Freesurfer

freesurfer@nmr.mgh.harvard.edu

10 participants
30093 discussions

Error in recon-all (Skull Stripping)
by Diana Tordesillas-Gutiérrez 09 Feb '17

09 Feb '17

Dear FreeSurfer developers, I am a new FreeSurfer user and I am trying to run it in a supercomputer for the first time. Recon-all finished with errors (see log attached) during skull stripping. I have searched the archives and only found that it could be a memory problem, but I am unsure about this. The version of freesurfer is: freesurfer-Linux-centos6_x86_64-stable-pub-v6.0.0-2 Thank you for your help, Diana -------------------------------------------------------- Diana Tordesillas Gutiérrez, PhD Neuroimaging Unit IDIVAL IDIVAL Building Avda. Cardenal Herrera Oria s/n 39011 Santander Spain Tel. +34 942202677 diana.tordesillas(a)humv.es diana.tordesillas(a)gmail.com ----------------------------------------------------------

2 2

Re: [Freesurfer] Expert file for recon-all -hires
by Antonin Skoch 09 Feb '17

09 Feb '17

Dear Paul, for the most recent version 6.0 the default is 10. Looking at the help of mris_inflate binary and also to the source code. Antonin Hello Falk Chris and Antonin, The default is 55 https://surfer.nmr.mgh.harvard.edu/fswiki/mris_inflate. Best, PaulOn Thu, Feb 9, 2017 at 10:42 AM, Falk Lüsebrink <falk.luesebr...(a)ovgu.de> wrote: > Dear Chris and Antonin, > > I found in some archived message or in the wiki that the default used to > be 50 iterations at some time in previous releases of FreeSurfer for > standard recon-all. Therefore, I changed it to 50 as default for every > resolution in the recent release. This works fine for example for up to > 0.5x0.5x0.5mm3 data. > > I'd also say, more is better and I have not experienced any drawbacks. > Processing time doesn't seem to increase noticeably, too. > > Best, > Falk > > ------------------------------ > *Von:* freesurfer-boun...(a)nmr.mgh.harvard.edu [freesurfer-boun...(a)nmr.mgh. > harvard.edu]" im Auftrag von "Antonin Skoch [a...(a)ikem.cz] > *Gesendet:* Donnerstag, 9. Februar 2017 16:20 > *An:* freesurfer(a)nmr.mgh.harvard.edu > *Betreff:* Re: [Freesurfer] Expert file for recon-all -hires > > Dear Chris, > > I have found that -n 15 is for our 0.7x0.7x0.7mm3 data not enough. > > I determined optimal value -n 30, empirically, by looking at the shape of > inflated surface with various number of iterations. Over -n 30 there was > almost no further progression in inflation. > > I suppose that better more than less. But maybe experts can correct me or > provide better suggestion. > > Antonin > > Hi all, > > If I'm reading this > <https://surfer.nmr.mgh.harvard.edu/fswiki/SubmillimeterRecon> correctly, > the current best practice for `-hires` is to include an expert file > containing "mris_inflate -n 15", and that this should work for all voxel > sizes (0.75mm)^3 - (1mm)^3? Or do we need to empirically determine the best > value for a given voxel size, and if so, what should be the criteria for > making this determination? > > Thanks, > Chris Markiewicz >

2 1

Full-Time Staff Position – Johns Hopkins University
by Arnold Bakker 09 Feb '17

09 Feb '17

Computational Neuroscience Research Position The Division of Psychiatric Neuroimaging in the Department of Psychiatry and Behavioral Sciences at Johns Hopkins University School of Medicine, has an opening for talented candidates with strong computational and programming skills. The successful candidate will assist the investigative team with a wide range of studies including those of Alzheimer’s disease, Parkinson’s disease, obsessive compulsive disorder, ataxia, obesity, feeding behavior and depression. Responsibilities will include acquisition and analysis of functional and structural brain imaging data, quality control of imaging data, maintenance of databases for statistical analysis, statistical analyses and assistance with manuscript preparation. A minimum commitment of two years is required. Please email a CV, a statement of research interests and career goals, and contact information for two or more references to: pni(a)jhu.edu. Qualifications Bachelors or Master’s of Science degree required. Strong background in biomedical engineering, computer science, physics, mathematics, statistics, computational neuroscience, or a related field is required. Experience with Matlab, Python, R, signal and image processing, statistics, databases, Linux and shell scripting is required. Experience in neuroimaging/neural data analysis using SPM or similar software is strongly preferred. Excellent organizational and interpersonal communication skills are required. Successful candidate must have great attention to detail and be able to work well independently as well as in a team.

1 0

Re: [Freesurfer] Expert file for recon-all -hires
by Antonin Skoch 09 Feb '17

09 Feb '17

Dear Chris, I have found that -n 15 is for our 0.7x0.7x0.7mm3 data not enough. I determined optimal value -n 30, empirically, by looking at the shape of inflated surface with various number of iterations. Over -n 30 there was almost no further progression in inflation. I suppose that better more than less. But maybe experts can correct me or provide better suggestion. Antonin Hi all, If I'm reading this < https://surfer.nmr.mgh.harvard.edu/fswiki/SubmillimeterRecon> correctly, the current best practice for `-hires` is to include an expert file containing "mris_inflate -n 15", and that this should work for all voxel sizes (0.75mm)^3 - (1mm)^3? Or do we need to empirically determine the best value for a given voxel size, and if so, what should be the criteria for making this determination?Thanks, Chris Markiewicz

3 2

Expert file for recon-all -hires
by Christopher Markiewicz 09 Feb '17

09 Feb '17

Hi all, If I'm reading this < https://surfer.nmr.mgh.harvard.edu/fswiki/SubmillimeterRecon> correctly, the current best practice for `-hires` is to include an expert file containing "mris_inflate -n 15", and that this should work for all voxel sizes (0.75mm)^3 - (1mm)^3? Or do we need to empirically determine the best value for a given voxel size, and if so, what should be the criteria for making this determination? Thanks, Chris Markiewicz

2 1

Edits to Aseg in the Longitudinal Stream
by Tamara Tavares 09 Feb '17

09 Feb '17

Hello, After I make edits to my aseg output in the longitudinal stream, do I just run the following command to recreate the final volumes: recon-long -autorecon2-noaseg -subjid <Subj_ID> I assume the -subjid is the name of the edited time point and <Subj_ID> is the name of the recreated volumes? Thank you, Tamara

2 1

Longitudinal stream with v6 but cross with v5.3
by Matthieu Vanhoutte 09 Feb '17

09 Feb '17

Dear FS's experts, I have run all my cross with v5.3 and hesitate to run all longitudinal process with v6. Does that make sense and will improve the process or would it be risky to do this ? Many thanks for your lights ! Best regards, Matthieu

2 2

Parallelization implementation in Freesurfer 6.0
by Francis Tyson Thomas 09 Feb '17

09 Feb '17

Hello Freesurfer team! Having gone through the instructions for making use of the parallelization in Freesurfer v6, I'm confused as to how exactly the calls for fine grained and coarse grained paralleization differ. As per the release notes, following is stated - *Parallelization: a new flag was introduced which enables two forms of compute parallelization that significantly reduces the runtime. As a point of reference, using a new-ish workstation (2015+), the recon-all -all runtime is just under 3 hours. When the -parallel flag is specified at the end of the recon-all command-line, it will enable 'fine-grained' parallelized code, making use of OpenMP, embedded in many of the binaries, namely affecting mri_em_register and mri_ca_register. By default, it instructs the binaries to use 4 processors (cores), meaning, 4 threads will run in parallel in some operations (manifested in 'top' by mri_ca_register, for example, showing 400% CPU utilization). This can be overridden by including the flag -openmp <num> after -parallel, where <num> is the number of processors you'd like to use (ex. 8 if you have an 8 core machine). Note that this parallelization was introduced in v5.3, but many new routines were OpenMP-parallelized in v6. The other form of parallelization, a 'coarse' form, enabled when the -parallel flag is specified, is such that during the stages where left and right hemispheric data is processed, each hemi binary is run separately (and in parallel, manifesting itself in 'top' as two instances of mris_sphere, for example). Note that a couple of the hemi stages (eg. mris_sphere) make use of a tiny amount of OpenMP code, which means that for brief periods, as many as 8 cores are utilized (2 binaries running code that each make use of 4 threads). In general, though, a 4 core machine can easily handle those periods. Be aware that if you enable this -parallel flag on instances of recon-all running through a job scheduler (like a cluster), it may not make your System Administrator happy if you do not pre-allocate a sufficient number of cores for your job, as you will be taking cycles from other cores that may be running jobs belonging to other cluster users.* Does this mean a command executed as "recon-all -s vol0 -i junk/orig/001/IM-0001-0001.dcm -all -parallel" would be a coarse grained parallelization and the command executed as "recon-all -s vol0 -i junk/orig/001/IM-0001-0001.dcm -all -parallel -openmp 8" be a fine grained parallelization? Thanks, Tyson

1 0

recon-all -autorecon2-cp -autorecon-wm -autorecon-pial issue
by 冉倩 08 Feb '17

08 Feb '17

Hello Freesurfer Developers, I have had last September Freesurfer course. When I operated my data following that stream I met some problem in reon-all -all step. Because I need put some screenshout to exaplain that problems, I add word document in attachment. Thanks very much! Best Regards, Qian Ran

3 15

ColorLut hippocampus subfields
by Arnaud Boré 08 Feb '17

08 Feb '17

Dear Freesurfer developpers, Is there a Freesurfer ColorLut that includes the new subfields of the hippocampus ? In the current colorlut there is no left/right for these subfields: parasubiculum, GC-DG, HATA, molecular layer and tail. In addition, I think it would be easier if the correct left/right labels would be used for the segmentation. Thank you in advance. Arnaud Boré

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