[Freesurfer] mri_glmfit error: negative DOF -10
Hello, I am using mri_glmfit to perform a correlation analysis between some surface-based DTI measures and a behavioural outcome (grip), aiming to correct for scannertyp (C,I,W), handedness (R,L) as discrete variables and age (AGE) and duration (Mo) as continuous covariates. Please find the fsgd file, the contrast file and the calculated Xg down below. However, I get the follow error: DOF = -10 ERROR: DOF = -10 (the complete output is attached) I would be grateful if somebody could help. I already tried it with demeaned data, z-transformed data, it doesn't help. Thank you for your help, Robert fsgd file: GroupDescriptorFile 1 Title OSGM Class CR Class CL Class WL Class WR Class IR Class IL Variables Age Mo Grip Input c02 CR 52 85 0.250 Input c04 CR 71 27 0.652 Input c05 CL 63 12 0.788 Input c06 WL 77 26 0.438 Input c07 WR 61 13 0.717 Input c10 IR 68 9 0.730 Input c11 IL 63 13 0.934 Input c12 IL 29 12 1.170 Input c13 IR 46 12 0.964 Input c14 IR 75 14 0.655 Input c15 IL 68 13 1.098 Input c16 IR 69 12 1.028 Input c18 WL 40 20 0.698 Input c19 WL 53 13 0.547 contrast file: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 design matrix: +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +52.0 +0.0 +0.0 +0.0 +0.0 +0.0 +85.0 +0.0 +0.0 +0.0 +0.0 +0.0 +0.25000 +0.0 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +71.0 +0.0 +0.0 +0.0 +0.0 +0.0 +27.0 +0.0 +0.0 +0.0 +0.0 +0.0 +0.65200 +0.0 +0.0 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +63.0 +0.0 +0.0 +0.0 +0.0 +0.0 +12.0 +0.0 +0.0 +0.0 +0.0 +0.0 +0.78800 +0.0 +0.0 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +77.0 +0.0 +0.0 +0.0 +0.0 +0.0 +26.0 +0.0 +0.0 +0.0 +0.0 +0.0 +0.43800 +0.0 +0.0 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +61.0 +0.0 +0.0 +0.0 +0.0 +0.0 +13.0 +0.0 +0.0 +0.0 +0.0 +0.0 +0.71700 +0.0 +0.0 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +68.0 +0.0 +0.0 +0.0 +0.0 +0.0 +9.0 +0.0 +0.0 +0.0 +0.0 +0.0 +0.73000 +0.0 +0.0 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +63.0 +0.0 +0.0 +0.0 +0.0 +0.0 +13.0 +0.0 +0.0 +0.0 +0.0 +0.0 +0.93400 +0.0 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +29.0 +0.0 +0.0 +0.0 +0.0 +0.0 +12.0 +0.0 +0.0 +0.0 +0.0 +0.0 +1.17000 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +46.0 +0.0 +0.0 +0.0 +0.0 +0.0 +12.0 +0.0 +0.0 +0.0 +0.0 +0.0 +0.96400 +0.0 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +75.0 +0.0 +0.0 +0.0 +0.0 +0.0 +14.0 +0.0 +0.0 +0.0 +0.0 +0.0 +0.65500 +0.0 +0.0 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +68.0 +0.0 +0.0 +0.0 +0.0 +0.0 +13.0 +0.0 +0.0 +0.0 +0.0 +0.0 +1.09800 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +69.0 +0.0 +0.0 +0.0 +0.0 +0.0 +12.0 +0.0 +0.0 +0.0 +0.0 +0.0 +1.02800 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +40.0 +0.0 +0.0 +0.0 +0.0 +0.0 +20.0 +0.0 +0.0 +0.0 +0.0 +0.0 +0.69800 +0.0 +0.0 +0.0 +0.0 +0.0 +1.0 +0.0 +0.0 +0.0 +0.0 +0.0 +53.0 +0.0 +0.0 +0.0 +0.0 +0.0 +13.0 +0.0 +0.0 +0.0 +0.0 +0.0 +0.54700 +0.0 +0.0 +0.0 output: gdfReadHeader: reading /home/schulz/dti/premotorstroke/wm-probe/fs_no3/glmout_corr_l/glmout/corr.fsgd INFO: DeMeanFlag keyword not found, DeMeaning will NOT be done. Continuous Variable Means (all subjects) 0 Age 59.6429 13.4462 1 Mo 20.0714 18.7444 2 Grip 0.762071 0.248599 Class Means of each Continuous Variable 1 CR 61.5000 56. 0.4510 2 CL 63. 12. 0.7880 3 WL 56.6667 19.6667 0.5610 4 WR 61. 13. 0.7170 5 IR 64.5000 11.7500 0.8442 6 IL 53. 12.6667 1.0673 INFO: gd2mtx_method is dods Reading source surface /home/schulz/freesurfer/subjects/premotorstroke/fsaverage/surf/lh.white Number of vertices 163842 Number of faces 327680 Total area 65416.648438 AvgVtxArea 0.399267
[Freesurfer] mri_vol2surf projection of white matter results
Dear Freesurfers, I transferred the white matter binary mask (derived from aseg, volume) to FSL and performed probabilistic tractography in it. Now I'd like to project the 'outer' white matter results (vol) onto the white surface. To test for the acuracy of this approach, I tried to project the white matter mask (binarized, volume) itself on the surface, proper projection should give only 1s (ones) on the surface, which holds true for the majority of vertices. I used the following command: mri_vol2surf --mov whitevolume.mgz --reg struct2freesurfer.dat --projdist-max 0 3 0.1 --interp nearest --hemi rh --out ... Is there a way to optimize this? The binarized white matter data (volume) is just below the white matter surface in most cases, sometimes, voxels are also above the surface. If I try trilinear I get values far away from 1 (hence values which are not included in the primary data (=binarized mask). Looking forward to your suggestions. Thank you very much, Best Robert -- Besuchen Sie uns auf: www.uke.de _ Universitätsklinikum Hamburg-Eppendorf; Körperschaft des öffentlichen Rechts; Gerichtsstand: Hamburg Vorstandsmitglieder: Prof. Dr. Martin Zeitz (Vorsitzender), Prof. Dr. Dr. Uwe Koch-Gromus, Joachim Prölß, Rainer Schoppik _ SAVE PAPER - THINK BEFORE PRINTING ___ 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.
[Freesurfer] mri_vol2surf projection of white matter results
Dear Freesurfers, I transferred the white matter binary mask (derived from aseg, volume) to FSL and performed probabilistic tractography in it. Now I'd like to project the 'outer' white matter results (vol) onto the white surface. To test for the acuracy of this approach, I tried to project the white matter mask (binarized, volume) itself on the surface, proper projection should give only 1s (ones) on the surface, which holds true for the majority of vertices. I used the following command: mri_vol2surf --mov whitevolume.mgz --reg struct2freesurfer.dat --projdist-max 0 3 0.1 --interp nearest --hemi rh --out ... Is there a way to optimize this? The binarized white matter data (volume) is just below the white matter surface in most cases, sometimes, voxels are also above the surface. If I try trilinear I get values far away from 1 (hence values which are not included in the primary data (=binarized mask). Looking forward to your suggestions. Thank you very much, Best Robert -- Besuchen Sie uns auf: www.uke.de _ Universitätsklinikum Hamburg-Eppendorf; Körperschaft des öffentlichen Rechts; Gerichtsstand: Hamburg Vorstandsmitglieder: Prof. Dr. Martin Zeitz (Vorsitzender), Prof. Dr. Dr. Uwe Koch-Gromus, Joachim Prölß, Rainer Schoppik _ SAVE PAPER - THINK BEFORE PRINTING ___ 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.
[Freesurfer] help needed with fsgd file and contrast matrix for correlation analysis
Dear all, I have 2 short questions regarding a correlation analysis I'd like to perform with mri_glmfit (surface-based DTI data are correctly preprocessed and concatenated into mgh file, statistics are performed on fsaverage space). 1) I'd like to look for correlation between the surface-data and a behavioural measure BEH in one group, correcting for gender and age. I set up the following fsgd: GroupDescriptorFile 1 Class Male Class Female Variables BEH Age Input subj1 Male 0.76 72 Input subj2 Male 0.44 67 Input subj3 Female 0.33 80 My contrast file is: 0 0 0.5 0.5 0 0 (referring to the FSGD examples Fsgd2G2V, GLM with 2*(2+1) columns). Is this setup correct? If not, how would I have to set up fsgd and C? 2) In a next step, I'd like to answer the question (now including another group2): Where does group1 (OLD) show increased correlation with BEH compared to group2 (YOUNG), correcting for GENDER and AGE within each group. I wonder because AGE is the relevant difference between both groups. GroupDescriptorFile 1 Class OldMale Class OldFemale Class YoungMale Class YoungFemale Variables BEH Age Input subj1 OldMale 0.76 72 Input subj2 OldMale 0.44 67 Input subj3 OldFemale 0.33 80 Input subj11 YoungMale 0.80 20 Input subj12 YoungMale 0.55 21 Input subj13 YoungFemale 0.44 31 My contrast file would look like: 0 0 0 0 0.5 0.5 -0.5 -0.5 0 0 0 0 (referring to GLM with 4*(2+1) columns). Correct? Thank you very much for your help! I am really not sure whether I did it correctly and whether I can trust my results. Best wishes, Robert -- Besuchen Sie uns auf: www.uke.de _ Universitätsklinikum Hamburg-Eppendorf; Körperschaft des öffentlichen Rechts; Gerichtsstand: Hamburg Vorstandsmitglieder: Prof. Dr. Martin Zeitz (Vorsitzender), Prof. Dr. Dr. Uwe Koch-Gromus, Astrid Lurati (komm.), Joachim Prölß, Matthias Waldmann (komm.) _ SAVE PAPER - THINK BEFORE PRINTING ___ 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.
Re: [Freesurfer] combination of different surfaces across the group for mri_glmfit statistics
Dear Douglas, thank you for your reply. Hi, I have two questions: 1. I have created specific surfaces in fsaverage rh for each of my 20 subjects. (xy.mgh) I'd like to run statistical analysis within the common surface area which is included in all subjects. Therefore I wonder how to combine those surfaces. My idea would be to add all surfaces and threshold the resulting surface by 20 == giving a final surface common to all of them. I don't understand what you mean. What are you thresholding? The thickness? No, to make it clear. Surfaces will include data derived from FSL fiber tracking to investigate the structural connectivity of this specific surface (brain area). So the actual surface is just a binary (1) mask. I'd like to know the common core surface hence the surface common to all 20 subjects. How can I calculate this final surface whose area can be found in all 20 subjects. It's just about the localization of the surface. It's about where this surface is located in all subjects. I just thought to add all surfaces with their 1s in ther vertices and see where I get 20. I hope this makes it clearer. 2. How can I limit mri_glmfit to this final surface to run the statistical analysis only in this area, not in all of the hemisphere? You can run mri_glmfit with a --mask option to select only the voxels you want to use mri_glmfit will be conducted based on the surface. so as --mask, I can use the finalmask.mhg? Thank you! doug Can anybody provide the appropriate commands for both issues? Thank you very much! Best wishes, Robert -- Pflichtangaben gemäß Gesetz über elektronische Handelsregister und Genossenschaftsregister sowie das Unternehmensregister (EHUG): Universitätsklinikum Hamburg-Eppendorf; Körperschaft des öffentlichen Rechts; Gerichtsstand: Hamburg Vorstandsmitglieder: Prof. Dr. Martin Zeitz (Vorsitzender), Prof. Dr. Dr. Uwe Koch-Gromus, Astrid Lurati (Kommissarisch), Joachim Prölß, Matthias Waldmann (Kommissarisch) Bitte erwägen Sie, ob diese Mail ausgedruckt werden muss - der Umwelt zuliebe. Please consider whether this mail must be printed - please think of the environment. ___ 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.
