Hi Doug, I was told you were away last week so I'm reposting this message. Thanks for sending me the information on the PET scripts. I ran gtmseg, and bbregister successfully (I inspected both and they look accurate), but I am getting a segfault with mri_gtmpvc at the auto mask step.
I copied the command and the error below. I thought that it could be a memory issue due to the size of the matrix (I ran it on a mac with 8gb) but I also tried to run it with the --tt-reduce and got the same error. The data is from an HRRT PET scanner with a PSF of 4mm. For the auto-mask input I used 6 .01, as your instructions were PSF+2. Is this correct? Thanks Jonathan mri_gtmpvc --i pet.nii.gz --psf 4 --auto-mask 6 .01 --seg /Users/jonathandubois/data/fssub/epc_111102/mri/gtmseg.mgz --reg pet_avt_reg.lta --default-seg-merge --mgx .01 --o gtmpvc.output --km-hb 11 12 13 50 51 52 --km-ref 8 47 --no-rescale Loading input pet.nii.gz done loading input 26 frames $Id: mri_gtmpvc.c,v 1.57 2015/05/29 19:51:24 greve Exp $ setenv SUBJECTS_DIR /Users/jonathandubois/data/fssub cd /Volumes/my_passport/external/Documents/fssub2/test mri_gtmpvc --i pet.nii.gz --psf 4 --auto-mask 6 .01 --seg /Users/jonathandubois/data/fssub/epc_111102/mri/gtmseg.mgz --reg pet_avt_reg.lta --default-seg-merge --mgx .01 --o gtmpvc.output --km-hb 11 12 13 50 51 52 --km-ref 8 47 --no-rescale sysname Darwin hostname Jons-MacBook-Air.local machine x86_64 user jonathandubois vgthresh 0.001000 nReplace 18 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 4 avail.processors, using 1 Creating output directory gtmpvc.output Loading seg for gtm /Users/jonathandubois/data/fssub/epc_111102/mri/gtmseg.mgz Loading seg ctab /Users/jonathandubois/data/fssub/epc_111102/mri/gtmseg.ctab Reading /Users/jonathandubois/data/fssub/epc_111102/mri/gtmseg.lta Replacing 18 Pruning ctab done with seg vol maxFWHM = 3.28205 (voxels), PadThresh=0.0001, nPad=10 Computing auto mask Segmentation fault: 11 Message: 13 > Date: Tue, 08 Sep 2015 14:34:11 -0400 > From: Douglas N Greve <gr...@nmr.mgh.harvard.edu> > Subject: Re: [Freesurfer] PET tools in FS dev v6.0 > To: freesurfer@nmr.mgh.harvard.edu > Message-ID: <55ef2a23.3070...@nmr.mgh.harvard.edu> > Content-Type: text/plain; charset=windows-1252; format=flowed > Yes there are both. For KM we have MRTM1 and MRTM2 so far. They are not > well documented yet. > 1. To start, run > gtmseg --s subject > This will take a couple of hours and produces some files needed for GTM > PVC (which is used for GTM, MG, RBV). > 2. You'd then register the PET to the anatomical with bbregister (with > --t2 weighting). Make sure to save the output as an LTA (--lta). I > usually use the mean TAC as the input. You can do this in parallel with #1. > 3. You'd then run mri_gtmpvc, something like > mri_gtmpvc --i tac.nii.gz --psf PSF --auto-mask PSF+2 .01 --seg > gtmseg.mgz --reg reg.lta --default-seg-merge --mgx .01 --o gtmpvc.output > --km-hb 11 12 13 50 51 52 --km-ref 8 47 --no-rescale > PSF is the point-spread FWHM of the scanner; reg.lta is the registration > from #2. --km-hb specifies the highbinding region for MRTM2. --km-ref > specifies the reference region. --mgx specifies to output a > muller-gartner map (not necessary for GTM ROI analysis). > 4. For the GTM (ROI) MRTM1 KM analysis, you would then run > mri_glmfit --y km.hb.tac.nii.gz --mrtm1 km.ref.tac.dat time.dat --o > mrtm1 --no-est-fwhm --nii.gz > where time.dat is a text file withe acquisition time of each time point > in the tac. > 5. For the MRTM2 analysis > set k2p = `cat mrtm1/k2prime.dat` > mri_glmfit --y km.hb.tac.nii.gz --mrtm2 km.ref.tac.dat time.dat $k2p --o > mrtm1 --no-est-fwhm --nii.gz > If you want to run a voxel-wise analysis, then you can use the mgx > volume as input (--y). Probably you'll want to sample this to the > surface using mri_vol2surf and the registration file aux/anat2pet.lta, > then smooth on the surface. > > > doug > > > > On 09/08/2015 02:11 PM, Jonathan DuBois wrote: > > Hi FS experts, > > > > In FS development package I see that there have been tools added for > > partial volume correction of PET data, but I was wondering if there > > were also kinetic modeling tools similar to those described in Doug's > > 2014 paper? If you could let me know what PET processing tools have > > been included in the dev package, I would be interested to test and > > compare them. > > > > Thanks, > > Jonathan > > > > > > _______________________________________________ > > Freesurfer mailing list > > Freesurfer@nmr.mgh.harvard.edu > > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
