Freesurfer

freesurfer@nmr.mgh.harvard.edu

19 participants
30088 discussions

Re: [Freesurfer] 4D multi-frame parcellation
by Greve, Douglas N.,Ph.D. 18 Nov '19

18 Nov '19

If you have an anatomical MR that you have run through recon-all, then just use that one instead of bert in your steps below ps. Please remember to post to the FS list and not to us personally On 11/12/19 8:41 PM, an wrote: > > External Email - Use Caution > > Many thanks. > > I got the 4D fMRI dataset from others and it is not raw data. In order > to run mri_vol2surf, I need to calculate the register.dat for the > required flag --srcreg so I tried in this way. > > FYI, the 4D fMRI dataset should have been registered with its > anatomical MR volume by using other tools. > > For now I only ran the recon-all on its anatomical MR volume without > running the functional stream as I don't have the raw data. > > Should I also run the raw fMRI data from scratch by using functional > stream? Or if I can run mri_vol2surf independently, which command > should I use to get a register.dat file? > In addition, the intensity of the 4D fMRI dataset I have is between > [-1, 1] and not integers, would this cause any problem? > > Best, > An > > > On 11/12/19 10:41 AM, Greve, Douglas N.,Ph.D. wrote: >> Don't do the 1st step. >> Why are you registering it to bert? I'm pretty sure bert did not >> participate in your fmri study >> When you run mri_vol2surf, it will probably work better with >> --projfrac 0.5 >> >> On 11/11/2019 1:32 PM, An wrote: >>> >>> External Email - Use Caution >>> >>> Hi Prof. Greve, >>> >>> Thanks for your reply and sorry for the confusion. >>> >>> I have a 4D fMRI dataset with 10 time frames, where the intensity of >>> each voxel is between [-1,1] and not integers. I want to map the >>> intensities of the volumes on its corresponding surface in order to >>> get the functional values on each vertex in each frame. I have >>> already run its corresponding anatomical image in freesurfer. >>> >>> To achieve it, I tried the following steps: >>> 1. conform the fmri series by using mri_convert >>> 2. register the fmri series with the anatomy image by using >>> bbregister: bbregister --mov /4Dvol.nii/ --s bert --reg register.dat >>> 3. assign values from volumes to each vertex by using mri_vol2surf: >>> mri_vol2surf --src /4Dvol.nii /--out/lhtest.mgz/ --srcreg >>> register.dat --hemi lh >>> I also tried the -regheader in mri_vol2surf: mri_vol2surf --src >>> /4Dvol.nii /--out/lhtest.mgz/ --regheader bert --hemi lh >>> >>> >>> The /4Dvol.nii /aligns very well with the orig.mgz in freeview. But >>> after step2, the registered volume looks wrong. I am wondering could >>> I use bbregister to register multi-frame fMRI series to a single >>> volume? >>> I also tried to use --regheader to replace the --srcreg file in >>> mri_vol2surf as the output register.dat in step2 is wrong. But the >>> output is still wrong. >>> >>> Many thanks. >>> >>> Best, >>> An >>> >>> Greve, Douglas N.,Ph.D. <DGREVE(a)mgh.harvard.edu >>> <mailto:DGREVE@mgh.harvard.edu>> 于2019年11月11日周一下午12:16写道： >>> >>> When you say it is 4D, what do you mean? That each label has its >>> own frame? >>> >>> On 11/6/19 5:59 PM, 曲岸 wrote: >>> > >>> > External Email - Use Caution >>> > >>> > Hi there, >>> > >>> > I am working with creating a surface parcellation from a >>> volumetric >>> > parcellation and have some questions as follows: >>> > >>> > 1. If the volumetric parcellation is a labeled 4D multi-frame >>> dataset, >>> > could I still use the mris_sample_parc? Or could mri_vol2surf >>> get the >>> > sampled labels for every vertex? >>> > >>> > 2. For the 4D multi-frame dataset and mri_vol2surf output >>> results, >>> > which GUI should I use to visualize it? I tried freeview for 4D >>> > multi-frame dataset but it looks weird. >>> > >>> > 3. Is there any command to separate multi-frame volume to >>> single frame >>> > volumes? >>> > >>> > Many thanks! >>> > >>> > >>> > Best, >>> > An >>> > >>> > _______________________________________________ >>> > Freesurfer mailing list >>> > Freesurfer(a)nmr.mgh.harvard.edu >>> <mailto:Freesurfer@nmr.mgh.harvard.edu> >>> > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>> >>> >>> _______________________________________________ >>> Freesurfer mailing list >>> Freesurfer(a)nmr.mgh.harvard.edu >>> <mailto:Freesurfer@nmr.mgh.harvard.edu> >>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>> >>> >>> _______________________________________________ >>> Freesurfer mailing list >>> Freesurfer(a)nmr.mgh.harvard.edu >>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >> >> >> _______________________________________________ >> Freesurfer mailing list >> Freesurfer(a)nmr.mgh.harvard.edu >> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >

3 9

Re: [Freesurfer] use of .m3z transform output from mri_cvs_register
by Vinny K 18 Nov '19

18 Nov '19

External Email - Use Caution Hi, I ran the mri_cvs_register between two subjectIDs with very good alignment between them. I had assumed that applying the outputted m3z transform, i.e.,final _CVSmorph_toTEMPLATE.m3z, to the moving subjectID (subjb) would normalize it to the target subjectID (subja). That is, the outputted image would be the same as the output image from mri_cvs_register ,final_CVSmorphed_toTEMPLAT_norm.mgz; however, the resulting image was totally out of alignment to the target subjectID. This is the command I had utilized to apply the m3z transform to the moving subjectID: mri_convert -at /home/vinny/Tools/freesurfer/subjects/subjb/cvs/final_CVSmorph_tosubja.m3z subjb.nii.gz subjb-to-subja_morph.nii.gz Thanks for your help, Vinny

1 1

Brain lobules pooling- request
by Samson Nivins 18 Nov '19

18 Nov '19

External Email - Use Caution Hi FreeSurfer Team, I am interested in individual lobular volume and thickness. So I have extracted based on the Q and A I created few script to run and extract the data. However, I got few questions regarding thickness and WM For extracting lobular GM I used mri_annotation2label --subject 1014 --hemi lh --lobesStrict lh.lobesStrict.annot mri_annotation2label --subject 1014 --hemi lh --annotation lobesStrict --outdir '/home/samson/Volume/freesurfer/subjects/recon-all_working/lh' for computing stats I used mris_anatomical_stats -l '/home/samson/Volume/freesurfer/subjects/recon-all_working/lh/lh.occipital.label' -b 1014 lh I am not sure, whether any command is there to export the data. And, the average cortical thickness which we obtain in the terminal window during individual extraction, does it correspond to the lobular thickness or as whole For example I extracted stats for left frontal cortex: I got the following details in Terminal window Number of vertices, Total surface area, total gray matter, average cortical thickness .... In this average cortical thickness - correspond to the ROI specifc are as a whole? Thanks for you help. Looking forward to hearing from you Cheers Best regards, Sam Samson Nivins | PhD Student A The University of Auckland, Liggins Institute Building 505, Level 2, 85 Park Road, Auckland Ph +(64) 21 256 5964 E samson.nivins(a)gmail.com<mailto:samson.nivins@gmail.com> [cid:image001.png@01D4C1F6.9273DF90]

2 1

Question about the input for longitudinal processing pipeline
by Darko Komnenić 18 Nov '19

18 Nov '19

External Email - Use Caution Dear Freesurfer experts, I wanted to run the Freesurfer's longitudinal processing pipeline (described here https://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalProcessing) on a series of MRI scans from a single patient. However, after looking at the scans, i realized that for most time points, the patient doesn't have a "simple/normal" T1 MPRAGE sequence available. Instead, they have a T1 sequence with a contrast agent, as well as T1 mp2rage sequences. Is it possible/advisable to run recon-all in general or this pipeline in particular on T1 MPRAGE scans that have a contrast agent, or that are mp2rage instead of MPRAGE? Thanks in advance! Best, Darko

2 1

Cortical thickness values on a vertex base level
by Aicha Dijkshoorn 18 Nov '19

18 Nov '19

External Email - Use Caution Dear developers, For a case-control study we are comparing the cortical thickness and cortical surface area on a vertex-base level, for which we found significant differences throughout the brain. Is it possible that, in addition to the significance values (p < .05 after FDR correction) on a vertex-base level, we can include the exact thickness or surface area value (or difference) of the significant clusters on a vertex base-level between the groups. In order words, to provide some measures to highlight the size of the significant cluster or the differences in cortical thickness / cortical surface area to provide clinically useful context. For instance, a command comparable to 'voxel size' in TBSS. As this is my first time using Freesurfer any help that you could offer on this matter would be highly appreciated. Best wishes, Brigitte Dijkshoorn

2 5

Questions for PetSurfer
by Soo-Jong Kim 18 Nov '19

18 Nov '19

External Email - Use Caution Dear FreeSurfer experts, I am a student in neuroimaging using FreeSurfer. I have some questions and I don't know where to send questions. That's why I found this mail and writing. I've studied PetSurfer for partial volume correction of PET. I saw the tutorial in website. Originally, I want to use RBV method for pvcorrected PET for SUVR analysis. After generating gtmseg and coregistration to make lta file, I followed this code and changed some. *mri_gtmpvc --i PET.nii --reg subj1.reg.lta --psf 5 --seg gtmseg.mgz * *--default-seg-merge --auto-mask PSF .01 --no-rescale --no-reduce-fov --rbv --o output_pvc* After this, rbv.nii.gz file was generated. and SUVR analysis was performed as reference region (cerebellar cortex). But Compared to original un-pvc PET, Cortex SUVR was reduced... In ideal pv-corrected PET, cortex SUVR was higher than un-pvc PET. Is is okay if I use the output file as rbv.nii.gz ? in Thomas et al, 2011, He used aparc+aseg.mgz file from FreeSurfer. and he merged some regions. Frontal, Temporal, Occipital and so on. (Only gray matter to correct PET using RBV method) In that case, What are the regions to make rbv.nii.gz method in PetSurfer? Can I use the rbv.nii.gz as partial volume corrected PET from RBV method? If not, Please tell me what is the correct command of mri_gtmpvc. I need only partial volume corrected PET using RBV method. Sincerely, Soo-Jong Kim

2 1

Multiple comparisons
by Juan Rivas 18 Nov '19

18 Nov '19

External Email - Use Caution I would like to know how the correction is made for multiple comparisons in the statistical analysis? We are using QDEC, with False Discovery Rate or Montecarlo Null-Z. Is there a better way to do that? Best, JC.

2 1

Question about design matrix for surface group analysis
by 邓刊 18 Nov '19

18 Nov '19

External Email - Use Caution Dear freesurfer users, I have a question about the design matrix.We want to analyse the relationship between the cortical thickness and the neuropsychological test. We have two groups:TLE and extralTLE,one discrete variable:gender,and two continuous variable: neuropsychological test and age. we make the FSGD files like this:Title OSGM Class MaleTLE Class MaleextraTLE Class FemaleTLE Class FemaleexatrTLE Variables Age Neuropsychology Input subject1 MaleTLE 30 5 Input subject2 MaleextraTLE 40 6 Input subject3 FemaleTLE 50 8 Input subject4 FemaleLextraTLE 60 7 Is it right? And we want to analyze the relationship between the cortical thickness and the neuropsychology regressing out the effects of gender and age, how to make the design matrix? Look forward to your reply. Sincerely, Kan Deng

2 3

Spatial smoothing for cortical folding measures
by huyang 18 Nov '19

18 Nov '19

External Email - Use Caution Dear FS experts, Is it necessary to do spatial smoothing for cortical folding measures (curv/sulc/smoothwm.FI.crv/pial_lgi, etc.)? These cortical folding measures are inherently smooth based on the computation methods used. If spatial smoothing is still needed, what's the recommended smoothing kernel size? Thanks! Yang Hu

3 2

Processor speed/cores vs cache size
by Pradyumna Bharadwaj 18 Nov '19

18 Nov '19

External Email - Use Caution Hello, For running recon-all or tracula on a local workstation, is it recommended to go for a higher processor speed/number of cores, or a larger cache size for the processor (Intel Xeon W series). Or is it likely to be a more complex interplay between the number of recon-all processes running at any given time? Thanks, Pradyumna

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