Freesurfer

freesurfer@nmr.mgh.harvard.edu

18 participants
30131 discussions

Specifying contrasts for an unbalanced design
by Trolle, Carl 20 Sep '22

20 Sep '22

Dear FreeSurfer team, We are investigating if there is a difference in cortical thicknesses between persons with a condition and controls, regressing out the effect of site, sex, age, but have some concerns on how to specify our contrasts. We don’t think that the age-slopes will be different between the patients and the controls but would like to check that formally using DODS before trying DOSS. Currently, we have only one female control from two of the sites and are left with 10 instead of 12 classes. Would be very grateful if you have time to check that we’ve got this right. Also, when we want to check for interactions between class and age to see if we can use DOSS instead, is the attached “interactionmatrix” correct? I’ve looked at the tutorials and this example<https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg49602.html>, but tried to adapt it to our situation. If we do find an interaction, would that cluster then show up on the cluster summary in the glmdir-folder? Thank you and sorry for all the questions, just wanting to make sure we are doing this correctly. Kind regards, Carl Trolle Research Fellow Spaulding Neuroimaging Lab FreeSurfer version: freesurfer-linux-centos7_x86_64-7.1.1-20200723-8b40551 Platform: CentOS Stream 8

2 1

password
by Nail Bulakbasi 20 Sep '22

20 Sep '22

External Email - Use Caution Dear administrator I downloaded the FS_7_2_0_Ubuntu_18_04_06.ova.7z file to run with Oracle virtual box but I need the password to open the file. I will be grateful if you send me the password. Thanks in advance ADDRESS: Nail BULAKBASI, M.D. Professor of Radiology & Neuroradiology (EDINR, EDiPNR) Director, Dr.Suat Günsel University of Kyrenia Hospital Chairman, Department of Radiology Turkish Republic of Northern Cyprus nail.bulakbasi(a)neu.edu.tr nail.bulakbasi(a)kyrenia.edu.tr nbulakbasi(a)yahoo.comADDRESS: P Please consider the environment before printing this email.

1 0

Help - segmentHA_T1_long.sh "list of time points not found"
by Raikkonen Kimi 19 Sep '22

19 Sep '22

External Email - Use Caution Dear FreeSurfer, Hi, I was trying to run longitudinal analysis of the hippocampus and amygdala segmentation pipeline after running the main FreeSurfer longitudinal pipeline. I ran segmentHA_T1_long.sh 101_MRI1.long.101_template segmentHA_T1_long.sh 101_MRI2.long.101_template segmentHA_T1_long.sh 101_MRI3.long.101_template etc. but I got the following error message: List of time points: /study/101_MRI1.long.101_template/base-tps does not exist I think I’ve headed the wrong direction, I was wondering how to do the list of time points for a subject. Thanks so much for your help in advance. Best regards, Alex

2 1

Zip folder passcode
by Chi-Wen Jao 19 Sep '22

19 Sep '22

External Email - Use Caution Dear freesurf developer I was trying to install FreeSurfer on a virtual machine and was asked for the password to open the archive in the zip folder FS_7_2_0_Ubuntu_18_04_06.ova.7z – I just wanted to ask for that password. Thank you for your help! Kind regards, Chi-Wen

1 0

Zip folder passcode
by Enya Weidner 19 Sep '22

19 Sep '22

External Email - Use Caution Hi everyone, I was trying to install FreeSurfer on a virtual machine and was asked for the password to open the archive in the zip folder FS_7_2_0_Ubuntu_18_04_06.ova.7z – I just wanted to ask for that password. Thank you for your help! Kind regards, Enya ---------------------- M.Sc. Enya Weidner Universität Bielefeld Fakultät für Psychologie und Sportwissenschaft AE02 Affektive Neuropsychologie Raum: T5-220 Telefon: 0521-106 4533 Email: enya.weidner(a)uni-bielefeld.de

1 0

Weill Cornell Medicine position for Neuroimaging Methods Specialist in NYC
by Tracy A. Butler 16 Sep '22

16 Sep '22

External Email - Use Caution Position Summary The Brain Health Imaging Institute (BHII) in the Department of Radiology of Weill Cornell Medicine is seeking highly qualified applicants for a permanent position as Neuroimaging Methods Specialist to support cutting edge current and future MRI and PET research by BHII and collaborating investigators. Position rank is open and can range from Staff Research Scientist through Assistant, Associate or Full Professor. A strong record of extramural funding, or high potential for future extramural funding, is highly desirable, but required only for tenure-track positions. The Neuroimaging Methods Specialist will have the opportunity interact with and learn from multidisciplinary BHII faculty including Neurologists, Radiologists, Neuroscientists, Engineers and Physicists. He/she will have access to state-of-the-art neuroimaging and computational facilities and other equipment including multiple MR (3T and 7T), PET/CT and PET/MR scanners, an on-site cyclotron and radiochemistry facilities, cluster of high computing nodes (with 1000 cores) and data storage of up to 6 pbyte. There is also access to a state-of-the art wet lab to enable validation and exploratory work. Weill Cornell Medicine, located in the upper east side of New York City, has a vibrant biomedical research community and strong ties to the main Cornell campus in Ithaca and the Tech campus on Roosevelt Island. The BHII, a multidisciplinary research group headed by Dr. Mony de Leon, uses MRI, PET and complementary methods to study brain anatomy, pathology and function in normal aging, and in disease states including Alzheimer’s Disease and Traumatic Brain Injury. Brain mechanisms of particular interest include brain fluid clearance, brain functional/structural networks, inflammation, cardiovascular effects on brain structure and function and iron deposition. In addition to neuroimaging, the BHII supports active investigator initiated clinical trials, fluid (blood and csf) biomarker development and testing, detailed cognitive and neurologic assessment and neuropathological examination. Responsibilities: * Develop and apply cutting-edge MRI and PET processing and analysis methods * Optimize standard operating procedures and automated pipelines for acquiring, transferring, de-identifying, processing, monitoring quality control, storing and backing-up neuroimage data. * Disseminate research results through publications in high-impact journals and presentations at top international conferences. * Collaborate with other BHII investigators to describe neuroimaging methods and results in manuscripts and grant applications. * Support, train and mentor postdoctoral fellows and students with medical or engineering background Required Qualifications: * PhD or equivalent degree in Electrical/Biomedical Engineering, Computer Science, Neuroscience, Applied Mathematics, Statistics or other related field. * Track record of publications in top scientific conferences and journals * Significant experience with common neuroimage processing and analyses tools, e.g. FSL, SPM, FreeSurfer, PMOD * Experience with one or several common coding /scripting tools e.g Bash, Python, Matlab, C, Java * Effective communication skills, both written and verbal * Track record of effectively working both independently and as part of a multidisciplinary team Desired Familiarity with one or more of the following fields is a plus: * PET image processing and kinetic modeling for established radiotracers (e.g. to identify amyloid, tau, microglial activation) as well as novel tracers. * Functional and Structural brain MRI * Brain anatomy and physiology * MRI and PET imaging of animal models * EEG To apply: Please send a one-page statement of career goals and research interests and full CV to  mas4023(a)med.cornell.edu<mailto:mas4023@med.cornell.edu> with subject title: Application for Neuroimaging Methods Specialist position. Weill Cornell Medicine: Founded in 1898, and affiliated with what is now New York-Presbyterian Hospital (NYPH) since 1927, Weill Cornell Medicine (WCM) is among the top-ranked clinical and medical research centers in the country. In addition to offering degree sin medicine, WCM also has PhD programs in biomedical research and education at the Weill Cornell Graduate School of Medical Sciences, and with neighboring Sloan-Kettering Institute and The Rockefeller University, has established a joint MD-PhD program for students to intensify their pursuit of Weill Cornell's triple mission of education, research, and patient care. WCM is divided into 24 basic science and patient care departments that focus on the sciences underlying clinical medicine and/or encompass the study, treatment, and prevention of human diseases. The basic science and clinical departments are located in buildings that straddle York Ave. between 68th and 72nd streets on Manhattan’s Upper East Side. Weill Cornell Medical College has 1781 full-time faculty (3582 total faculty) distributed across 8 basic science and 15 clinical departments. WCM maintains major affiliations with Memorial Sloan-Kettering Cancer Center, The Rockefeller University, the Hospital for Special Surgery, as well as with the metropolitan-area institutions that constitute NYP Healthcare Network. Weill Cornell Medicine is a comprehensive academic medical center that's committed to excellence in patient care, scientific discovery, and the education of future physicians in New York City and around the world. Our doctors and scientists-faculty from Weill Cornell Medical College, Weill Cornell Graduate School of Medical Sciences, and the Weill Cornell Physician Organization-are engaged in world-class clinical care and cutting-edge research that connect patients to the latest treatment innovations and prevention strategies. Located in the heart of the Upper East Side's scientific corridor, Weill Cornell Medicine's powerful network of collaborators extends to its parent university Cornell University; to Qatar, where an international campus offers a U.S. medical degree; and to programs in Tanzania, Haiti, Brazil, Austria and Turkey. Our medical practices serve communities throughout New York City, and our faculty provide comprehensive care at NewYork-Presbyterian Hospital/Weill Cornell Medical Center, NewYork-Presbyterian/Lower Manhattan Hospital, and NewYork-Presbyterian/Queens. At Weill Cornell Medicine, we work together to treat each individual, not just their conditions or illnesses, as we strive to deliver the finest possible care for our patients - the center of everything we do. Diversity is one of Weill Cornell Medicine’s core values and is essential to achieving excellence in patient care, research, and education. We welcome applications from candidates who share our commitment to fostering a culture of fairness, equity, and belonging. Weill Cornell Medicine is an Equal Employment Opportunity Employer, providing equal employment opportunities to all qualified applicants without regard to race, sex, sexual orientation, gender identity, national origin, color, age, religion, protected veteran or disability status, or genetic information. Tracy Butler, MD Medical Director, Brain Health Imaging Institute Associate Professor of Neurology in Radiology and Psychiatry Weill Cornell Medicine Department of Radiology 407 East 61st street, 2nd Floor New York, NY 10065

1 0

Re: [Freesurfer] How to modify affine transform if we flip the image in y direction?
by Arman Avesta 16 Sep '22

16 Sep '22

External Email - Use Caution Thank you so much, Douglas! Turned out that negating the second column of the affine matrix was the right thing to do :) My code was not working for another reason (the code wasn't reading the nifti header correctly). You're right: the affine matrix is the multiplication of scaling, rotation, shear, and translation matrices. So if the original affine matrix is M, the scaling matrix is S, and the corrected affine matrix is Mc: Mc = M @ S If we define S matrix as: S = np.array([[1, 0, 0, 0], [0, -1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]]) The result on Mc will be negating the second column, which we can directly do by: Mc = MMc[:, 1] = - M[:, 1] instead of multiplying M and S. Thank you for your input, I appreciate it :) Arman Avesta the easiest way to think about it is in terms of matrices. You have an affine matrix M already that converts from col,row,slice to xyz. When you flip the 2nd dim, it is like you are applying a matrix R to the new col,row,slice to get the old col,row,slice, so your new affine matrix will be M*R. R will be something like1 0 0 0 0 -1 0 Nrows 0 0 1 0 0 0 1 On 9/15/2022 4:39 PM, Arman Avesta wrote: External Email - Use Caution Hi gang, I hope you're enjoying the last stretch of summer! I am writing a python code that reads NIfTI images, changes their coordinate system into LAS (*MailScanner has detected a possible fraud attempt from "secure-web.cisco.com" claiming to be* http://secure-web.cisco.com/1BgqieiGDbxuT4tCHyJZWr1ABBl6JirOFjUmCoUV-_5mJyi… <http://secure-web.cisco.com/1_56Y7I55cGEKACYrc6S3VTmRRwExp_9ZrFfAHl414WhHLP…>), and then saves the corrected NIfTI mage.The function that corrects the coordinate system is copied below. I am not sure if I am changing the *affine transform* (from the MRI space to the scanner space) correctly.For instance, to go from LPS to LAS, I should flip the image in 2nd dimension. The piece that I'm not sure about is if I should negate the 2nd column of the affine matrix or if the affine correction is more complicated:image = np.flip(image, 1) affine[:, 1] = - affine[:, 1] I would really really appreciate any help here! I'm stuck :D Cheers, Arman ----------------------------------------------------------------- def correct_coordinates(image, affine, coords): """ This function re-orients the MRI volume into the standard radiology system, i.e. ('L','A','S'), and also corrects the affine transform for the volume (from the MRI volume space to the scanner space). Inputs: - image: MRI volume - affine: affine transform for the MRI volume (e.g. [[-5, 0, 0, 1] [0, -5.2, 7.8 1.6] [0, 6.9, 5.9, -7.7] [0, 0, 0, 1]] - coords: the coordinate system of the MRI volume (e.g. ('L', 'P', 'S') Outputs: - image: corrected MRI volume in the standard radiology coordinate system: ('L', 'A', 'S') - affine: corrected affine transform (from MRI volume space to the scanner space). """ if coords == ('L', 'A', 'S'): return image, affine if coords == ('R', 'A', 'S'): image = np.flip(image, 0) affine[:, 0] = - affine[:, 0] return image, affine if coords == ('L', 'P', 'S'): image = np.flip(image, 1) affine[:, 1] = - affine[:, 1] # print('func coords: ', coords) # print('func corr coords: ', nib.aff2axcodes(affine)) # print('function affine: \n', affine) return image, affine if coords == ('L', 'S', 'A'): image = np.swapaxes(image, 1, 2) affine[:, [1, 2]] = affine[:, [2, 1]] return image, affine if coords == ('L', 'I', 'A'): # L,I,A --> L,A,I image = np.swapaxes(image, 1, 2) affine[:, [1, 2]] = affine[:, [2, 1]] # L,A,I --> L,A,S image = np.flip(image, 2) affine[:, 2] = - affine[:, 2] return image, affine if coords == ('L', 'P', 'S'): image = np.flip(image, 1) affine[:, 1] = - affine[:, 1] return image, affine if coords == ('L', 'I', 'P'): # L,I,P --> L,I,A image = np.flip(image,2) affine[:, 2] = - affine[:, 2] # L,I,A --> L,A,I image = np.swapaxes(image, 1, 2) affine[:, [1, 2]] = affine[:, [2, 1]] # L,A,I --> L,A,S image = np.flip(image, 2) affine[:, 2] = - affine[:, 2] return image, affine if coords == ('P', 'I', 'R'): # P,I,R --> A,I,R image = np.flip(image, 0) affine[:, 0] = - affine[:, 0] # A,I,R --> A,S,R image = np.flip(image, 1) affine[:, 1] = - affine[:, 1] # A,S,R --> A,S,L image = np.flip(image, 2) affine[:, 2] = - affine[:, 2] # A,S,L --> L,A,S image = np.moveaxis(image, [0, 1, 2], [1, 2, 0]) affine[:, [0, 1, 2]] = affine[:, [1, 2, 0]] return image, affine raise Exception('coords not identified: please go to correct_coordinates function ' 'and define the coordinate system.')

1 0

Password for virtual disk image (vdi)
by Christian Stald Skoven 16 Sep '22

16 Sep '22

External Email - Use Caution I would like a/the password for extracting the 7z-file of the ubuntu virtual disk image (vdi). Thank you in advance :) Br, Christian Stald Skoven Ph.D., Post Doc, Preclinical MRI CFIN, AUH

1 0

Registration did not converge in 5 steps
by tom.willems＠unibe.ch 16 Sep '22

16 Sep '22

External Email - Use Caution Dear Freesurfer Experts I am trying use segmentHA_T1.sh: During the registration of imageDump.mgz to target hippoAmygBinaryMask_autoCropped.mgz, I get the warning: * WARNING: Registration did not converge in 5 steps! * Problem might be ill posed. * Please inspect output manually! Also I can’t seem to find „imageDump_coregistered.mgz“, to check the aligned result (as advised in the log file) The script runs fine and the outputs look ok, too; still I was wondering what is going wrong here. Many Thanks in advance! Tom

1 0

How to modify affine transform if we flip the image in y direction?
by Arman Avesta 16 Sep '22

16 Sep '22

External Email - Use Caution Hi gang, I hope you're enjoying the last stretch of summer! I am writing a python code that reads NIfTI images, changes their coordinate system into LAS (http://secure-web.cisco.com/1_56Y7I55cGEKACYrc6S3VTmRRwExp_9ZrFfAHl414WhHLP…) and then saves the corrected NIfTI mage. The function that corrects the coordinate system is copied below. I am not sure if I am changing the affine transform (from the MRI space to the scanner space) correctly. For instance, to go from LPS to LAS, I should flip the image in 2nd dimension. The piece that I'm not sure about is if I should negate the 2nd column of the affine matrix or if the affine correction is more complicated:image = np.flip(image, 1) affine[:, 1] = - affine[:, 1]I would really really appreciate any help here! I'm stuck :D Cheers,Arman -----------------------------------------------------------------def correct_coordinates(image, affine, coords): """ This function re-orients the MRI volume into the standard radiology system, i.e. ('L','A','S'), and also corrects the affine transform for the volume (from the MRI volume space to the scanner space). Inputs: - image: MRI volume - affine: affine transform for the MRI volume (e.g. [[-5, 0, 0, 1] [0, -5.2, 7.8 1.6] [0, 6.9, 5.9, -7.7] [0, 0, 0, 1]] - coords: the coordinate system of the MRI volume (e.g. ('L', 'P', 'S') Outputs: - image: corrected MRI volume in the standard radiology coordinate system: ('L', 'A', 'S') - affine: corrected affine transform (from MRI volume space to the scanner space). """ if coords == ('L', 'A', 'S'): return image, affine if coords == ('R', 'A', 'S'): image = np.flip(image, 0) affine[:, 0] = - affine[:, 0] return image, affine if coords == ('L', 'P', 'S'): image = np.flip(image, 1) affine[:, 1] = - affine[:, 1] # print('func coords: ', coords) # print('func corr coords: ', nib.aff2axcodes(affine)) # print('function affine: \n', affine) return image, affine if coords == ('L', 'S', 'A'): image = np.swapaxes(image, 1, 2) affine[:, [1, 2]] = affine[:, [2, 1]] return image, affine if coords == ('L', 'I', 'A'): # L,I,A --> L,A,I image = np.swapaxes(image, 1, 2) affine[:, [1, 2]] = affine[:, [2, 1]] # L,A,I --> L,A,S image = np.flip(image, 2) affine[:, 2] = - affine[:, 2] return image, affine if coords == ('L', 'P', 'S'): image = np.flip(image, 1) affine[:, 1] = - affine[:, 1] return image, affine if coords == ('L', 'I', 'P'): # L,I,P --> L,I,A image = np.flip(image,2) affine[:, 2] = - affine[:, 2] # L,I,A --> L,A,I image = np.swapaxes(image, 1, 2) affine[:, [1, 2]] = affine[:, [2, 1]] # L,A,I --> L,A,S image = np.flip(image, 2) affine[:, 2] = - affine[:, 2] return image, affine if coords == ('P', 'I', 'R'): # P,I,R --> A,I,R image = np.flip(image, 0) affine[:, 0] = - affine[:, 0] # A,I,R --> A,S,R image = np.flip(image, 1) affine[:, 1] = - affine[:, 1] # A,S,R --> A,S,L image = np.flip(image, 2) affine[:, 2] = - affine[:, 2] # A,S,L --> L,A,S image = np.moveaxis(image, [0, 1, 2], [1, 2, 0]) affine[:, [0, 1, 2]] = affine[:, [1, 2, 0]] return image, affine raise Exception('coords not identified: please go to correct_coordinates function ' 'and define the coordinate system.') Arman Avesta, MD PhD student in the combined radiology residency & PhD training program Yale School of Medicine & Yale Graduate School of Arts and Sciences 195 Church St 6th Floor, New Haven, CT 06510 arman.avesta(a)yale.edu medicine.yale.edu/profile/arman_avesta

2 1

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