Sorry, hit the send before I was ready. I was also going to point out that mri_glmfit will zero-out a voxel unless it finds that all inputs are non-zero (pruning). It looks like it could not find any voxels that met this criterion. doug
Douglas N Greve wrote:
Did the data set change between the two analyses? The
KingLab wrote:
Hello Doug- In going through the threads for freesurfer, I found the following which also describes the issue I am having. However, the error arises for me while running a collection of brains through qdec, so I have not been running the script directly in order to simply add the --no-prune tag. I am running the qdec analysis on new measurements created specifically for analysis in our lab. Someone prior to me set up everything for the qdec analysis to work fine with another database, and was halfway through another brain database when I picked up. For this database, the prior person completed the measurement entitled middle30.asc, which works fine. When using the same .dat file for the database, and the only difference is using the measure of middle45.asc, no data is mapped to the brains. In looking through the code, the same error appeared to happen in the mri_glmfit command in the same place. -qcache has been run for all of the brains for the new measurement, and the middle30.asc files appear similar in structure to the middle45.asc files, as would be expected. Why would it be pruning the one measurement and not the other? Where can I look to solve the issue to run the data in qdec to compare the measurements? If I just run mri_glmfit with --no-prune, will qdec read it in without overwriting it automatically? I have a feeling I missed something in preparing the new region that the person I took over for has done for the other measurements, but I'm not sure what it is or where to look for it. Any help would be appreciated. Thanks! Below I have included the output for both measurements. The codes appear identical up until after the pruning command comes about which provides 0 voxels. middle30.asc
lh-Avg-Intercept-middle30.asc -----------------------
Does the average middle30.asc differ from zero?
1.000 1.000;
lh-Diff-Normal-AD-Intercept-middle30.asc -----------------------
Does the average middle30.asc differ between Normal and AD?
1.000 -1.000;
ninputs = 188
Checking inputs
nframestot = 188
Allocing output
nframes = 188
Writing to /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/y.mgh
gdfReadHeader: reading /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/qdec.fsgd
INFO: demeaning continous variables
Continuous Variable Means (all subjects)
Class Means of each Continuous Variable
1 Diagnosis2Normal
2 Diagnosis2AD
INFO: gd2mtx_method is dods
Reading source surface /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/surf/lh.white
reading group avg surface area 822 cm^2 from file
Reading in average area /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/surf/lh.white.avg.area.mgh
Number of vertices 163842
Number of faces 327680
Total area 65416.648438
AvgVtxArea 0.399267
AvgVtxDist 0.721953
StdVtxDist 0.195470
$Id: mri_glmfit.c,v 1.138.2.12 2009/06/03 00:06:45 greve Exp $
cwd /Volumes/MacintoshHD2/ADNI-PreProcessed
cmdline mri_glmfit --y /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/y.mgh --fsgd /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/qdec.fsgd dods --glmdir /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled --surf fsaverage lh --label /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/label/lh.cortex.label --C /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/contrasts/lh-Avg-Intercept-middle30.asc.mat --C /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/contrasts/lh-Diff-Normal-AD-Intercept-middle30.asc.mat
sysname Darwin
hostname b417w74-106.med.utah.edu
machine i386
user kinglab
FixVertexAreaFlag = 1
UseMaskWithSmoothing 1
OneSampleGroupMean 0
y /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/y.mgh
logyflag 0
usedti 0
FSGD /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/qdec.fsgd
labelmask /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/label/lh.cortex.label
maskinv 0
glmdir /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled
IllCondOK 0
DoFFx 0
Creating output directory /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled
Loading y from /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/y.mgh
INFO: gd2mtx_method is dods
Saving design matrix to /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/Xg.dat
Matrix condition is 1
/Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/label/lh.cortex.label /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/label/lh.cortex.label
Found 149955 points in label.
Pruning voxels by thr: 0.000000
Found 149953 voxels in mask
Saving mask to /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/mask.mgh
search space = 75250.1
DOF = 186
Starting fit and test
Fit completed in 0.0740167 minutes
Computing spatial AR1 on surface
INFO: fwhmFromAR1: Fixing group surface area
Residual: ar1mn=0.998860, ar1std=0.000986, gstd=11.981826, fwhm=28.215043
Writing results
lh-Avg-Intercept-middle30.asc
maxvox sig=1e+10 F=86188.1 at index 0 0 0 seed=1324082303lh-Diff-Normal-AD-Intercept-middle30.asc
maxvox sig=9.81891 F=46.013 at index 70329 0 0 seed=1324082303mri_glmfit done
ninputs = 2
Checking inputs
nframestot = 2
Allocing output
nframes = 2
Writing to /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/contrasts.sig.mgh
reading group avg surface area 822 cm^2 from file
Reading in average area /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/surf/lh.white.avg.area.mgh
reading group avg surface area 1041 cm^2 from file
Reading in average area /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/surf/lh.pial.avg.area.mgh
reading group avg surface area 822 cm^2 from file
reading colortable from annotation file...
colortable with 36 entries read (originally /autofs/space/amaebi_026/users/buckner_cortical_atlas/scripts/colortable_final.txt)
reading colortable from annotation file...
colortable with 36 entries read (originally /autofs/space/amaebi_026/users/buckner_cortical_atlas/scripts/colortable_final.txt)
gdfReadHeader: reading /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/y.fsgd
INFO: ignoring tag Creator
INFO: ignoring tag SUBJECTS_DIR
INFO: ignoring tag SynthSeed
INFO: demeaning continous variables
Continuous Variable Means (all subjects)
Class Means of each Continuous Variable
1 Diagnosis2Normal
2 Diagnosis2AD
============================================================
Completed loading of analyzed data.
Middle45.asc
lh-Avg-Intercept-middle45.asc -----------------------
Does the average middle45.asc differ from zero?
1.000 1.000;
lh-Diff-Normal-AD-Intercept-middle45.asc -----------------------
Does the average middle45.asc differ between Normal and AD?
1.000 -1.000;
ninputs = 188
Checking inputs
nframestot = 188
Allocing output
nframes = 188
Writing to /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/y.mgh
gdfReadHeader: reading /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/qdec.fsgd
INFO: demeaning continous variables
Continuous Variable Means (all subjects)
Class Means of each Continuous Variable
1 Diagnosis2Normal
2 Diagnosis2AD
INFO: gd2mtx_method is dods
Reading source surface /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/surf/lh.white
reading group avg surface area 822 cm^2 from file
Reading in average area /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/surf/lh.white.avg.area.mgh
Number of vertices 163842
Number of faces 327680
Total area 65416.648438
AvgVtxArea 0.399267
AvgVtxDist 0.721953
StdVtxDist 0.195470
$Id: mri_glmfit.c,v 1.138.2.12 2009/06/03 00:06:45 greve Exp $
cwd /Volumes/MacintoshHD2/ADNI-PreProcessed
cmdline mri_glmfit --y /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/y.mgh --fsgd /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/qdec.fsgd dods --glmdir /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled --surf fsaverage lh --label /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/label/lh.cortex.label --C /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/contrasts/lh-Avg-Intercept-middle45.asc.mat --C /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/contrasts/lh-Diff-Normal-AD-Intercept-middle45.asc.mat
sysname Darwin
hostname b417w74-106.med.utah.edu
machine i386
user kinglab
FixVertexAreaFlag = 1
UseMaskWithSmoothing 1
OneSampleGroupMean 0
y /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/y.mgh
logyflag 0
usedti 0
FSGD /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/qdec.fsgd
labelmask /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/label/lh.cortex.label
maskinv 0
glmdir /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled
IllCondOK 0
DoFFx 0
Creating output directory /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled
Loading y from /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/y.mgh
INFO: gd2mtx_method is dods
Saving design matrix to /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/Xg.dat
Matrix condition is 1
/Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/label/lh.cortex.label /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/label/lh.cortex.label
Found 149955 points in label.
Pruning voxels by thr: 0.000000
Found 0 voxels in mask
Saving mask to /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/mask.mgh
search space = 0
DOF = 186
Starting fit and test
Fit completed in 0.0002 minutes
Computing spatial AR1 on surface
INFO: fwhmFromAR1: Fixing group surface area
WARNING: ar1 = nan <= 0. Setting fwhm to 0.
Residual: ar1mn=nan, ar1std=nan, gstd=0.000000, fwhm=0.000000
Writing results
lh-Avg-Intercept-middle45.asc
maxvox sig=0 F=0 at index 0 0 0 seed=1323815913lh-Diff-Normal-AD-Intercept-middle45.asc
maxvox sig=0 F=0 at index 0 0 0 seed=1323815913mri_glmfit done
ninputs = 2
Checking inputs
nframestot = 2
Allocing output
nframes = 2
Writing to /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/contrasts.sig.mgh
reading group avg surface area 822 cm^2 from file
Reading in average area /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/surf/lh.white.avg.area.mgh
reading group avg surface area 1041 cm^2 from file
Reading in average area /Volumes/MacintoshHD2/ADNI-PreProcessed/fsaverage/surf/lh.pial.avg.area.mgh
reading group avg surface area 822 cm^2 from file
reading colortable from annotation file...
colortable with 36 entries read (originally /autofs/space/amaebi_026/users/buckner_cortical_atlas/scripts/colortable_final.txt)
reading colortable from annotation file...
colortable with 36 entries read (originally /autofs/space/amaebi_026/users/buckner_cortical_atlas/scripts/colortable_final.txt)
gdfReadHeader: reading /Volumes/MacintoshHD2/ADNI-PreProcessed/qdec/Untitled/y.fsgd
INFO: ignoring tag Creator
INFO: ignoring tag SUBJECTS_DIR
INFO: ignoring tag SynthSeed
INFO: demeaning continous variables
Continuous Variable Means (all subjects)
Class Means of each Continuous Variable
1 Diagnosis2Normal
2 Diagnosis2AD
============================================================
Completed loading of analyzed data.
Thank you!
Kyle Hansen
Research Assistant - AIAL
University of Utah
Hi Kevin, the problem is that once you threshold for >0, there are no voxels that have non-zero values in all of the input subjects. Such voxels are "pruned" because there is not a good model for them. The GLM code does not do a different design at each voxels. In general, doing such thresholding is bad statistically. If you really want to do it, run it with --no-prune.
doug
kevin bickart wrote:
Hi all,
I am conducting a contrast of z (r) maps from 3 seed-based resting-state networks (e.g., 1 > 2 and 3). To eliminate the influence of anticorrelations on the contrast results, I thresholded the z (r) maps at 0 before running mri_glmfit, thereby eliminating voxels with negative z (r) values. The command completed successfully but the resulting sig maps did not contain voxels with any values. I have attached the fsgd and contrast files and pasted the command and output below. Can anyone identify the problem here? Thanks, Kevin
*Potentially problematic portion of output:* Pruning voxels by thr: 0.000000 Found 0 voxels in mask Saving mask to LR_Amy_lOFC2_glm2/mask.mgh search space = 0.000000
*Entire command and output: * [kreature:kbickart_fcmap_ASN_Anchor_AmyPks_buckner_YC] (nmr-stable5.1-env) mri_glmfit --y LR_Amy_lOFCpk2 _vs_vmPFCpk2_and_cACCpk2_pos_4D.nii.gz --allowsubjrep --fsgd /autofs/homes/006/kbickart/AmySubNet_glm/pa ired3.fsgd doss --C /autofs/homes/006/kbickart/AmySubNet_glm/paired.mat --glmdir LR_Amy_lOFC2_glm2 gdfReadHeader: reading /autofs/homes/006/kbickart/AmySubNet_glm/paired3.fsgd INFO: DeMeanFlag keyword not found, DeMeaning will NOT be done. Continuous Variable Means (all subjects) 0 roi1_vs_roi2_vs_roi3 0 0.707107 Class Means of each Continuous Variable 1 Class_081028_BC77WB 0.0000 2 Class_081028_XE45TK 0.0000 3 Class_081028_YY28TK 0.0000 .... $Id: mri_glmfit.c,v 1.196.2.6 2011/05/05 20:54:25 greve Exp $ cwd /autofs/cluster/animal/rs-fmri/YC/GROUP_STATS/kbickart/FCMAP/kbickart_fcmap_ASN_Anchor_AmyPks_buckne r_YC cmdline mri_glmfit --y LR_Amy_lOFCpk2_vs_vmPFCpk2_and_cACCpk2_pos_4D.nii.gz --allowsubjrep --fsgd /autof s/homes/006/kbickart/AmySubNet_glm/paired3.fsgd doss --C /autofs/homes/006/kbickart/AmySubNet_glm/paired .mat --glmdir LR_Amy_lOFC2_glm2 sysname Linux hostname kreature machine x86_64 user kbickart FixVertexAreaFlag = 1 UseMaskWithSmoothing 1 OneSampleGroupMean 0 y /autofs/cluster/animal/rs-fmri/YC/GROUP_STATS/kbickart/FCMAP/kbickart_fcmap_ASN_Anchor_AmyPks_buckn er_YC/LR_Amy_lOFCpk2_vs_vmPFCpk2_and_cACCpk2_pos_4D.nii.gz logyflag 0 usedti 0 FSGD /autofs/homes/006/kbickart/AmySubNet_glm/paired3.fsgd glmdir LR_Amy_lOFC2_glm2 IllCondOK 0 ReScaleX 1 DoFFx 0 Creating output directory LR_Amy_lOFC2_glm2 Loading y from /autofs/cluster/animal/rs-fmri/YC/GROUP_STATS/kbickart/FCMAP/kbickart_fcmap_ASN_Anchor_Am yPks_buckner_YC/LR_Amy_lOFCpk2_vs_vmPFCpk2_and_cACCpk2_pos_4D.nii.gz llINFO: gd2mtx_method is doss Saving design matrix to LR_Amy_lOFC2_glm2/Xg.dat Normalized matrix condition is 1 Matrix condition is 44.5 Pruning voxels by thr: 0.000000 Found 0 voxels in mask Saving mask to LR_Amy_lOFC2_glm2/mask.mgh search space = 0.000000 DOF = 177 Starting fit and test Fit completed in 0.0148667 minutes Computing spatial AR1 in volume. fMRIspatialAR1(): hit 0 voxels WARNING: no voxels in AR1 computation Residual: ar1mn = (nan,nan,nan) fwhm = (nan,nan,nan) nan Writing results paired maxvox sig=0 F=0 at index 0 0 0 seed=1321328111 mri_glmfit done
-- Kevin Bickart MD-PhD Student Department of Anatomy and Neurobiology Boston University School of Medicine
live UP to it
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu mailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center gr...@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html http://www.nmr.mgh.harvard.edu/facility/filedrop/index.html
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu mailto: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.