The symmetric GTM is the default (the non-sym is not implemented).
I highly recommend that you don't use melzter (which is why I did not have docs
for it:). But if you must run it, add --melzter BinThresh MaskThresh nDilations
BinThresh is used to create a binary mask by thresholding the WM+GM tissue
fraction maps at BinThresh. This is a number between 0 and 1
nDilations is the number of times that the binary mask is dilated to expand it
a bit (don't use more than 2)
MaskThresh - the binarized mask is smoothed and then divided into the PET image
to do the correction. Voxels whose smoothed value is less than MaskThresh is
set to 0. MaskThresh should be between 0 and 1
this is all kind of messy and unprincipled and the results are very sensitive
to the values you set above . which is why I avoid it.
On 1/27/2020 9:23 PM, Kate Marvel wrote:
External Email - Use Caution
Thanks Douglas.
I checked the mri_gtmpvc --help command but the arguments for symmetric
Geometric Transfer Matrix and two-component model (Meltzer)
are missing. See the output below:
USAGE: mri_gtmpvc
--i inputvol : source data to PVC
--frame F : only process 0-based frame F from inputvol
--psf psfmm : scanner PSF FWHM in mm
--seg segfile : anatomical segmentation to define regions for GTM
--reg reg.lta : LTA registration file that maps anatomical to PET
--regheader : assume input and seg share scanner space
--o outdir : output directory
--mask volfile : ignore areas outside of the mask (in input vol space)
--auto-mask FWHM thresh : automatically compute mask
--no-reduce-fov : do not reduce FoV to encompass automask
--C contrast.mtx : univariate contrast to test (ascii text file)
--default-seg-merge : default schema for merging ROIs
--merge-hypos : merge left and right hypointensites into to ROI
--merge-cblum-wm-gyri : cerebellum WM gyri back into cerebellum WM
--tt-reduce : reduce segmentation to that of a tissue type
--replace Id1 Id2 : replace seg Id1 with seg Id2
--replace-file : file with a list of Ids to replace
--reg-identity : assume that input is in anatomical space
--rescale Id1 <Id2...> : specify reference region(s) used to rescale
(default is pons)
--no-rescale : do not global rescale such that mean of reference region is
scaleref
--scale-refval refval : scale such that mean in reference region is refval
--no-tfe : do not correction for tissue fraction effect (with --psf 0 turns
off PVC entirely)
--rbv : perform RBV PVC
--rbv-res voxsize : set RBV voxel resolution (good for when standard res
takes too much memory)
--mg gmthresh RefId1 RefId2 ...: perform Mueller-Gaertner PVC, gmthresh is
min gm pvf bet 0 and 1
--mg-ref-cerebral-wm : set MG RefIds to 2 and 41
--mg-ref-lobes-wm : set MG RefIds to those for lobes when using wm subseg
--mgx gmxthresh : GLM-based Mueller-Gaertner PVC, gmxthresh is min gm pvf
bet 0 and 1
--km-ref RefId1 RefId2 ... : compute reference TAC for KM as mean of given
RefIds
--km-hb RefId1 RefId2 ... : compute HiBinding TAC for KM as mean of given
RefIds
--ss bpc scale dcf : steady-state analysis spec blood plasma concentration,
unit scale
and decay correction factor. You must also spec --km-ref. Turns off
rescaling
--X : save X matrix in matlab4 format as X.mat (it will be big)
--y : save y matrix in matlab4 format as y.mat
--beta : save beta matrix in matlab4 format as beta.mat
--X0 : save X0 matrix in matlab4 format as X0.mat (it will be big)
--save-input : saves rescaled input as input.rescaled.nii.gz
--save-eres : saves residual error
--save-yhat : saves yhat
--save-yhat-with-noise seed nreps : saves yhat with noise, seed < 0 for TOD
--save-yhat-full-fov : saves yhat in full FoV (if FoV was reduced)
--save-yhat0 : saves yhat prior to smoothing
--synth gtmbeta C R S MBs seed nreps : synthesize volume with gtmbeta as
input
spec all other inputs the same; CRS are PSF for col, row, slice, MBs
motion blur slope
seed=0 for no noise, -1 for TOD seed. Turns on VFC for synth
but returns it to its specified value for analysis
--synth-only : exit after doing synthesis (implies --synth-save)
--synth-save : with --synth saves synthesized volume to outdir/synth.nii.gz
--threads N : use N threads (with Open MP)
--max-threads : use the maximum allowable number of threads for this computer
--max-threads-minus-1 : use one less than the maximum allowable number of
threads for this computer
--sd SUBJECTS_DIR
--vg-thresh thrshold : threshold for 'ERROR: LTAconcat(): LTAs 0 and 1 do
not match'
--gdiag diagno : set diagnostic level
--debug turn on debugging
--checkopts don't run anything, just check options and exit
--help print out information on how to use this program
--version print out version and exit
Thanks,
On Mon, Jan 27, 2020 at 6:03 PM Greve, Douglas N.,Ph.D.
<dgr...@mgh.harvard.edu<mailto:dgr...@mgh.harvard.edu>> wrote:
Run it with --help. Also look at the PETsurfer wiki page for typical usage
On 1/27/2020 3:54 PM, Kate Marvel wrote:
External Email - Use Caution
Hi Freesurfer community,
I would like to know the arguments to use for symmetric Geometric Transfer
Matrix and two-compartment model (Meltzer Method) in the petsurfer command
"mri_gtmpvc".
Thank you,
--
kate
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu<mailto:Freesurfer@nmr.mgh.harvard.edu>
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu<mailto:Freesurfer@nmr.mgh.harvard.edu>
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
--
kate
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu<mailto: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
