Freesurfer

freesurfer@nmr.mgh.harvard.edu

4 participants
30076 discussions

Freesurfer mri_convert
by Zoe O'Brien-Moran 27 Mar '22

27 Mar '22

External Email - Use Caution Dear FS Experts I am having continued issues with the mri_normalize function on MPRAGE data I get an error message: 3d normalization pass 1 of 2 error: No such file or directory error: MRInormFindControlPoints: could not find enough control points error: No such file or directory error: MRInormFindControlPoints failed Can you advise on any reason this may be happening?

2 1

Permission denied error from -localGI
by Billah, Tashrif 27 Mar '22

27 Mar '22

It does not exist and cannot be created in a shared cluster. Is there a way to define a different folder for /scratch/tp24093955057547375.load_mgh.m.mgh? =============== Douglas N. Greve Wed, 09 Mar 2022 10:04:31 -0800 Do you have permission to /scratch? Does it exist? =============== Billah, Tashrif Wed, 02 Mar 2022 06:50:41 -0800 Hi all, We are trying to run recon-all with -localGI but running into an unknown permission denied error: $ recon-all -s sub-1001 -localGI ``` make_outer_surface('/path/to/FS_SUBJECTS_DIR/sub-1001/surf/tmp-mris_compute_lgi-lh.pial/lh.pial.filled.mgz',15,'/path/to/FS_SUBJECTS_DIR/sub-1001/surf/tmp-mris_compute_lgi-lh.pial/lh.pial-outer'); exit ================= < M A T L A B (R) > Copyright 1984-2017 The MathWorks, Inc. R2017a (9.2.0.556344) 64-bit (glnxa64) March 27, 2017 >> reading filled volume... /bin/bash: /scratch/tp24093955057547375.load_mgh.m.mgh: Permission denied ERROR: could not open /scratch/tp24093955057547375.load_mgh.m.mgh for reading ERROR: loading /path/to/FS_SUBJECTS_DIR/sub-1001/surf/tmp-mris_compute_lgi-lh.pial/lh.pial.filled.mgz as MGH Struct contents reference from a non-struct array object. Error in make_outer_surface (line 25) volume=vol.vol; >> ERROR: make_outer_surface did not create output file '/path/to/FS_SUBJECTS_DIR/sub-1001/surf/tmp-mris_compute_lgi-lh.pial/lh.pial-outer'! recon-all -s sub-1001 exited with ERRORS at Wed Mar 2 09:02:21 EST 2022 ``` Our MATLAB environment is properly setup and we regularly run other FreeSurfer-MATLAB commands. So I do not think that is causing the issue. Best, Tashrif

2 1

vertexwise correlation between the two maps and atlases in volume (MNI) to surface
by stdp82＠virgilio.it 27 Mar '22

27 Mar '22

External Email - Use Caution Hi list, a) I have two maps, one's of rs-fMRI functional connectivity (by FS-FAST) and one's of receptor density (from a PET-Atlas). Both the maps are in surface space. How can I measure the vertexwise correlation between the two maps? b) How can I move some atlases in volume (MNI) to surface? I have tried to use: mri_vol2surf --src ATLAS_vol.nii --regheader fsaverage --projfrac 0.5 --hemi lh --out ATLAS_surface.nii but the results are inconsistent. Thanks Stefano

2 1

Turning Significant Cluster (sig.cluster.summary or sig.mgh) into binary .nii ROI
by Emily Popa 27 Mar '22

27 Mar '22

External Email - Use Caution Hello, I have been struggling to figure out a way to get the significant results of a freesurfer group analysis (just one significant cluster) from either the sig.cluster.summary or sig.mgh file into a binarized ROI in .nii format. I used mri_glmfit to perform the cortical thickness analysis on sMRI data and found the significant cluster only in the left hemisphere. I tried to follow some instructions I found on the FreeSurfer site, however, at the first step/line below (mri_volsynth) the –delta-crsf input was referred to on the site as the vertex number, but on the –help documentation, it was specified that this input should be: col row slice frame : 0-based. Looking at the cluster summary file, for the significant cluster there is a maximum vertex, but I am not sure if this is the vertex number I should be using. mri_volsynth --template $SUBJECTS_DIR/fsaverage/surf/lh.thickness --pdf delta --delta-crsf 5 9 5 1 --o delta.sm00.mgh mris_fwhm --smooth-only --subject fsaverage --hemi lh --i delta.sm00.mgh --niters 10 --o delta.sm10.mgh mri_binarize --i delta.sm10.mgh --min 10e-5 --o delta.sm10.bin.mgh When trying the first line above for mri_volsynth, I get this error: Segmentation fault: 11 (so I think my input –deta-crsf is likely wrong). If there is a better way to convert this ROI to a mask, please let me know! To give additional context, I am hoping to get a binary mask of this ROI that I can use as a mask for an ASL analysis in FSL (to limit the analysis to this ROI). I plan to do this by registering the binary ROI mask to the FSL standard group template (by using the transformation matrix from registering the fsaverage template to FSL’s standard group template first), and then getting the ROI mask from FSL’s group template space into native ASL space. If this plan sounds reasonable to you, it would also be very helpful to know. I really appreciate your time and help and look forward to hearing back! Best, Emily S. Popa, M.S. Staff Research Associate II, Neuroimaging Pacific Brain Health Center | Pacific Neuroscience Institute Foundation | Providence Saint John’s Health Center 1301 20th St. #250 Santa Monica, CA. 90404 (408) 750-7971 (M) [cid3519*image001.jpg(a)01D7EC78.96AA1AC0]<https://secure-web.cisco.com/1hLjiH2EMqrNb1Aj_EN8nMGnwBm1r3goPYCE5vd4lR-hGE…> [cid3519*image002.jpg(a)01D7EC78.96AA1AC0] <https://secure-web.cisco.com/11wThjE0wBKB7uH7Mdk8M0tt4s97K_cFB0xuOQfaRwID0m…> [cid3519*image003.jpg(a)01D7EC78.96AA1AC0] <https://secure-web.cisco.com/1oGIwvc1noJjNXOvlfvxCGrBAHvKAjS0PHn5htsutbjwTJ…> [cid3519*image004.jpg(a)01D7EC78.96AA1AC0] <https://secure-web.cisco.com/1cMaCmazMIVCk8v3EhpuiTkZaPsebaExskfX49kF6WWp7P…> [cid3519*image005.jpg(a)01D7EC78.96AA1AC0] <https://secure-web.cisco.com/17e7d2R-5gmmayIc-Okzb6HMA5_U4NT4cVQtWz_nLOjiv5…>

2 2

Question about calculation of pial area in DKatlas
by Dongyue Zhou 26 Mar '22

26 Mar '22

External Email - Use Caution Dear FreeSurfer experts: I want to calculate mean pial surface area in DKatlas in individual space.And I know the function read_annotation.m in $FREESURFER_HOME/matlab for loading this atlas.And if I can load sub/surf/?h.area.pial as the area in each vertice and calculate mean pial area by dividing sum PSA of all vertices in one ROI by the number of vertices in this ROI?But why my result is not the same as result in sub/stats/?h.aparc.stats ? Does the column named surface area represent mean PSA in each ROI?Bests,Dongyue

2 1

fMRI-EEG Postdoctoral Position at Albert Einstein COM in NYC
by Mark Wagshul 25 Mar '22

25 Mar '22

External Email - Use Caution Drs. Elyse Sussman and Mark Wagshul, of the Departments of Neuroscience and Radiology at the Albert Einstein College of Medicine in NY, are seeking a full-time postdoctoral research associate to work on an NIH-funded project exploring the neural bases of chemotherapy-induced cognitive deficits in survivors of childhood leukemia. The project uses EEG- and fMRI-based measures of cognitive function to understand the treatment-related deficits in children ages 5-12 years. The position is available starting May 2022; the position is for the duration of the project (March 2025), renewed annually based on performance and funding. Our lab combines advanced event-related potential EEG and functional MRI brain imaging methods to understand cognitive function, and how it is impacted by treatment for cancer. In particular, we use EEG- and MRI-based measures of cognitive tasks to study brain pathways affected by cancer treatment, comparing these measures with standard neuropsychological testing. By combining the superior temporal resolution of EEG with the millimeter spatial resolution of fMRI, along with advanced techniques for extracting effective connectivity (e.g. neural dynamics and causality), we aim to identify the specific cognitive domains and related brain pathways altered by cancer treatment. These studies will ultimately lead to objective measures of brain function which can drive development of better therapy modalities which do not impact cognitive skills in the way that current therapies do. Successful candidates for the position will have a PhD in physics, biomedical engineering, neuroscience or a related discipline, a strong background in either cognitive neuroscience or MRI physics, expertise in image and EEG data processing, as well as data and statistical analysis. Programming experience should include the commonly used packages for image and EEG processing, such as C++, Matlab, and FSL. The candidate should have significant experience in EEG/ERP data collection and analysis methods and/or imaging in a clinical setting and a strong background in the development and/or application of novel methods for human brain imaging, such as DTI/fMRI/structural MRI data analyses. Candidates with experience only in EEG-based or fMRI-based methods will be considered. The candidate will allocate time to both the EEG and MRI labs. Prior experience with auditory-based ERP/fMRI is not necessary, but candidates with such experience will be given preference. A strong work ethic, excellent organizational and communication skills, and ability to work independently is expected. Previous experience working with children in a research setting is a plus. The candidate will be expected to be involved in all aspects of the project, including EEG, image and data analysis, statistical analysis, as well as interfacing with the clinical team. Competitive salary and benefits will be offered based on qualifications. The Gruss MRRC is located on the main campus of the Albert Einstein College of Medicine in Bronx, NY. The Center has a multinuclear 3.0 Tesla Whole Body Philips Elition human system, along with on-site support from a Philips Clinical Scientist, and a 9.4 T Direct Drive Varian animal system. Image processing facilities including a computing cluster, multiple Linux workstations and PCs with imaging and general software installed. Please visit our website for more details (http://secure-web.cisco.com/1pGDF8ABU3mUM3q0mm1bnhU-FyjF46OfHfxWPLuYsMKf3FT…) The Albert Einstein College of Medicine is a leading academic institution, with over $170M in NIH funding. The candidate will have access to state-of-the-art brain imaging (MRI, EEG) and computing facilities. Einstein is one of the nation's premier institutions for medical education, basic research, and clinical investigation. It is home to some 2,000 faculty members, 750 M.D. students, 350 Ph.D. students attending the Sue Golding Graduate Division - including 125 in the combined M.D./Ph.D. program - and 325 postdoctoral investigators training at our Belfer Institute for Advanced Biomedical Studies. More than 7,000 Einstein alumni are among the nation's foremost clinicians, biomedical scientists, and medical educators. Interested candidates should send their CV, a brief statement of their research interests and career goals, and the names and phone numbers of three references to Mark Wagshul (mark.wagshul(a)einsteinmed.edu<mailto:mark.wagshul@einsteinmed.edu>) or Elyse Sussman (elyse.sussman(a)einsteinmed.edu<mailto:elyse.sussman@einsteinmed.edu>). Review of applications will begin immediately and continue until the position is filled. Mark ____________________ Mark Wagshul, PhD Associate Professor Gruss Magnetic Resonance Research Center Albert Einstein College of Medicine Bronx, NY 10461 Ph: 718-430-4011 FAX: 718-430-3399 Email: mark.wagshul(a)einsteinmed.org<mailto:mark.wagshul@einsteinmed.org> [einstein-logo-rgb] This email message and any accompanying attachments may contain privileged information intended only for the named recipient(s). If you are not the intended recipient(s), you are hereby notified that the dissemination, distribution, and or copying of this message is strictly prohibited. If you receive this message in error, or are not the named recipient(s), please notify the sender at the email address above, delete this email from your computer, and destroy any copies in any form immediately.

1 0

Curvature differences between 7T and 3T images
by Samuel Paré 24 Mar '22

24 Mar '22

External Email - Use Caution Hello, We are looking at some of our data acquired on a 7T magnet. We ran recon-all with the “highres” flag. The obtained results for cortical thickness and cortical volume are very comparable to data that we obtain when using 3 T brain images. However, the results for cortical curvature are significantly different from what we get by our 3 T data. I attach the output of mean cortical curvature values using the Destrieux atlas for a 3T and 7T dataset on healthy controls. As you can notice, they differ roughly by a factor of 2. Any explanation for what may cause this large difference? Thanks, Samuel [image: image.png] -- *Samuel Paré, MSc* Laboratoire du développement et plasticité du système visuel École d'optométrie Université de Montréal

1 0

Updated Ubuntu 18 VM image with Freesurfer 7.2.0 release pre-installed
by fsbuild 23 Mar '22

23 Mar '22

External Email - Use Caution The Ubuntu 18 VM image with the terminal shell environment set out of the box to use the pre-installed freesurfer 7.2.0 release - is back online.  While the functionality of the release is unchanged, the new image is:- named with 7_2_0 to distinguish it from future releases- a bit smaller in size and may run faster depending upon your hardware- expanded a bit differently Please see the wiki page for details, https://secure-web.cisco.com/1yIC-PbRjEF1VSkLF8fML6h4gueEGDMqWNEbYP-pGBkGm0… Happy to address any inquiries. - R.

1 0

mri_convert: possible bug converting float to int
by Gesierich, Benno Gerhard Dr. 22 Mar '22

22 Mar '22

External Email - Use Caution Hello FreeSurfer Developers, I converted a label map (<in volume>), given as Nifti image of type float32 and located in Freesurfer conform space, back to native space using the native T1 in Nifti format and with type int16 as reference <reference volume> . During conversion, the datatype was automatically converted to that of the reference T1. Unfortunately, the labels changed during this conversion, probably because of applying a scaling (slope), which is not saved correctly to the Nifti header. I come to this conclusion, because the maximum label in the output has the maximum value, possible with int16 datatype, i.e. 32767 Here the exact command I used: mri_convert -rt nearest -rl <reference volume> <in volume> <out volume> The problem of changed labels in the output can be avoided by using the —no change option, which keeps the datatype of the input: mri_convert --nochange -rt nearest -rl <reference volume> <in volume> <out volume> I used Fresurfer 7.2.0 Best Regards, Benno Gesierich

1 0

TRACULA longitudinal - weird tracts
by Donovan Roediger 21 Mar '22

21 Mar '22

External Email - Use Caution Hello Anastasia, I ran TRACULA (7.2.0) twice on the same dataset, once as cross-sectional and once as longitudinal. The cross sectional tracts look OK, but the longitudinal output looks very bad for every subject. Below are summaries of my workflows along with links to screenshots of the resulting tracts. Can you tell what I'm doing to mess up the longitudinal stream? Cross sectional TRACULA: 1) Structural and diffusion data preprocessed using HCP 4.0.1 (which includes cross-sectional recon-all with FS 6.0) 2) Thalamic segmentation was run using FreeSurfer 7.2.0. 3) Cross sectional TRACULA (7.2.0) was run using the HCP-cleaned diffusion data. Default settings used except doeddy=0 because eddy was run as part of HCP preprocessing. IMAGE: https://secure-web.cisco.com/13BZ6YlVnYcAOxNoC1FBf0K8tfLvVUgG_XzJ51MPZJ5fE6… Longitudinal TRACULA: 1) Structural and diffusion data preprocessed using HCP 4.0.1 (which included cross-sectional recon-all with FS 6.0). 2) Longitudinal recon-all was run under FreeSurfer 6.0 starting with the HCP FreeSurfer cross-sectional outputs. 3) Thalamic segmentation was run on the ".long" subjects using FreeSurfer 7.2.0. 4) Longitudinal TRACULA (7.2.0) was run using the HCP-cleaned diffusion data. Default settings used except doeddy=0 because eddy was run as part of HCP preprocessing. IMAGE: https://secure-web.cisco.com/1us4Dr6uNFzuOYt72IbrjF1_SvArR22k16JUpH4ZBPgF2W… Thanks so much! Donovan -- Donovan Roediger Neuroimaging Analyst | Psychiatry | psychiatry.umn.edu University of Minnesota Medical School | med.umn.edu | 612-626-5064 staff bio <https://secure-web.cisco.com/1IdZ5gFMBKeiXOXZZl8u_Lmu2Itqgjp8vqJbAswnHrkh98…> | he, him, his

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