Freesurfer March 2025

freesurfer@nmr.mgh.harvard.edu

49 participants
63 discussions

EasyReg/SynthSeg singular matrix error
by Suresh, Krish 17 Mar '25

17 Mar '25

External Email - Use Caution Hi, Thank you in advance for your assistance, I am new to freesurfer. I am trying to use EasyReg to register an axial T1 post-contrast series and an axial high-resolution T2 series, so that I can propagate a tumor segmentation from the T1 to the T2. I am getting the error message copied below. One problem I see is that the SynthSeg segmentation is not accurate, see screenshot. I am running freesurfer version 7.4.1 on WSL 2/Ubuntu 22. I would appreciate any pointers you may have, thank you very much. Krish mri_easyreg --ref t1.nii --flo t2.nii --ref_seg t1_seg.nii --flo_seg t2_seg.nii --ref_reg t1_reg.nii --flo_reg t2_reg.nii --fwd_field fwd_field.nii --bak_field bak_field.nii --threads -1 2025-03-16 21:49:05.684219: I tensorflow/core/util/port.cc:110] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`. 2025-03-16 21:49:05.687469: I tensorflow/tsl/cuda/cudart_stub.cc:28] Could not find cuda drivers on your machine, GPU will not be used. 2025-03-16 21:49:05.728495: I tensorflow/tsl/cuda/cudart_stub.cc:28] Could not find cuda drivers on your machine, GPU will not be used. 2025-03-16 21:49:05.729021: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations. To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags. 2025-03-16 21:49:06.476538: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT using all available threads ( 12 ) Segmenting reference image Reading reference image Setting up segmentation net Inference / segmentation 1/1 [==============================] - 11s 11s/step Postprocessing Saving result Segmenting floating image Reading floating image Inference / segmentation 1/1 [==============================] - 19s 19s/step Postprocessing Saving result Linear registration Computing centroids and estimating affine transform Traceback (most recent call last): File "/usr/local/freesurfer/7.4.1/python/scripts/mri_easyreg", line 1809, in <module> main() File "/usr/local/freesurfer/7.4.1/python/scripts/mri_easyreg", line 180, in main Mref = getM(atlasCOG[:, ok > 0], refCOG[:, ok > 0]) File "/usr/local/freesurfer/7.4.1/python/scripts/mri_easyreg", line 1198, in getM x = np.matmul(np.linalg.inv(np.matmul(np.transpose(A), A)), np.matmul(np.transpose(A), b)) File "<__array_function__ internals>", line 180, in inv File "/usr/local/freesurfer/7.4.1/python/lib/python3.8/site-packages/numpy/linalg/linalg.py", line 552, in inv ainv = _umath_linalg.inv(a, signature=signature, extobj=extobj) File "/usr/local/freesurfer/7.4.1/python/lib/python3.8/site-packages/numpy/linalg/linalg.py", line 89, in _raise_linalgerror_singular raise LinAlgError("Singular matrix") numpy.linalg.LinAlgError: Singular matrix [cid:546a1e09-0300-496f-9c0d-58eb45c8e243]

2 1

aparc2aseg not working
by Jin Liu 17 Mar '25

17 Mar '25

External Email - Use Caution Hello FreeSurfer team, I’m trying to create a lobe segmentation for white matter. I first created the annotation files using mri_annotation2label with --lobesStrict for each hemisphere, and then tried to map the lobe parcellation onto segmentation volume by using mri_aparc2aseg. But somehow the resulting image was not segmented by lobes. It contains cortical parcellation by lobe, detailed subcortical segmentation by structure, and white matter is marked as a whole (cerebral white matter). I then checked the annotation files created by lobesStrict and they look fine, without any subcortical labels. I wonder where I made a mistake, or if there is an easier way to do lobe segmentation. The code I used is as follows: mri_annotation2label --subject 1300 --hemi lh --labelbase ./1300/label/aparc-lh --lobesStrict ./1300/label/lh.aparc.lobesStrict mri_annotation2label --subject 1300 --hemi rh --labelbase ./1300/label/aparc-rh --lobesStrict ./1300/label/rh.aparc.lobesStrict mri_aparc2aseg --s 1300 --annot aparc.lobesStrict --o ./1300/mri/lobesStrict.mgz Best, Jin

2 1

Recon-all timing with fs v.8.0 and parallel recon-all
by Paola Di Filippo 14 Mar '25

14 Mar '25

External Email - Use Caution Hi everyone. I am using Freesurfer v.8.0 on a MacBook Pro with M4 Pro Chip, 48 GB RAM, with MacOs Sequoia 15.2. I tried performing a couple of recon-all, they were completed without hard failures but with the following timing: recon-all-run-time-hours 13.590 recon-all-run-time-hours 8.973 I was wondering if considering my hardware characteristics and the FS version I am using, this is normal or there’s something that I am doing wrong. I have attached my log of both recons. Also, is the following command (Im aware it requires both Xcode and parallel installation) still valid in order to perform recon-all of multiple subjects at the same time with FS v. 8.0? How many subjects would you recommend to analyze simultaneously, and what kind of timing should I be expecting, considering that I have 14 logical cores? Sorry about the many questions and many thanks in advance ls *.nii | parallel --jobs 8 recon-all -s {.} -i {} -all -qcache --

1 0

Re: [Freesurfer] Memory Allocation Issue with mri_histo_atlas_segment_fireants
by fsbuild 14 Mar '25

14 Mar '25

External Email - Use Caution  The notification, Allocation of 5033164800 exceeds 10% of free system memor … makes me think it’s asking to allocate an additional ~4G of RAM on top of what everything running on the machine is already using.  That apparently generates the 10% of free memory warning, and then the system actually fails to allocate memory to run a new process. If you are running other software applications that use appreciable system memory, then I would try not running anything else other than the mri_histo_atlas_segment_fireants command.  If that still fails to allocate memory as the only running application, then I suggest switching to a machine with more memory. - R. On Mar 14, 2025, at 07:35, Payam Sadeghi Shabestari <payam.sadeghishabestari(a)uzh.ch> wrote:        External Email - Use Caution        Hello FS Dev Team,I'm attempting to run mri_histo_atlas_segment_fireants on Ubuntu 20.04.6, but I encountered a "Cannot allocate memory" error. Could this be due to insufficient RAM? I would appreciate your assistance.Best regards, Payam Normalizing intensitiesSubdividing brainstem into left and right halves, and cropping bottom if needed   Running EasyReg2025-03-14 10:50:24.182715: I tensorflow/tsl/cuda/cudart_stub.cc:28] Could not find cuda drivers on your machine, GPU will not be used.2025-03-14 10:50:24.227568: I tensorflow/tsl/cuda/cudart_stub.cc:28] Could not find cuda drivers on your machine, GPU will not be used.2025-03-14 10:50:24.228099: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.To enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.2025-03-14 10:50:24.983225: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRTUserWarning: int_downsize is deprecated, use the int_resolution parameter.UserWarning: unet_half_res is deprecated, use the svf_resolution parameter.2025-03-14 10:51:31.533186: W tensorflow/tsl/framework/cpu_allocator_impl.cc:83] Allocation of 5033164800 exceeds 10% of free system memory.2025-03-14 10:54:04.568813: W tensorflow/tsl/framework/cpu_allocator_impl.cc:83] Allocation of 5033164800 exceeds 10% of free system memory.  Creating and applying mask for left hemisphereReading in atlasTraceback (most recent call last):  File "/usr/local/freesurfer/8.0.0/python/packages/ERC_bayesian_segmentation//scripts/segment_fireants.py", line 364, in <module>    for n, (prior_indices, prior_values) in enumerate(label_loader):  File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 438, in __iter__    return self._get_iterator()  File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 386, in _get_iterator    return _MultiProcessingDataLoaderIter(self)  File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 1039, in __init__    w.start()  File "/usr/local/freesurfer/8..0.0/python/lib/python3.8/multiprocessing/process.py", line 121, in start    self._popen = self._Popen(self)  File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/multiprocessing/context.py", line 224, in _Popen    return _default_context.get_context().Process._Popen(process_obj)  File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/multiprocessing/context.py", line 277, in _Popen    return Popen(process_obj)  File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/multiprocessing/popen_fork.py", line 19, in __init__    self._launch(process_obj)  File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/multiprocessing/popen_fork.py", line 70, in _launch    self.pid = os.fork()OSError: [Errno 12] Cannot allocate memory _______________________________________________Freesurfer mailing listFreesurfer@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

1 0

Memory Allocation Issue with mri_histo_atlas_segment_fireants
by Payam Sadeghi Shabestari 14 Mar '25

14 Mar '25

External Email - Use Caution Hello FS Dev Team, I'm attempting to run mri_histo_atlas_segment_fireants on Ubuntu 20.04.6, but I encountered a "Cannot allocate memory" error. Could this be due to insufficient RAM? I would appreciate your assistance. Best regards, Payam Normalizing intensities Subdividing brainstem into left and right halves, and cropping bottom if needed Running EasyReg 2025-03-14 10:50:24.182715: I tensorflow/tsl/cuda/cudart_stub.cc:28] Could not find cuda drivers on your machine, GPU will not be used. 2025-03-14 10:50:24.227568: I tensorflow/tsl/cuda/cudart_stub.cc:28] Could not find cuda drivers on your machine, GPU will not be used. 2025-03-14 10:50:24.228099: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations. To enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags. 2025-03-14 10:50:24.983225: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT UserWarning: int_downsize is deprecated, use the int_resolution parameter. UserWarning: unet_half_res is deprecated, use the svf_resolution parameter. 2025-03-14 10:51:31.533186: W tensorflow/tsl/framework/cpu_allocator_impl.cc:83] Allocation of 5033164800 exceeds 10% of free system memory. 2025-03-14 10:54:04.568813: W tensorflow/tsl/framework/cpu_allocator_impl.cc:83] Allocation of 5033164800 exceeds 10% of free system memory. Creating and applying mask for left hemisphere Reading in atlas Traceback (most recent call last): File "/usr/local/freesurfer/8.0.0/python/packages/ERC_bayesian_segmentation//scripts/segment_fireants.py", line 364, in <module> for n, (prior_indices, prior_values) in enumerate(label_loader): File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 438, in __iter__ return self._get_iterator() File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 386, in _get_iterator return _MultiProcessingDataLoaderIter(self) File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 1039, in __init__ w.start() File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/multiprocessing/process.py", line 121, in start self._popen = self._Popen(self) File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/multiprocessing/context.py", line 224, in _Popen return _default_context.get_context().Process._Popen(process_obj) File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/multiprocessing/context.py", line 277, in _Popen return Popen(process_obj) File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/multiprocessing/popen_fork.py", line 19, in __init__ self._launch(process_obj) File "/usr/local/freesurfer/8.0.0/python/lib/python3.8/multiprocessing/popen_fork.py", line 70, in _launch self.pid = os.fork() OSError: [Errno 12] Cannot allocate memory

1 0

Research Assistant Position in the Buckeye Brain Aging Lab
by Graham, Jillian 13 Mar '25

13 Mar '25

External Email - Use Caution Dear all, We're hiring a full-time research assistant. Please see below! Applications are being sought for a full-time research assistant position in the Buckeye Brain Aging Lab (B-BAL; thebbal.com) directed by Dr. Scott Hayes. The B-BAL studies the relationships between aging, physical activity, cognition, and the brain [using Magnetic Resonance Imaging (MRI) metrics]. We also examine neural correlates of memory in healthy and clinical populations. The Research Assistant 1 assists in the general coordination and conduct of research in the B-BAL, including: recruitment, screening, and scheduling and coordination of participant appointments (cognitive, MRI, and physical fitness assessments), subject payments, and submission of IRB amendments; assists with research lab purchases and maintains lab supplies; administration of neuropsychological tests and/or questionnaires to young and older adults; assists with cognitive, MRI, and physical fitness data collection, including obtaining informed consent, cognitive testing (computer and paper and pencil based testing), MRI data acquisition, and record keeping; assists with cognitive and MRI data processing and analyses, supervises undergraduate research assistants. Desired qualifications include: Bachelor's degree in psychology, neuroscience, computer science, data science, engineering, or physics; Additional desired (but not required) skills: programming experience in Bash, Python, and/or Matlab, as well as experience in neuroimaging data analyses; knowledge of software packages, such as FreeSurfer, FSL, AFNI, SPM, CONN, R. The position is available for one year with the possibility of extension for additional years based on satisfactory performance and availability of funds. If you are interested in being considered for the position, please submit a cover letter, current CV, course transcripts, and a writing sample (research paper from course, poster presented at a conference) as a single pdf file to bbal(a)osu.edu If you have questions about the position or the lab, please email bbal(a)osu.edu. Lab website here: https://secure-web.cisco.com/1ormq1uhzhqmauEuhNCDwXNMI_0Kqf2X4Uq-oWXTYA0jMe… The target state date is May 1, 2025, but this is negotiable. The B-BAL | Buckeye Brain Aging Lab - U.OSU<https://secure-web.cisco.com/1ormq1uhzhqmauEuhNCDwXNMI_0Kqf2X4Uq-oWXTYA0jMe…> The Buckeye Brain Aging Lab (B-BAL), directed by Dr. Scott M. Hayes, works primarily with healthy young, middle-aged, and older adults, as well as individuals with mild cognitive impairment (MCI) and neurodegenerative disorders such as Alzheimer’s disease. u.osu.edu Best, [cid:8857d538-8197-4236-8e5c-9a2588c0d194] Jillian Graham Research Assistant The Buckeye Brain Aging Lab Department of Psychology 1835 Neil Avenue, Columbus, OH 43210 graham.1492(a)osu.edu/ osu.edu<http://secure-web.cisco.com/1CuyboOisXUdjDn-hrgdBDcslI6TY7acxPwgq6MR_nmfSaY…> Pronouns: she/her/hers

1 0

[FreeSurfer] Registration of non-mri image to same registered space of mri_image (recon_all)
by Chavez, Caitlyn 13 Mar '25

13 Mar '25

External Email - Use Caution Hello FreeSurfer Community, I am looking for any help/information/documentation in replicating the transforms performed in FreeSurfer recon_all on a non-mri volume image. I am trying to align the voxel space of a non-mri image with the masking (aseg, seg, parc) outputs of the registered (recon_all) mri_image. I am sure it is not as simple as running the non-mri image through FreeSurfer recon_all, but I am unsure how to align the images with the exact same parameters previously used. If there is a method of extracting the parameters and the algorithm from FreeSurfer to register the non-mri images using python I would also be interested in learning about this as well. Any help/information/documentation is greatly appreciated. Thank you in advance for your help! Best, Caitlyn Additional information: recon-all -i subject_1.nii.gz -subjid subject_1 -all ##(Successful run from the command line ) Program versions used: 7.2.0 (freesurfer-linux-centos7_x86_64-7.2.0-20210720-aa8f76b) 7.2.0 Non-mri image info Can be converted to .npy or .nii format

2 1

Proposed Known Issue: v8.0.0 and v8.0.0-beta
by Abhishek Roy 13 Mar '25

13 Mar '25

External Email - Use Caution Dear team and subscribers, I'd like to highlight an issue I've had with recon-all in v8.0.0 and v8.0.0-beta. I've also logged this as a github issue: https://secure-web.cisco.com/1jJPFLQEv-RDDChH3FDSznGS6PwsMf1JbYPxvNx-0gCoDY… The value of `SegmentedTotalIntraCranialVolume` is always 0.00 . Found out why. It seems the csvprint module needs python2. this has been updated to python3 in commit 9a59bec<https://secure-web.cisco.com/1ye7ehFogBTOSZ5-SI8taauhvvbvu-gHsmX1xAYZrH1X12…>. However this commit did not make it into the v8.0.0 or v8.0.0-beta version <https://secure-web.cisco.com/1ye7ehFogBTOSZ5-SI8taauhvvbvu-gHsmX1xAYZrH1X12…> It seems that the -motioncor step in autorecon1 in recon-all prints the value of Total Intracranial Volume from the mri/synthseg.vol.csv into the text file stats/synthseg.tiv.dat. This file is read by mri_segstats later in the process to make the aseg.stats file. When csvprint is broken, the file stats/synthseg.tiv.dat is empty, so the value of SegmentedTotalIntraCranialVol is 0 in aseg.stats. I believe this should be logged as a known issue for these software versions because the software incorrectly reports the value of sTIV in aseg.stats. A hacky temporary fix would be to copy the csvprint python file from the mentioned commit, and overwrite the version in one's freesurfer install at $FREESURFER_HOME/bin/csvprint Abhishek Roy, Oxford Brain Diagnostics.

1 0

Freesurfer Anatomical Parcellation Issues
by hmarzban＠student.ubc.ca 12 Mar '25

12 Mar '25

External Email - Use Caution Hello, I hope you are well. I was wondering if there is an easy way to correct for missegmentation of the grey matter parcellation. For example, would it be possible to correct for the overestimation of the pre-postcentral gyri (see picture below)? Since it seems as if the parcellation is extending into the meninges. Warmly, Hengameh

1 0

autorecon1_norm.mgz not found
by Ni, Zhen (NIH/NINDS) [C] 12 Mar '25

12 Mar '25

External Email - Use Caution Hello, I'm using the command "recon-all -autorecon1 -subjid $SUBJECTNAME". After about 2 hours, there are a few mgz and other files generated in the /mri directory, however, I got an error message: WARNING: entowm has very low volumes so not using it to edit GA of cortex label #-------------------------------------- #@# CC Seg Mon Mar 10 19:13:32 EDT 2025 \n seg2cc --s sub-002 \n ERROR: cannot find /Applications/freesurfer/8.0.0/subjects/sub-002/mri/norm.mgz Darwin nzs-MacBook-Air.local 23.6.0 Darwin Kernel Version 23.6.0: Fri Jul 5 17:54:20 PDT 2024; root:xnu-10063.141.1~2/RELEASE_X86_64 x86_64 recon-all -s sub-002 exited with ERRORS at Mon Mar 10 19:13:33 EDT 2025 I wonder if you have any idea to fix the problem. Zhen

2 6

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Freesurfer March 2025