group is greater than the other (Blue = Patients > HC and Red = HC > Patients). > > Then, I processed to extract individual values from each cluster in order to plot them. When I extract data from the blue clusters and plot them, the two groups show no difference in thickness at all, whereas when I look at volume, HC show more than patients in blue clusters (see attached figure). I would assume that both figures would show patients > HC based on the negative statistic. > > Am I interpreting the colors wrong? Or am I doing something wrong? > > These are my steps 1) After applying montecarlo correction, I drew my ROIs to extract the data from; 2) map it onto every single subject; and then 3) used mris_anatomical_stats to extract the data from each subject. > > Any suggestions are welcome. > > Best, > Omar > > Jose O. Maximo, Ph.D. | Postdoctoral Fellow > Department of Psychiatry & Behavioral Neurobiology > UAB School of Medicine > University of Alabama at Birmingham > Cell Phone: (619) 252-3492 > Email: jomaximo at uabmc.edu<https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer><mailto:jomaximo at uabmc.edu<https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer>>

2 3

FS6.0 on 2yo subjects
by Alexopoulos, Dimitrios 18 Sep '19

18 Sep '19

External Email - Use Caution We have acquired 0.8mm isotropic T1-mprage sequences on 2 year old subjects. Although the output looks good (see below), is it recommended to use FS6.0 on a young cohort? [cid:image005.jpg@01D56E40.7241E0E0][cid:image006.jpg@01D56E40.7241E0E0] [cid:image007.jpg@01D56E40.7241E0E0] ________________________________ The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.

3 2

Research Assistant Position (infant/pediatric MR imaging) Boston Children's Hospital/Harvard Medical School
by Gaab, Nadine 18 Sep '19

18 Sep '19

The Gaab Lab (www.gaablab.com) is looking for a Clinical Research Assistant II to assist with all study phases of functional and structural MRI for our longitudinal infant MRI study. This position is ideal for anyone considering future graduate study in cognitive (developmental) neuroscience, computer science, cognitive science or neuroscience. This Clinical Research Assistant II will be responsible for: Subject recruitment for pediatric research studies which will include close interaction with schools and daycare centers in the Boston area. Scheduling of pediatric (infant) research subjects. Maintaining of database, implementing and maintaining of analysis software; pediatric testing (including MRI/fMRI scanning of young children, infants, and their parents) and data analyses (psychometrics, psychophysics and functional magnetic imaging), stimuli design and administrative work (e.g., preparation of internal review board proposals). Under general supervision coordinating activities of daily operations of clinical research studies. Screening, consenting, recruiting, and selecting patients for studies. Preparing IRB and/or CCI documentation and submissions in conjunction with the Principal Investigator and/or other members of the research team. Conducting data collection activities and preparing analytical reports regarding results of studies. Assisting in writing articles and summary papers for various publications. To qualify, you must have: A Bachelor's Degree in computer science, psychology, cognitive science, neuroscience, math, or related field required and one year of relevant work experience. Comfort with a heterogeneous computing environment (Linux/Unix, Mac, and Windows) as well as some programming experience preferred; knowledge of psychological experiment software (E Prime, Presentation, etc.) and/or fMRI analyses software helpful. Demonstrable pre-existing interest in cognitive neuroscience desired; knowledge of neuroanatomy or statistics helpful. Prior experience working with infants and small children (e.g. in a preschool or daycare setting or within a research project) will be a plus. Must be self-motivated and able to work in a fast-paced, changing environment and must like working with children. Boston Children’s Hospital offers competitive compensation and unmatched benefits, including a rotating days/evenings and week-end schedule, affordable health, vision and dental insurance, generous levels of time off, 403(b) Retirement Savings plan, Pension, Tuition Reimbursement, cell phone plan discounts and discounted rates on T-passes (50% off). Boston Children’s Hospital is an Equal Opportunity / Affirmative Action Employer. Qualified applicants will receive consideration for employment without regard to their race, color, religion, national origin, sex, sexual orientation, gender identity, protected veteran status or disability. Please apply through the Boston Children’s Hospital job portal. The ReqID is 52838BR: http://www.childrenshospital.org/career-opportunities "Nothing in science has any value to society if it is not communicated." Anne Roe (1904-1991) Nadine Gaab, PhD Associate Professor of Pediatrics Boston Children's Hospital/Harvard Medical School Department of Medicine/Division of Developmental Medicine Laboratories of Cognitive Neuroscience Mail stop code: BCH3178 1 Autumn Street (Office 643); Boston, MA 02115 nadine.gaab(a)childrens.harvard.edu<mailto:nadine.gaab@childrens.harvard.edu> phone: 857-218-3021 www.gaablab.com<http://www.gaablab.com/> www.bostonearlyliteracyscreener.com<http://www.bostonearlyliteracyscreener.com/> Twitter: Nadine Gaab (@GaabLab)

1 0

mkcontrast-sess question
by Nasiriavanaki, Zahra 18 Sep '19

18 Sep '19

Dear Freesurfer experts Hi I have a question about weighting different contrasts using mkcontrast-sess commmand. I have single subject functional data with 18 different contrasts, and I am willing to weight the different contrasts as (for example) ascending linear. Something like: mkcontrast-sess -an $analysis -a 1 -a 2 -a 3 -a 4 -a 10 -a 11 -a 12 -a 13 -a 6 -a 7 -a 8 -a 9 -a 15 -a 16 -a 17 -a 18 In which contrast number two weights twice as contrast number one and contrast number 3 weights twice as contrast number 2 and so on. Is there any way I can do such a thing with mkcontrast-sess? or do you have any suggestion for such an analysis? Thanks Mona Zahra (Mona) Nasiriavanaki Postdoctoral Research Fellow Martinos Center for Biomedical Imaging Massachusetts General Hospital 149 13th Street, 149-2615 Charlestown, MA, USA, 02129

2 3

GNU R package to read and write FreeSurfer data
by Tim Schäfer 18 Sep '19

18 Sep '19

External Email - Use Caution Dear list, I needed some GNU R functions which read and write MGH/MGZ files including all the header information, so I wrote them and also put them into a package to share them. Support for binary 'curv' format morphometry files (like 'surf/lh.area') and annotations (like 'labels/lh.aparc.annot') in also included. The package is on CRAN: https://cran.r-project.org/package=freesurferformats Docs, examples and vignettes are included. Unit tests can be found (with the full source code) on GitHub: https://github.com/dfsp-spirit/freesurferformats Best, Tim -- Dr. Tim Schäfer Postdoc Computational Neuroimaging Department of Child and Adolescent Psychiatry, Psychosomatics and Psychotherapy University Hospital Frankfurt, Goethe University Frankfurt am Main, Germany

1 0

How to fasten recon-all execution time without using CUDA
by Dev vasu 18 Sep '19

18 Sep '19

External Email - Use Caution Dear all, I am using - openmp 30 to run recon-all, i see now change in the execution time period. Could you please highlight me a method to reduce the execution time for recon-all, if possible some parallel script to do that same. Thanks Vasudev

4 3

Names of the segmentation volumes in lh.hippoSfLabels-T1.v10.mgz
by Kai Liu 18 Sep '19

18 Sep '19

External Email - Use Caution Hi FS experts, I am using the module of segmentation of hippocampal subfields. About the output of lh.hippoSfLabels-T1.v10.mgz, could you pls provide the correspondence between the values of segmentation volumes in the image file and the anatomical names in the file of lh.hippoSfVolumes-T1.v10.txt? I am pretty confused because there are totally 12 regions as described in the .txt file, and I was anticipating 12 values corresponding to 12 segmentation volumes in the .mgz file (like aseg). However, I found 14 discontinuous values in the .mgz ranging from 203 to 226. Looking forward to your kind reply. Kai

1 0

mri_convert for Philips dicoms
by Mun, Anna 17 Sep '19

17 Sep '19

Dear FreeSurfer team, I am working with data sent to us from other research site. They use Philips scanner and I can not run mri_convert on the dicoms. Is there other way to mri_convert and reconstruct Philips dicoms. error log is: WARNING: tag SeriesNumber not found WARNING: tag image orientation not found ERROR: GetDICOMInfo(): dcmGetDWIParams() 7 break DICOMRead.c:5228 Ref Series No = 0 ERROR: bits = 0 not supported. Send email to freesurfer(a)nmr.mgh.harvard.edu Thank you in advance! Best, Anna

2 1

Freeview error
by Miguel Ángel Rivas Fernández 17 Sep '19

17 Sep '19

External Email - Use Caution Dear Freesurfer devs, I tried to open the brainmask.mgz file in freeview but I obtained this error: mghRead(/usr/local/freesurfer/subjects/4227/mri/brainmask.mgz, -1): could not open file MRIread failed: Unable to read from /usr/local/freesurfer/subjects/4227/mri/brainmask.mgz mghRead(/usr/local/freesurfer/subjects/4227/mri/brainmask.mgz, -1): could not open file Any suggestion about what could be the problem?, maybe the write permissons of the freesurfer folder? I am using the freesurfer-Linux-centos6_x86_64-stable-pub-v6.0.0-2beb96c version that is installed in a Ubuntu 18.04 virtual machine. It is very strange because few weeks ago everything it was working correctly. Thanks in advance. Best regards, -- *Miguel Ángel Rivas Fernández*

2 6

mris_sphere question
by Lauricella, Michael 17 Sep '19

17 Sep '19

External Email - Use Caution Dear Freesurfer Experts, I'm fairly new to freesurfer, but have a set-up in which I have been able to run many successful recon-all runs. One problem that I have sometimes, maybe about 20% of the time, is that the run will freeze and crash at this step: mris_sphere -rusage /mnt/freesurfer_subjs/SUBJECT_ID/touch/rusage.mris_sphere.lh.dat -seed 1234 ../surf/lh.inflated ../surf/lh.sphere here's the end of the logfile: "== Number of threads available to mris_sphere for OpenMP = 2 == scaling brain by 0.287... MRISunfold() max_passes = 1 ------- tol=5.0e-01, sigma=0.0, host=mgt-i, nav=1024, nbrs=2, l_area=1.000, l_dist=1.000 using quadratic fit line minimization complete_dist_mat 0 rms 0 smooth_averages 0 remove_neg 0 ico_order 0 which_surface 0 target_radius 0.000000 nfields 0 scale 1.000000 desired_rms_height -1.000000 momentum 0.900000 nbhd_size 7 max_nbrs 8 niterations 25 nsurfaces 0 SURFACES 3 flags 0 (0) use curv 0 no sulc 0 no rigid align 0 mris->nsize 2 mris->hemisphere 0 randomSeed 1234 -------------------- mrisRemoveNegativeArea() pass 1: epoch 1 of 3 starting distance error %21.21 pass 1: epoch 2 of 3 starting distance error %20.95 unfolding complete - removing small folds... starting distance error %20.57 removing remaining folds... final distance error %20.59 MRISunfold() return, current seed 1234" Does anyone have any ideas why this might be happening? Maybe this is a computationally intensive step? If I run a few recon-alls concurrently it almost always crashes at this step. The command I run is: recon-all -all -openmp 2 -qcache -nuintensitycor-3T -s SUBJECT_ID -sd /mnt/freesurfer_subjs/ Thank you so much for your help! Best, Michael

2 4

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