Freesurfer August 2018

freesurfer@nmr.mgh.harvard.edu

104 participants
167 discussions

Re: [Freesurfer] Question about Freeview, 20% threshold
by Yendiki, Anastasia 28 Aug '18

28 Aug '18

The output of TRACULA is a probability distribution that tells you how likely each voxel is to be part of the tract that you’re looking at. Voxels with low probability values are not very likely to be part of the tract. Setting the threshold in freeview to 20% means that you only want to display the voxels that have probability values greater than 20% of the voxel with the greatest probability to be part of this tract. From: <freesurfer-bounces(a)nmr.mgh.harvard.edu<mailto:freesurfer-bounces@nmr.mgh.harvard.edu>> on behalf of "Daumail, Loic Jean" <LDAUMAIL(a)mgh.harvard.edu<mailto:LDAUMAIL@mgh.harvard.edu>> Reply-To: Freesurfer support list <freesurfer(a)nmr.mgh.harvard.edu<mailto:freesurfer@nmr.mgh.harvard.edu>> Date: Friday, July 20, 2018 at 10:13 AM To: "freesurfer(a)nmr.mgh.harvard.edu<mailto:freesurfer@nmr.mgh.harvard.edu>" <freesurfer(a)nmr.mgh.harvard.edu<mailto:freesurfer@nmr.mgh.harvard.edu>> Subject: [Freesurfer] Question about Freeview, 20% threshold Dear FreeSurfer experts, I am currently using TRACULA( from freeSurfer 6.0), and on the freeview visualization of merged files of the probability distributions of the tracts, I don't understand the thing about the "20% threshold". Indeed, on the tracula tutorial, it is written that the tracts are displayed in a way that they are at 20% of their maximum threshold. However, what does this mean? Is it the 20% of the most probable paths for each tracts? The adjustment bar on the side of the screen permits to vary this threshold, but when we reduce the level of the threshold, the size of the tracts increase, which seems in contradiction with the previous statements... I then suppose that my guess must be wrong? Thank you for your time and consideration! Sincerely, Loïc

1 0

[TRACULA] eddy_correct /affine registration matrices / motion/ eddy currents
by Daumail, Loic Jean 28 Aug '18

28 Aug '18

Dear FreeSurfer experts, I would like to understand better the principle of eddy_correct: -in which way the eddy currents caused by motion are set apart from the eddy currents caused by fast switching diffusion-encoding gradients? -and then, how are the affine registration matrices reflecting motion computed by eddy_correct considering the previous issues? -does eddy_correct even computes these affine registration matrices(dwi.ecclog)? or is it another tool of dmri_motion? Thanks, Loïc Daumail Graduate intern at Athinoula A. Martinos Center for Biomedical Imaging MGH/HST

2 1

error opening file dlabel/mni/lh.atr_PP_avg33_mni_bbr_end1_dil.nii.gz
by Montserrat Domingo Ayllón 28 Aug '18

28 Aug '18

External Email - Use Caution Hi everybody! I have an error in one of my subjects preprocessing error opening file dlabel/mni/lh.atr_PP_avg33_mni_bbr_end1_dil.nii.gz In the dlabel folder of this subject, the mni folder has the half of the items that the rest of subjects have (2036 vs 4568). It is a subject with severe atrophy, therefore, I don't know if the trac-preproc has exited with errors due to this atrophy or because of other facts that I'm not able to detect. Thank you in advance, Montse

2 1

Parcellation labels error
by Ali,Mohamed Tarek Mohamed 27 Aug '18

27 Aug '18

External Email - Use Caution Hello FreeSurfer Developers, While I am using [vec,labels,cnames] = read_annotation('lh.aparc.annot'), I am getting values in labels which are not found in cnames.table(:,5). These values are 0. I assigned the values to the unknown area. However, when I looked at the number of vertices per brain regions I found that corpuscallosum has no vertices in the DK atlas. This is my first week using a neuroimaging software and Freesurfer is my first software to work on, so I just don't know if that's ok or if there is something wrong. I am attaching the output of a python code I wrote to count the number of vertices per brain region in the left hemisphere of a structural MRI image from our dataset. Thanks in advance, Mohamed T. Ali Ph.D. student, Bioengineering, University of Louisville.

2 4

correlation between blood biomarkers and longitudinal decrease in cortical thickness
by Foad Taghdiri 27 Aug '18

27 Aug '18

External Email - Use Caution Hello, I have done freesurfer (v6) longitudinal preprocessing for 20 subjects (baseline and 2-year follow-up). We also have levels of some blood biomarkers (e.g. tau, neurofilament light) for these subjects at the baseline. I am interested to figure out if the level of blood biomarkers at the baseline is related to the amount of cortical thinning over time, and if so, which brain regions’ cortical thinning are significantly correlated with the blood biomarker. I was wondering if you could please guide me through this analysis and let me know how I can do it. Thanks Best, Foad

2 1

recon-all with mri_ca_normalize (longitudinal)
by miracle ozzoude 27 Aug '18

27 Aug '18

External Email - Use Caution Hello FreeSurfer team, In order to create the longitudinal base after norm.mgz, Is the command below the right way to end recon-all after creating norm.mgz? recon-all -s subjectID -autorecon1 -autorecon2 -i T1image -canorm. Thank you. Best, Paul

2 1

Follow-Up: Longitudinal Processing - Retaining Manual Edits
by KALEY E ELLIS 27 Aug '18

27 Aug '18

External Email - Use Caution Dear FreeSurfer Developers, I posted this question last week, but did not receive a response. Please let me know if I can provide any additional information. We are using your longitudinal processing pipeline (version 6.0) for our study (n=140, 3 timepoints). Our scanner was upgraded a year ago and the a filter was used on the raw images to adjust for hyperintensities in the images due to using a multi-channel coil. At this point, our bases have already been processed and consist of three timepoints with a mix of filters: T1 no filter, T2 no filter, T3 with filter. Since we have invested many resources into the manual editing process, we want to try to retain all the manual edits that have been done to the base. We still have the original scan and the filtered scan for each subject. We have some ideas on how we might be able to retain the edits, but would you be able to provide input regarding whether this seems reasonable? Your feedback is greatly appreciated! Plan A: 1) Create a new T3 cross-sectional with the correct nifti 2) Run recon-all on the existing base Q: Will this retain the previous manual edits AND incorporate the new cross-sectional image? Plan B: 1) Create a new T3 cross-sectional with the correct nifti 2) Create a new base Q: Is it possible to transfer data from old manual edits to a new base? If so, how? Thank you! Kaley Associate Research Specialist Center for Healthy Minds Waisman Center for Brain Imaging and Behavior University of Wisconsin-Madison

2 1

error with mri_ca_register in long
by miracle ozzoude 27 Aug '18

27 Aug '18

External Email - Use Caution Hello All, When running the longitudinal pipeline for version 6, i encountered this error " TransformSample: gcam has not been inverted!; Numerical result out of range". I found a thread with similar error but the problem wasn't solved. https://mail.nmr.mgh.harvard.edu/pipermail/freesurfer/2017-April/051240.html . Just like the previous thread, it happens whenever I include the "-bigventricle" flag. Below is a sample of my recon-all scripts from cross-sectional to longitudinal. The error.log and recon.log files are too large to send. Thank you. Best, Paul Cross-sectional recon-all -s ${SUBJECTS_DIR}/$session -autorecon-all -i ${filename} -hippocampal-subfields-T1 -parallel -openmp 20 -bigventricles -3T -time Longitudinal recon-all -base ${session} -tp "${tp1}" -tp "${tp2}" -all -parallel -openmp 20 -bigventricles -3T -time recon-all -long ${tp1} ${session} -all -parallel -openmp 20 -bigventricles -3T -time recon-all -long ${tp2} ${session} -all -parallel -openmp 20 -bigventricles -3T -time

2 2

Resample data for longitudinal comparisons
by James Gullickson 27 Aug '18

27 Aug '18

External Email - Use Caution All, I would like to compare cortical thickness between two timepoints. Timepoint 1 has 1mm^3 data. Timepoint 2 has 0.8mm^3 data. I realize that there are limitations with this change in resolution between timepoints and that no analysis is going to be perfect, but I am trying to come up with the best possible solution. Would it work to resample both timepoints' raw data into 0.9mm^3 ( mri_convert --conform_size 0.9 invol outvol ), and then run the 0.9mm data through Freesurfer 6.0 submilimeter recon? If so, are there any special considerations I need to take into account after the initial recon, or would longitudinal processing after this point be per normal? Thanks, James

2 1

Longitudinal Processing question
by Nillo, Ryan Michael R 27 Aug '18

27 Aug '18

External Email - Use Caution Hello FreeSurfer experts, I have a few questions on the longitudinal processing stream. I am interested in studying longitudinal morphometry changes in a pediatric population. The age range of this population is between 4-20 years with age deltas between a few months to < 2 years. I want to know what are appropriate/acceptable age deltas used for longitudinal processing. I looked through the archives and one researcher asked about age deltas. One of the answers was that “if the time delta is relatively small, this has good chances to work”. What does “relatively small” mean? If the time points exit the longitudinal pipeline without errors and the surfaces look correct, does this mean the data is relevant?? When I correct for head size, should I use the eTIV values from the cross-sectional directories or the longitudinal directories? Thanks in advance, Ryan Michael Nillo Staff Research Associate I University of California San Francisco Department of Radiology and Biomedical Imaging

2 1

← Newer
1
2
3
4
5
6
7
8
9
...
17
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 August 2018