Freesurfer

freesurfer@nmr.mgh.harvard.edu

17 participants
30099 discussions

output from mri_glmfit
by Peggy Skelly 06 Jan '16

06 Jan '16

Hello Freesurfer statistical experts, I need some assistance in explaining the output generated by mri_glmfit (then followed by correction for multiple comparisions, mri_glmfit-sim). We fit one model and got some statistically significant clusters. Then, we ran another model, the same plus an additional nuisance variable. It had similar significant clusters. How do we compare the results from the 2 different models? Which model is better? In other statistic programs, we could examine an R^2 value, which can be interpreted as the percent of the response variable variation that is explained by the model. The model with the higher R^2 would be considered better since it fits the data better. But the output from mri_glmfit has rvar.mgh - the sum of the squares of the residual error divided by the DOF. So it would seem that a lower rvar would indicate a better fit. Can rvar.mgh be used to compute an R^2, because the R^2 concept is more widely understood. Or, is there a way to explain rvar in lay terms similar to R^2? Thanks, Peggy

2 1

how to display individual Desikan parcels?
by Mihaela Stefan 05 Jan '16

05 Jan '16

Hello Freesurfers, I would like to be able to display just a few parcels from the Desikan atlas on the fsaverage inflated surface. For example, I would like to display only the superiorfrontal and the posteriorcingulate. I tried to recreate the parcels myself by drawing labels, but they don't look pretty (see the attachment). I traced the labels in qdec (based on the Annotation/Opacity tab) and then I mapped them to fsaverage and visualized them with freeview. There's probably a better way to do that but I couldn't figure it out. Can anyone help me with this? Thanks! Mihaela

2 6

.Img format
by Hassan bakhshi 05 Jan '16

05 Jan '16

Hello , I have an image with .img format and an image with .hdr format. Freesurfer cant run them. What im suppose to di0o??? I just want to view the image.

2 1

mri_watershed for infants or NHP
by Mahmoud 05 Jan '16

05 Jan '16

Dear experts, I'm trying to adopt the HCP pipeline to our own data but at it is originally developed for human adults it fails frequently on my data. A big chunk of the pipeline is based on FreeSurfer so I thought it should be most relevant to ask my question here. the main problem I have here is that mri_watershed fails so the rest of the pipeline cannot proceed. The errors are usually like below: ****************TEMPLATE DEFORMATION**************** second estimation of the COG coord: x=128,y=131, z=124, r=39161 iterations ^^^^^^^^ couldn't find WM with original limits - expanding ^^^^^^ GLOBAL CSF_MIN=1, CSF_intensity=2, CSF_MAX=3 , nb = 38180 RIGHT_CER CSF_MIN=1, CSF_intensity=2, CSF_MAX=3 , nb = -1043712479 LEFT_CER CSF_MIN=1, CSF_intensity=2, CSF_MAX=3 , nb = 1117744481 RIGHT_BRAIN CSF_MIN=1, CSF_intensity=2, CSF_MAX=3 , nb = -1064196139 LEFT_BRAIN CSF_MIN=1, CSF_intensity=2, CSF_MAX=3 , nb = 1050954328 OTHER CSF_MIN=1, CSF_intensity=2, CSF_MAX=3 , nb = 1062732290 mri_watershed Error: GLOBAL region of the brain empty ! ################################################################## or *************************WATERSHED************************** Sorting... mri_watershed Error: Problem in the COG calculation: T1.mgz may not contain properly normalized white-matter. Confirm that T1.mgz has enough white-matter voxels of value 110. If white-matter contains few if any voxels of value 110, try adding wm control-points to nu.mgz, and re-run recon-all. ################################################################# I really appreciate any help or direction. Thank you! Mahmoud

1 0

Re: [Freesurfer] Individual regions of CorticalParcellation_Yeo2011
by Parker, Richard 05 Jan '16

05 Jan '16

Hi Professor Yeo, Sorry for the late reply. In the end I ended up giving in and performing my functional connectivity analysis in MNI space, like any rational person would. As a result, I ended up using the MNI152 versions of network masks you provided. Thanks for all your help, and for providing the split region ROI files! Best wishes, Richard Parker Richard Parker Postgraduate Research Student Maurice Wohl Clinical Neuroscience Institute King's College London Telephone: 07943515208 Email: Richard.r.parker(a)kcl.ac.uk

1 0

Freesurfer 6.0 -hires mri_segment potential issue
by Ajay Kurani 05 Jan '16

05 Jan '16

Hello Freesurfer experts, I am using the nightly build of Freesurfer 6.0 (Dec 18, 2015) on Ubuntu 14.04 with a high resolution image (0.5mm isotropic). I set OMP_NUM_THREADS=16 and then passed the -openmp 16 and -itkthreads 16 flags in my autorecon 2 line command. Autorecon1 and the beginning parts of Autorecon2 ran in a reasonable amount of time. Currently the program is running the mri_segment -keep -mprage brain.mgz wm.seg.mgz line command however it has been at this step for approximately 5 days! I used top line command to verify that the program is still running and it seems to be still running. I wasn't sure if this time for this step is typical for a high-res image (as this is my first time processing data higher than 1mm iso), or if there is a potential bug. I ran this same subject on another computer using the same version of Freesurfer and Ubuntu (8 core machine) and mri_segment (~ step 15) has been running for over a day so far. Any feedback would be appreciated as to if this amount of processing time is typical for this type of resolution, or if it could be a potential bug and that I should wait for a new beta release? Thanks, Ajay

2 1

Is the x-component of the b vectors extracted by mri_convert inverted?
by Emad Ahmadi 05 Jan '16

05 Jan '16

Hi FreeSurfer Experts, Happy New Year! I am using mri_convert to extract b vectors (bvecs) from diffusion images obtained at Bay 4. Then I use TRACULA for reconstructing white-matter pathways. The shape of the reconstructed pathways are odd, so I checked if the extracted bvecs are correct. To check if the extracted bvecs are correct, I used Diffusion Toolkit together with the extracted bvecs to do deterministic tractography. Then I checked the shape of the extracted nerve fibers. I have put a slide deck in the following dropbox folder showing that the shape of the extracted nerve fibers don’t look right: https://www.dropbox.com/s/1bt6on0rwm9url6/x_inversion.pptx?dl=0 <https://www.dropbox.com/s/1bt6on0rwm9url6/x_inversion.pptx?dl=0> Then I inverted the x-components of the extracted bvecs, and did the tractography again. This corrected the shape of the extracted nerve fibers (see the slide deck). I additionally had a look at the extracted dtift_V1 vectors under the two conditions, i.e. before and after x-component inversion. The results showed the same thing: x inversion corrects the images. I just wanted to see if others have had the same problem and confirm that I should invert the x-component of the bvecs extracted via mri_convert (I had the same experience with dcm2nii tool, where I had to invert the z-component of the extracted bvecs). I hope a wonderful new year for everyone! Emad Emad Ahmadi, MD ----------------------------------------------- Research Fellow in Radiology Massachusetts General Hospital Harvard Medical School 25 New Chardon Street, Suite 400 Boston, MA 02114 Tel: 617 726 5237 Email: emad(a)nmr.mgh.harvard.edu <mailto:emad@nmr.mgh.harvard.edu>

1 0

Data differ depending on OSX version
by Kasper Jessen 05 Jan '16

05 Jan '16

Dear freesurfer, As i understand from a previous forum thread, it is recommended that all the data is run on the same computer with the same operating system. However, what is the reason for the differing results when I am running the same 10 subjects on a MacPro and iMac respectively (please see below): *MacPro:* build-stamp.txt: freesurfer-Darwin-lion-stable-pub-v5.3.0 13.4.0 Darwin Kernel Version 13.4.0 *iMac: * build-stamp.txt: freesurfer-Darwin-lion-stable-pub-v5.3.0 11.4.2 Darwin Kernel Version 11.4.2 All the values are different. For example, the lh_caudalmiddlefrontal_thickness is on the iMac 2.472 and on the MacPro 2.537. Is it because of a randomization generator in FreeSurfer? Best regards, Kasper Jessen

2 3

recon-all get stucked
by salvatoreandrea.92＠libero.it 05 Jan '16

05 Jan '16

Hi Freesurfer experts! I am running recon-all with a Toshiba Satellite (16 Gb RAM and 4 CPU). My OS is Ubuntu 14.04 LTS (Trusty). Unfortunately recon-all gets stucked at the following: CORRECTING DEFECT 0 (vertices=30357, convex hull=5730) XL defect detected... I am using the Freesurfer development version, because I need to use Recon Edit. Attached you find the recon-all log. Best regards, Salvatore

2 1

control point edit
by maryam koochaki 04 Jan '16

04 Jan '16

Hi,I have some subjects which need some point seeding. For the first round they accept control points perfectly, but or more improvement they seem that they will not improve any more!!Should I leave it as it is? I attached one image as an example. Thanks,Maryam

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