External Email - Use Caution
Hello,
I am trying to run the command trac-all -prep -c dmrircmultiscan.example -jobs prep.txt using a config file made for a cross-sectional study with multiple scans. Each subjects has 2 scans. When I run the command I get this error message.
length of pathway list (/proj/mihaliklab/usasoc/FREESURFER_HOME_LOC/trctrain/hcp/pathlist.txt) and endpoint label ID list ERROR: () do not match
Under the hood I noticed a gmids variable could be set. I this something I need to do for this data structure? I have previously not had issues using one scan.
Below is the config file text. Thank you in advance!
# Rotate diffusion gradient vectors to match eddy-current compensation? # Only used if doeddy = 1 or 2 # Default: 1 (yes) # set dorotbvecs = 1
# Intra-subject (diffusion-to-T1) registration method # 1: Affine with a correlation ratio cost # 2: Affine with a mutual information cost # 3: Affine with a boundary-based cost (default) # set intrareg = 3
# Degrees of freedom for intra-subject registration # Can be 6 (rigid), 9 (rigid+scaling), or 12 (full affine) # Default: 6 for infants, 9 otherwise # set intradof = 9
# Maximum rotation angle (degrees) for intra-subject registration # Default: 3 for infants, 90 otherwise # set intrarot = 90
# Inter-subject registration method # 1: Affine T1-to-T1 with a correlation ratio cost # 2: Affine T1-to-T1 with a mutual information cost # 3: Affine T1-to-T1 with a robust cost (default for infants) # 4: Nonlinear T1-to-T1 with CVS # 5: Nonlinear FA-to-FA with SyN (default) # set interreg = 5
# Target brain for inter-subject registration # Default for affine T1-to-T1: # $FSLDIR/data/standard/MNI152_T1_1mm_brain.nii.gz # Default for nonlinear T1-to-T1: # $FREESURFER_HOME/subjects/cvs_avg35/ # Default for nonlinear FA-to-FA: # $FREESURFER_HOME/trctrain/hcp/MGH35_HCP_FA_template.nii.gz # # I think this fixed my issue with syntax # set intertrg = /proj/mihaliklab/usasoc/FREESURFER_HOME_LOC/trctrain/hcp/MGH35_HCP_FA_template.nii.gz
# Whole-brain segmentation used to extract the anatomical neighborhood priors # of the pathways of interest # Must exist in each subject's FreeSurfer recon directory, under: # $SUBJECTS_DIR/$subjid/mri/$segname.mgz # Default: aparc+aseg # set segname = aparc+aseg
# Use the thalamic nuclei segmentation? # This is highly recommended to use for any pathways that terminate in # or neighbor the thalamus # When used, it is merged with the whole-brain segmentation above # Must exist in each subject's FreeSurfer recon directory, under: # $SUBJECTS_DIR/$subjid/mri/ThalamicNuclei.v12.T1.FSvoxelSpace.mgz # Default: 1 (yes) # set usethalnuc = 0
# Use brain mask extracted from T1 image instead of low-b diffusion image? # Has no effect if there is no T1 data # Default: 1 (yes) # set usemaskanat = 1
# Fractional intensity threshold for BET mask extraction from low-b images # This mask is used only if usemaskanat = 0 # Default: 0.3 # #set thrbet = 0.5
# Paths to reconstruct # Default: All paths in $FREESURFER_HOME/trctrain/hcp/pathlist.txt # set pathlist = /proj/mihaliklab/usasoc/FREESURFER_HOME_LOC/trctrain/hcp/pathlist.txt
# Number of path control points # It can be a single number for all paths or a different number for each of the # paths specified in pathlist (recommended, with more points for longer paths) # Default: As in $FREESURFER_HOME/trctrain/hcp/pathlist.txt #
set ncpts = /proj/mihaliklab/usasoc/FREESURFER_HOME_LOC/trctrain/hcp/pathlist.txt
# List of training subjects # This text file lists the locations of training subject directories # Default: $FREESURFER_HOME/trctrain/trainlist.txt # set trainfile = /proj/mihaliklab/usasoc/FREESURFER_HOME_LOC/trctrain/hcp/trainlist.txt
# Number of "sticks" (anisotropic diffusion compartments) in the bedpostx # ball-and-stick model # Default: 2 # set nstick = 2
# Number of MCMC burn-in iterations # (Path samples drawn initially by MCMC algorithm and discarded) # Default: 200 # set nburnin = 200
# Number of MCMC iterations # (Path samples drawn by MCMC algorithm and used to estimate path distribution) # Default: 7500 # set nsample = 7500
# Frequency with which MCMC path samples are retained for path distribution # Default: 5 (keep every 5th sample) # set nkeep = 5
# Reinitialize path reconstruction? # This is an option of last resort, to be used only if one of the reconstructed # pathway distributions looks like a single curve. This is a sign that the # initial guess for the pathway was problematic, perhaps due to poor alignment # between the individual and the atlas. Setting the reinit parameter to 1 and # rerunning "trac-all -prior" and "trac-all -path", only for the specific # subjects and pathways that had this problem, will attempt to reconstruct them # with a different initial guess. # Default: 0 (do not reinitialize) # set reinit = 0
Jake R. Powell, MS, LAT, ATC Doctoral Student - Human Movement Science
Matthew Gfeller Sport-Related Traumatic Brain Injury Research Center 2207 Stallings Evans Sports Medicine Center The University of North Carolina at Chapel Hill P 919-962-0409 E jrpowell@unc.edumailto:jrpowell@unc.edu