Freesurfer

freesurfer@nmr.mgh.harvard.edu

16 participants
30084 discussions

Important 8.2.0 release updates for arm64 Mac machines
by fsbuild 05 Jun '26

05 Jun '26

External Email - Use Caution If you are a Mac user with a silicon (arm64) machine with an M series chip (M1 … M4) running the 8.2.0 Freesurfer release, then please apply the latest updates by downloading the 8.2.0 patch release script and running it as described here,https://secure-web.cisco.com/17cvaN4h3NMsl7K7bR6zvjEQxNQNiLWoGZ7144oqZ… This should provide a working/updated version of SynthSeg, e..g., which is used by recon-all.  Please be sure to accept all the requests to update files. This update only applies to the 8.2.0 release and does not apply to Linux users of Mac users with older hardware using Intel processors. - R.

1 0

bringing longitudinal pipeline results back to native T1w space
by Yakupov, Renat /DZNE 04 Jun '26

04 Jun '26

External Email - Use Caution Good day all, I'd like to revisit the topic of bringing the segmentation and parcellation results of the longitudinal processing back to native T1w image space. I've used the suggested commands from this wiki https://secure-web.cisco.com/12JK6OH3yluIisG2SSs-MiTTRFfsBTwA7OHLDD0T4LKePO… with a small modification of using transformation files from <baseid>/mri/transforms directory (since they are absent in <tpid>mri/transforms directory). However, the output *-in-rawavg fies still have 256x256x256 resolution, rather than the native T1w resolution. How can I get the aseg and aparc results in native T1w space and resolution? Thank you in advance! Best regards, Renat.

1 0

Re: comparison of FS v 8.0.0 with earlier versions
by fsbuild 02 Jun '26

02 Jun '26

External Email - Use Caution Starting with 8.0.0, the recon-all pipeline was changed by default to use python code with deep learning algorithms via use of the SynthSeg, SynthStrip and SynthMorph commands.  This change should reduce the time needed to run a recon-all on a single CPU from about 8 hours down to ~2.5 hours.  This change however requires more RAM, e.g., at least 32G total system RAM; that may not be enough for recon-all running on some laptops including machines running MacOS.  Without enough memory, the OS will likely kill the recon-all process causing it to prematurely exit, and you will get some kind of system/fault error in the recon-all.log file. You can read about all the changes in the release notes page for 8.2.0, 8.1.0 and 8.0.0 (in that order from the top and then scrolling on down),https://secure-web.cisco.com/1D1wxOT0U-uFIizZOp_WPrw9ovGJ3pBOyXu3NR73… Releases can be downloaded from,https://secure-web.cisco.com/1J_8lzKD8I8vBHFQBalZC_mEPr3j-YcAz2NXavu9u… If you are running on Linux, I would start with the current 8.2.0 release as that will have the most recent bug fixes. There is currently an issue with SynthSeg on arm64 Macs, so for darwin_arm64 try the 8.1.0 release instead.   If you install 8.2.0 you should apply the latest patch updates as described under,https://secure-web.cisco.com/1o1ymBvn7L-cM9TppcvD--YG4ZaDPL4Dw4P1yr_8… I think it still works to revert your installed 8.x.x release back to the old recon-all pipeline behavior by passing the argument --no-v8  on the recon-all command line.  Keep in mind this should increase the recon-all run time back up to however long it used to take on your machine with older Freesurfer releases.  - R. On Jun 2, 2026, at 17:55, CRISTIAN MOJICA <comojica319(a)g.ucla.edu> wrote:        External Email - Use Caution        Hello FreeSurfer experts, Is anyone aware of any resources/studies that compare FreeSurfer 8.0.0 to earlier versions, or can anyone provide their experience with FreeSurfer 8.0.0 compared to earlier versions? I'm primarily curious about recon -all and its performance. Thank you in advance._______________________________________________ Freesurfer mailing list -- freesurfer(a)nmr.mgh.harvard.edu To unsubscribe send an email to freesurfer-leave(a)nmr.mgh.harvard.edu https://secure-web.cisco.com/1CQvmv0hUEWLZ1hizdo4vLfOCdBdp_sOh3Acuu_AGc-w9Q…

1 0

comparison of FS v 8.0.0 with earlier versions
by CRISTIAN MOJICA 02 Jun '26

02 Jun '26

External Email - Use Caution Hello FreeSurfer experts, Is anyone aware of any resources/studies that compare FreeSurfer 8.0.0 to earlier versions, or can anyone provide their experience with FreeSurfer 8.0.0 compared to earlier versions? I'm primarily curious about recon -all and its performance. Thank you in advance.

1 0

3 PhD positions in the BRIDGE-AI Doctoral Network
by Koen Van Leemput 02 Jun '26

02 Jun '26

External Email - Use Caution Are you passionate about pushing the limits of AI in neuroscience, and eager to make a real-world clinical impact? The BRIDGE-AI Doctoral Network is hiring ambitious PhD candidates across Europe to advance trustworthy AI for longitudinal neuroimaging, improving the way neurological diseases are monitored over time.* We are currently inviting applications for the last 3 open PhD positions, located in Germany and Finland*: * Deep collaboration between hospitals, universities & industry * Training events in leading European innovation hubs * Ample clinical neuroimaging data and expert guidance * A competitive salary while you research. Apply by 30 Jun 2026 (but applying earlier increases your chances!) Learn more & apply: https://secure-web.cisco.com/1UvMp3hoZMJb4yFegzuGdbiiNdOoG-BIuUKZp8LYxjdcxR…

1 0

Question on FreeSurfer recon-all workflow for brain tumor MRI
by nooshin Safari 01 Jun '26

01 Jun '26

External Email - Use Caution Hello, We are processing T1-weighted MRI from brain tumor patients and aim to run high-quality FreeSurfer recon-all for morphometric analysis. We are considering this workflow: TumorSynth tumor mask → VBG/virtual-brain correction → recon-all. Could you advise whether this is methodologically sound? Should tumor cases be handled with VBG/lesion filling before recon-all, by passing a mask directly, or another approach? Any recommended flags, QC steps, manual edits, example commands, or references would be appreciated. Best, Nooshin

2 2

Re: Singularity/Apptainer Image for FreeSurfer 8.2.0 on RHEL 8 HPC
by fsbuild 25 May '26

25 May '26

External Email - Use Caution We support Rocky8 Linux now instead of CentOS8/RHEL 8.  The Rocky8 installer may however work on RHEL 8.   There is also a CentOS7 installer for Freesurfer 8.2.0 which could also work on RHEL 8,https://secure-web.cisco.com/1vmYDp5mv23tqWzQ6UYVDN6np38kpfPQa_nY4_VDIYVD… but we no longer support the CentOS7 installer. There is a 8.2.0 docker file and an 8.0.0 singularity file for use with Rocky8 linux listed under,https://secure-web.cisco.com/1otR-Nsj7mSUktWs9KYtmVJiwIwW_YtvtnKZ2IcR… If you want to use something newer than 8.0.0 with that singularity file, you could try editing that exiting 8.0.0 singularity file to instead use the 8.2.0 Rocky8 installer freesurfer-CentOS7-8.2.0-1.x86_64.rpm and likewise change all 8.0.0 entries to 8.2.0 in that file.   You can download the Rocky8 8.2.0 installer from,https://secure-web.cisco.com/14IoQ-fNJng0Jj1BRFCbUBdge8u1ckTL1UZp8lV_w… Alternately if you search for "use dockerfile to create singularity image” you will find various ways to create a docker image from a  docker container and and subsequently use the docker image to create a singularity image. - R On May 17, 2026, at 16:04, nooshin Safari <noonoosh91(a)gmail.com> wrote:

2 1

Postdoc position (MGH): the noradrenergic system in Alzheimer’s disease using cognitive computational modeling and multimodal neuroimaging
by Jacobs, Heidi I.,PHD 22 May '26

22 May '26

Postdoctoral fellow position: Understanding how the noradrenergic system contributes to Alzheimer’s disease using cognitive computational modeling and multimodal neuroimaging at Massachusetts General Hospital The Athinoula A. Martinos Center for Biomedical Imaging, located within the Department of Radiology at Massachusetts General Hospital (MGH) and Harvard Medical School (HMS) in Boston, Massachusetts, has an opening for a highly qualified postdoctoral individual to work with Dr. Heidi Jacobs. The Jacobs lab is focused on improving the early detection and early treatment of Alzheimer’s disease. The lab focuses on the neuromodulatory subcortical nuclei, in particular the locus coeruleus, using a variety of approaches, including 3T MRI, 7T MRI, PET imaging, pupil measurements, physiological recordings, blood-based markers, cognitive assessments and perturbational approaches, such as pharmacological manipulations and non-invasive vagus nerve stimulation. We are seeking a postdoctoral fellow with a background in cognitive computational, and/or biophysical modeling, including frameworks such as predictive coding and reinforcement learning, to investigate the neural mechanisms underlying how the noradrenergic locus coeruleus system contributes to tau progression and cognitive decline. Although these approaches have a strong tradition in psychiatry, they are increasingly recognized as promising tools for understanding Alzheimer's disease: the noradrenergic system has been computationally linked to the regulation of belief updating and exploration/exploitation balance, and given that the noradrenergic locus coeruleus is one of the first systems to undergo neurodegenerative and functional changes in Alzheimer's disease, these frameworks offer a principled and largely untapped lens through which we can study cognitive decline and disease progression. The Jacobs Lab is actively expanding in this direction and is looking for candidates who are eager to bring this expertise, help shape this line of work, and forge collaborations that bridge the computational and clinical neuroscience communities. This position is particularly suited to candidates with demonstrated experience linking such models to in vivo human neuroimaging and who are excited by the opportunity to apply these methods in a neurodegenerative disease context. The project will involve integrating multimodal imaging (e.g., sMRI, fMRI, (f)MRS, PET) with pupillometry and advanced modeling approaches, including dynamic causal modeling (DCM) and reinforcement learning, in observational and pharmacological manipulation studies. In parallel, the lab develops approaches to characterize heterogeneity and temporal trajectories of disease progression using large-scale, well-characterized longitudinal datasets. The lab also evaluates specific monoaminergic PET tracers to characterize age- and Alzheimer's disease-related molecular changes in the noradrenergic and serotonergic systems. The successful candidate will have the opportunity to contribute to these ongoing projects depending on expertise and interests. The Jacobs Lab consists of a multidisciplinary team of two instructors, three postdocs, two PhD students, two clinical research coordinators, and one project manager. The work in this project will be performed at the Athinoula A. Martinos Center for Biomedical Imaging, Charlestown Navy Yard. The candidate will be part of ongoing collaborations and work in a stimulating and inspiring environment. Our research group values open communication, motivation, optimism, mutual respect, teamwork, and innovative thinking. Dr. Jacobs is highly committed to individualized mentoring to help trainees achieve the most out of their postdoctoral candidature and to move forward in their career. At a minimum, the training environment will include professional development, one-on-one meetings, group lab meetings, frequent seminars and journal clubs, opportunities to present their work at national and international scientific conferences (i.e. Human Amyloid Imaging, Alzheimer’s Association International Conference), collaborations with partners within and outside MGH, and manuscript preparation for publication in high-impact journals. Dr. Jacobs has a strong funding record and will strongly encourage and provide mentorship in grant applications. Her trainees have a strong publication record, obtained competitive awards/grants, and many have secured faculty or senior positions. Requirements: Candidates should hold a Ph.D. in Cognitive Neuroscience, Computational Modeling, Bio/medical Physics, Biomedical Engineering, Neuroimaging, or related disciplines. A strong publication record, including at least three first-author peer-reviewed publications with demonstrated research experience in neuroimaging combined with computational modeling is required. Essential qualifications include: strong expertise in fMRI analyses and other neuroimaging methods (e.g. MRS, PET), advanced programming skills (e.g., Python, MATLAB), experience with computational modeling (e.g. cognitive computational modeling), a solid background in signal processing. Candidates must be able or willing to learn to relate neuroimaging findings within a pathobiological framework for neurodegenerative diseases. The candidate should be highly motivated, a strong communicator (verbally and in scientific writing), have an internal drive to learn independently, and be comfortable working as part of a larger collaborative team. A real necessity is a passion for science and a positive attitude. Experience in Alzheimer’s disease/aging/memory-related research would be beneficial. Start date is negotiable. Apply: The successful candidate will have joint appointments at MGH and HMS. If interested, please send your CV, letter describing interests, background, major achievements, skills, goals and contact information for three professional references. Please send application materials to Dr. Heidi Jacobs, Associate Professor of Radiology, at hjacobs(a)mgh.harvard.edu<mailto:hjacobs@mgh.harvard.edu> (www.heidijacobs.org) ___________________________________________ Heidi Jacobs, PhD Associate Professor of Radiology Massachusetts General Hospital | Harvard Medical School Department of Radiology Athinoula A. Martinos Center for Biomedical Imaging 149 13th Street Charlestown, MA 02129 hjacobs(a)mgh.harvard.edu www.heidijacobs.org<http://www.heidijacobs.org> Learn about what we do and why: https://shorturl.at/fEFa8 [Image removed by sender. A blue and black logo Description automatically generated] [Image removed by sender. signature_3581815700] Social media: Linkedin: https://www.linkedin.com/in/heidijacobs/ Bluesky: https://bsky.app/profile/heidijacobslab.bsky.social [Please note: I might send an email out of office hours, at weekends or during holidays. This suits me – but doesn’t mean I expect you to do the same]

1 0

Singularity/Apptainer Image for FreeSurfer 8.2.0 on RHEL 8 HPC
by nooshin Safari 18 May '26

18 May '26

External Email - Use Caution Dear FreeSurfer team, We need a Singularity/Apptainer image for the latest stable version of FreeSurfer to be run on RHEL 8 HPC. We would like to work with a Singularity/Apptainer image for FreeSurfer 8.2.0. We will be running it on the institutional machine at MD Anderson Cancer Center. The operating system is Red Hat Enterprise Linux 8.7 (Ootpa). Given that the Apptainer image for FreeSurfer 8.2.0 does not appear to be available online, could you please let us know how long it would take to create one? We have a grant deadline in 45 days, and having this image available would significantly help us meet that deadline. Thank you for your guidance. Best regards, Nooshin

2 1

Singularity/Apptainer Image for FreeSurfer 8.2.0 on RHEL 8 HPC
by nooshin Safari 18 May '26

18 May '26

2 1

← Newer
1
2
3
4
5
6
7
...
3009
Older →

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