Freesurfer

freesurfer@nmr.mgh.harvard.edu

4 participants
30090 discussions

'Best practice' for dealing with lesioned brains
by King, Daniel (Research Student) 11 Jul '17

11 Jul '17

Dear Freesurfer Team, I am attempting to complete a surface reconstruction on a dataset with traumatic lesions both to the grey and white matter. I am hoping to get some opinions on what would be considered as the 'Best practice' pipeline for dealing with such a dataset. We have access to T1 and T2-FLAIR images for a moderately large dataset and thus require a pipeline as fully-automated as is possible whilst still relatively accurate. I have seaerched the mailing list archives but many just relate to white matter lesions in MS for example. My current thoughts were to use the T1 and Fair images with the -FLAIRpial flag in recon-all to improve pial segmentation, but also using an automated lesion segmentation tool to conduct lesion filling. However, I know this is more effective for white than Gray matter lesions due to the greater intensity variability in the Gray matter compared to the white matter. Any reccomendations would be greatly appreciated. Cheers Dan

1 0

Extract smoothed cortical thickness whole-brain vertex-wise
by Nicolette Schwarz 10 Jul '17

10 Jul '17

Hello everyone, I'm pretty new to FS and can't seem to find the solution for my problem. I am trying to write whole-brain vertex-wise cortical thickness (CT) data from each participant into ascii files. In my original analysis, each participant's average CT in clusters derived from a whole-brain GLM were extracted and used in a second analysis. I'd like to extract CT vertex-wise to perform a whole-brain version of this second analysis. I need the smoothed and "fsaveraged" (to get the same # of vertices across participants in a template space) CT values to match the cluster data and that seems to be tricky. This works: mris_convert -c lh.thickness lh.sphere.reg test1.thickness.asc This doesn't work: mris_convert -c lh.thickness.fwhm10.fsaverage.mgh lh.sphere.reg test2.lh.thickness.asc # ERROR: number of vertices in lh.thickness.fwhm10.fsaverage.mgh does not match surface (163842,124874) I presume the issue is that ?h.thickness.fwhm10.fsaverage.mgh is a volume file while ?h.sphere.reg is a surface file. So I've tried to convert volume to surface (created a folder test1/surf and copied both fsaverage/surf/lh.sphere.reg and sub/surf/lh.thickness.fwhm10.fsaverage.mgh there) using different combinations of mri_vol2surf, e.g.: mri_vol2surf --srcsubject test1 --src lh.thickness.fwhm10.fsaverage.mgh --trgsubject test1 --surfreg lh.sphere.reg --regheader test1 --hemi lh --o test.surf This doesn't work as it wants me to specify a type but I have no clue what the type for the surface is (or better, it doesn't seem to be an option going by mris_convert). Other combinations asked for --reg files and I played with tkmedit2 but none of this is working out. Also, is ?h.sphere.reg an appropriate choice and would it have to be smoothed to FWHM=10 prior to mris_convert? Trying to solve this has me going in all kinds of circles and I'm unsure about this approach anyway. Hence, I would truly appreciate any help. Thank you very much in advance, Nic -- Nicolette Schwarz, Ph.D. Postdoctoral Research Fellow http://heartbrain.hms.harvard.edu

2 5

questions mri_glmfit-sim
by Van Der Deijl, Rosanne 10 Jul '17

10 Jul '17

Dear Freesurfer team, I have two more questions about mri_glmfit-sim. 1. In the summary file I get a list of all the inputs/outputs, but I don't understand what all of them mean. Could you explain the ones I highlighted in bold? 2. Is it possible to a priori set the area threshold, so that I only get clusters larger than a certain size? # Input rh.CALMlong.thickness-spc.fwhm15/CALMlong_contrast2/sig.mgh # Frame Number 0 # srcsubj fsaverage # hemi rh # surface white # group_avg_surface_area 82167.6 # group_avg_vtxarea_loaded 1 # annot aparc # SUBJECTS_DIR /local_mount/space/avadhuti/1/users/gdesbord/CALM_data// # SearchSpace_mm2 75762.5 # SearchSpace_vtx 148554 # Bonferroni 0 # Minimum Threshold 1.3 # Maximum Threshold infinity # Threshold Sign abs # AdjustThreshWhenOneTail 1 # Area Threshold 0 mm^2 # CSD thresh 1.300000 # CSD nreps 10000 # CSD simtype null-z # CSD contrast NA # CSD confint 90.000000 # Overall max 1.70172 at vertex 163606 # Overall min -2.48953 at vertex 90839 # NClusters 32 # FixMNI = 0 Thank you very much in advance. Kind regards, Rose

1 0

error with mri_glmfit --frame-mask
by Sadie Marvel 10 Jul '17

10 Jul '17

?Hi list, I am trying to run mri_glmfit on a group of subjects using the --frame-mask flag because I want the design to be changed based on which subjects have a value at a single voxel. This is because I have one subject in my group whose signal was cut off through parts of their occipital and temporal lobes. This is the command I am using: mri_glmfit --y ces.nii --wls cesvar.nii --osgm --no-prune --frame-mask masks.nii --surface fsaverage rh --glmdir my-glm2.wls --nii.gz? where masks.nii is a file that contains each individual's binary mask. However, when I use this command, I get this error: ERROR: MRInormWeights: value less than or eq to 0.? which I assumed may have something to do with my ces/cesvar files being a different size than my masks file; however, all three of the files have the same dimensions (163842x1x1x26 where 26 is the number of subjects in my group) Any idea why I am getting this error? I also tried creating a file of binary masks by first concatenating ces.nii and cesvar.nii and binarizing it, but I am getting the same error. Thanks!

1 0

Fwd: surface inaccuracies around the optic chiasm
by Gabor Perlaki 10 Jul '17

10 Jul '17

Somehow, my last email didn't appear on the Freesurfer forum. I'm resending it without the attachments. Dear Bruce, I'm not interested in the chiasm, but the medialorbitofrontal cortex would be interesting. In most of the subjects pial and white surfaces intrude into the area around the optic chiasm, which is clearly not a cortical region. However, I think that until this region is segmented as a subcortical label (i.e. opic chiasm) or nothing in the aparc.DKTatlas+aseg.mgz file my cortical measures won't be affected, but if it's labeled as medialorbitofrontal cortex or wm then I should correct it. 1. Could you please confirm that these surface inacuracies don't matter as long as the area is not segmented as a cortical region (i.e. medialorbitofrontal) or white matter in the aparc.DKTatlas+aseg.mgz file? 2. I think I should correct the wm.mgz, but I'm not sure about how to correct it without loosing my earlier corrections. I'm ready with all of the other corrections inclding wm, brainmask and aseg corrections as well? I'm trying to resend some images showing the missegmented area. Best Regards, Gabor

2 1

Re: [Freesurfer] help with code interpretation
by Leanne Hunt 10 Jul '17

10 Jul '17

Hi, I am currently using free surfer to analyse my brain imaging and i was wondering if you could tell me what G_S stands for in the rh.aparc.a2009s.stats and lh.aparc.a2009s.stats folders? for example: G_and_S_frontomargin Many thanks Leanne

2 2

surface inaccuracies around the optic chiasm
by Gabor Perlaki 09 Jul '17

09 Jul '17

Dear all, I use Freesurfer 6. In some of my subjects the surfaces (both pial and white) intruding into the optic chiasm. However, when I check the aparc.DKTatlas+aseg.mgz it seems that although the surfaces are wrong, the region is segmented as optic chiasm. Can I ignore this type of surface inaccuracies similarly to surface inacurracies around the amygdala and hippocampus? Best Regards, Gabor

2 3

recon-all on newborn
by Das S. 08 Jul '17

08 Jul '17

Dear Freesurfer expert, I want to perform recon-all pipeline on neonates data.Can I do so? I was going through some mail archives and there it is said that on T1 image of neonates data we can't perform recon-all. But again there is another paper where it says Freesurfer can perform segmentation of neonates image. Any input will be appreciated. Thanks S Das

4 10

convert .mgz to .nii format
by Almarshad F. 08 Jul '17

08 Jul '17

Hi everyone; is it possible to convert T1.mgz back to T1.nii.gz I'm using the examples in the tutorial data which does not have the original T1.nii i tried to use the command mri_convert but i don't think the file it created is correct ! any advice regards to this issue will be very helpful .. kindest regards, Fatima

2 1

Regarding_ Midbrain_misalignment
by Dr Sampada Sinha 07 Jul '17

07 Jul '17

Dear freesurfer experts, I am trying to over brainstem mask of my patients' group on mmi_colin_t1hires atlas. I had no problems in overlaying thalamus or amygdala mask, but somehow the midbrain mask is shifted from the actual place (please see attachment). Can you please let me know how do I correct this error? Thanks for all your help. Kind regards, Sam -- Dr Sampada Sinha Research medical scientist (B) Department of Geriatric Mental Health (DGMH) King George Medical University Lucknow-226003 , Uttar Pradesh India

3 7

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