That depends on what you expect from your data. If you do not have an apriori hypothesis about the sign of the contrast you are testing, then use abs (unsigned). Eg, if your hypothesis is Controls have a thicker cortex than ADs and you set up your contrast to be gamma = ThicknessControls - ThicknessADs, then you might expect gamma > 0. By selecting a sign (pos or neg), you get an advantage in the significance. But the hypothesis MUST be a priori. You should not play with your data to see what gives you the best result.
On 05/26/2016 03:07 PM, Hao wen wrote:
Hi, Doug: I check out the documentation for this command: 'mri_surfcluster', and I also played with another comparision, with the --cache value with 1.3 and 4, and I get fewer cluster with 1.3, so right now I got the idea for the --cache value, but Im still confused about how do you decide which sign to use? I know the what the values do for the cluster, but what about the sign? Merci Hao ________________________________________ 发件人: freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu 代表 Hao wen hao.freesurfer@hotmail.com 发送时间: 2016年5月26日 20:16:18 收件人: freesurfer@nmr.mgh.harvard.edu 主题: [Freesurfer] 答复: 答复: 答复: 答复: Cluster-wise correction error with mri_glmfit-sim
Hi, Doug: I find something, here r my command:
mri_glmfit-sim \ --glmdir lh.AD_CN_NO_COVARIATE.glmdir \ --cache 4 pos \ --cwp 0.05
so I changed the sign, at first, it was 'neg', after correction, i got no cluster, but with 'pos', I got Ncluster = 9, and with 'abs', I got Ncluster = 8, So i am a little lost, I checked out the explaination for the 'mri_glmfit-sim ', so my question is that what is the differences for the sign, and what are the thresholds for --cache and --cwp seperately? Hope to hear from you soon! Merci Best Hao ________________________________________ 发件人: freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu 代表 Douglas N Greve greve@nmr.mgh.harvard.edu 发送时间: 2016年5月26日 19:27:53 收件人: freesurfer@nmr.mgh.harvard.edu 主题: Re: [Freesurfer] 答复: 答复: 答复: Cluster-wise correction error with mri_glmfit-sim
They use two different methods. SurfStat uses gaussian random fields and FS uses monte carlo simulation. How significant is your cluster in FS?
On 05/26/2016 01:17 PM, Hao wen wrote:
Hi: Because I make the group analysis with SurfStat, and I get the cluster after the clusterwise correction, so I think in Freesurfer, we should have the similiar result:)
发件人: freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu 代表 Douglas N Greve greve@nmr.mgh.harvard.edu 发送时间: 2016年5月26日 18:22:58 收件人: freesurfer@nmr.mgh.harvard.edu 主题: Re: [Freesurfer] 答复: 答复: Cluster-wise correction error with mri_glmfit-sim
why do you think there shold be a cluster?
On 05/26/2016 12:21 PM, Hao wen wrote:
Hi, I tried every value for the --cache and --cwp, as what i understood, the --cache is the threshold for the vertex-wise correction and the --cwp is the clusterwise correction, and i even tried the 0.99 with the --cwp, which should give me all my cluster, but still, I got nothing, I checked the 'mri_glmfit-sim', and it seems not so clear for me, do you there where is the problem?
Merci Hao ________________________________________ 发件人: freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu 代表 Douglas N Greve greve@nmr.mgh.harvard.edu 发送时间: 2016年5月26日 17:54:58 收件人: freesurfer@nmr.mgh.harvard.edu 主题: Re: [Freesurfer] 答复: Cluster-wise correction error with mri_glmfit-sim
It looks like it worked. there is not an error there. It did not find any voxels above your chosen level of signficance, but that is not an error per se
On 05/26/2016 11:19 AM, Hao wen wrote:
Doug:
Thank you for your reply, it does not work, also, I think the problem is that for me after the clusterwise correction, I got no corrected cluster, I am not sure if it the problem that I set the --cache value and the --cwp value, but I changed the value, it always return me nothing!
This is the msg, it said that found 0 value :
Computing voxel-wise significance CSDpvalMaxSigMap(): found 0/163842 above 0 Adjusting threshold for 1-tailed test. If the input is not a -log10(p) volume, re-run with --no-adjust. Searching for Clusters ... thmin=1.300000 (0.998970), thmax=-1.000000 (-1), thsignid=-1, minarea=0.000000 Found 2 clusters Max cluster size 132.153870 Pruning by CW P-Value 0.001 Saving thresholded output to lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.masked.mgh Saving cluster numbers to lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.ocn.mgh Saving cluster pval lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.mgh
Do you have any idea?
Best
Hao
*发件人:* freesurfer-bounces@nmr.mgh.harvard.edu freesurfer-bounces@nmr.mgh.harvard.edu 代表 Douglas Greve greve@nmr.mgh.harvard.edu *发送时间:* 2016年5月26日 17:06:48 *收件人:* freesurfer@nmr.mgh.harvard.edu *主题:* Re: [Freesurfer] Cluster-wise correction error with mri_glmfit-sim Instead of --2spaces, use --bonferroni 2
On 5/26/16 10:30 AM, Hao wen wrote:
Hello:
Freesurfers, follow the tutorial '
# FsTutorial https://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial # GroupAnalysis https://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/GroupAnalysis?action=fullsearch&context=180&value=linkto%3A%22FsTutorial/GroupAnalysis%22', my task is to find the thickness differences between groups(Controls(30) & AD(30)), and it works well for the first part which is to fit the linear model and get the uncorrected P_values, and I got the expected P-value map, but when I followed the clusterwise correction for multiple comparision, with this command below:
*mri_glmfit-sim *
- --glmdir lh.gender_age.glmdir *
- --cache 4 neg *
- --cwp 0.05*
- --2spaces*
I got two questions:
- with the flag --2spaces, i got the error that this command not found;
- also, with or without the --2spaces, after the freeview open, i
got another error, which said did not find any volume geometry information in the surface, I googled, and I know Doug has answered in this post: http://www.mail-archive.com/freesurfer%40nmr.mgh.harvard.edu/msg42911.html, but I think my freesurfer is the newest version:freesurfer-5.3.0-15052013-x86_64-centos!
As below, these are my command and the output:
*§§Command:* export SUBJECTS_DIR=/aramis/dataARAMIS/users/junhao.wen/Back_up/Data_AD_CN/ADNI_60_AD_CN_object_recon_all_output/ cd $SUBJECTS_DIR/glm/AD_CN_NO_COVARIATE mri_glmfit-sim \ --glmdir lh.AD_CN_NO_COVARIATE.glmdir \ --cache 1.3 neg \ --cwp 0.05\ --2spaces
# load the cluster annotation in freeview freeview -f $SUBJECTS_DIR/fsaverage/surf/lh.inflated:overlay=lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.mgh:overlay_threshold=2,5:annot=lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.ocn.annot -viewport 3d
*§§Output:*
cmdline mri_glmfit --y lh.AD_CN_NO_COVARIATE.thickness.15.mgh --fsgd AD_CN_NO_COVARIATE.fsgd dods --C group.diff.mtx --surf fsaverage lh --cortex --glmdir lh.AD_CN_NO_COVARIATE.glmdir SURFACE: fsaverage lh log file is lh.AD_CN_NO_COVARIATE.glmdir/cache.mri_glmfit-sim.log
cd /aramis/dataARAMIS/users/junhao.wen/Back_up/Data_AD_CN/ADNI_60_AD_CN_object_recon_all_output/glm/AD_CN_NO_COVARIATE /aramis/dartagnan2/Software/FreeSurfer/freesurfer-5.3.0-15052013-x86_64-centos/bin/mri_glmfit-sim --glmdir lh.AD_CN_NO_COVARIATE.glmdir --cache 1.3 neg --cwp 0.05
$Id: mri_glmfit-sim,v 1.36.2.5 2012/10/01 22:31:37 greve Exp $ Thu May 26 16:20:39 CEST 2016 Linux HP1973 3.11.0-26-generic #45~precise1-Ubuntu SMP Tue Jul 15 04:02:35 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux wen setenv SUBJECTS_DIR /aramis/dataARAMIS/users/junhao.wen/Back_up/Data_AD_CN/ADNI_60_AD_CN_object_recon_all_output/ FREESURFER_HOME /aramis/dartagnan2/Software/FreeSurfer/freesurfer-5.3.0-15052013-x86_64-centos
Original mri_glmfit command line: cmdline mri_glmfit --y lh.AD_CN_NO_COVARIATE.thickness.15.mgh --fsgd AD_CN_NO_COVARIATE.fsgd dods --C group.diff.mtx --surf fsaverage lh --cortex --glmdir lh.AD_CN_NO_COVARIATE.glmdir
DoSim = 0 UseCache = 1 DoPoll = 0 DoPBSubmit = 0 DoBackground = 0 DiagCluster = 0 gd2mtx = dods fwhm = 22.776285 CSD /aramis/dartagnan2/Software/FreeSurfer/freesurfer-5.3.0-15052013-x86_64-centos/average/mult-comp-cor/fsaverage/lh/cortex/fwhm23/neg/th13/mc-z.csd mri_surfcluster --in lh.AD_CN_NO_COVARIATE.glmdir/group.diff/sig.mgh --csd /aramis/dartagnan2/Software/FreeSurfer/freesurfer-5.3.0-15052013-x86_64-centos/average/mult-comp-cor/fsaverage/lh/cortex/fwhm23/neg/th13/mc-z.csd --mask lh.AD_CN_NO_COVARIATE.glmdir/mask.mgh --cwsig lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.mgh --vwsig lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.voxel.mgh --sum lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.summary --ocn lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.ocn.mgh --oannot lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.ocn.annot --annot aparc --csdpdf lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.pdf.dat --cwpvalthresh 0.05 --o lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.masked.mgh --no-fixmni --surf white Creating CDFs from CSD files csd->threshsign = -1 thsign = neg, id = -1 version $Id: mri_surfcluster.c,v 1.51.2.3 2012/05/31 22:10:05 greve Exp $ hemi = lh srcid = lh.AD_CN_NO_COVARIATE.glmdir/group.diff/sig.mgh srcsubjid = fsaverage srcsurf = white srcframe = 0 thsign = neg thmin = 1.3 thmax = -1 fdr = -1 minarea = 0 xfmfile = talairach.xfm nth = -1 outid = lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.masked.mgh MGH ocnid = lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.ocn.mgh MGH sumfile = lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.summary subjectsdir = /aramis/dataARAMIS/users/junhao.wen/Back_up/Data_AD_CN/ADNI_60_AD_CN_object_recon_all_output/ FixMNI = 0 Found 149955 vertices in mask Found 149955 vertices in mask Found 149955 points in clabel. ------------- XFM matrix (RAS2RAS) --------------- /aramis/dataARAMIS/users/junhao.wen/Back_up/Data_AD_CN/ADNI_60_AD_CN_object_recon_all_output//fsaverage/mri/transforms/talairach.xfm 1.000 0.000 0.000 0.000; 0.000 1.000 0.000 0.000; 0.000 0.000 1.000 0.000; 0.000 0.000 0.000 1.000;
Reading source surface /aramis/dataARAMIS/users/junhao.wen/Back_up/Data_AD_CN/ADNI_60_AD_CN_object_recon_all_output//fsaverage/surf/lh.white Done reading source surface Reading annotation /aramis/dataARAMIS/users/junhao.wen/Back_up/Data_AD_CN/ADNI_60_AD_CN_object_recon_all_output//fsaverage/label/lh.aparc.annot reading colortable from annotation file... colortable with 36 entries read (originally /autofs/space/terrier_001/users/nicks/freesurfer/average/colortable_desikan_killiany.txt) Computing metric properties Loading source values number of voxels in search space = 149955 Done loading source values (nvtxs = 163842) overall max = 8.58718 at vertex 22205 overall min = -1.57802 at vertex 24019 surface nvertices 163842 surface area 65417.097656 surface area 65416.648438 Computing voxel-wise significance CSDpvalMaxSigMap(): found 0/163842 above 0 Adjusting threshold for 1-tailed test. If the input is not a -log10(p) volume, re-run with --no-adjust. Searching for Clusters ... thmin=1.300000 (0.998970), thmax=-1.000000 (-1), thsignid=-1, minarea=0.000000 Found 2 clusters Max cluster size 132.153870 Pruning by CW P-Value 0.05 Saving thresholded output to lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.masked.mgh Saving cluster numbers to lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.ocn.mgh Saving cluster pval lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.mgh Constructing output annotation Writing annotation lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.ocn.annot mri_segstats --seg lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.ocn.mgh --exclude 0 --i /aramis/dataARAMIS/users/junhao.wen/Back_up/Data_AD_CN/ADNI_60_AD_CN_object_recon_all_output/glm/AD_CN_NO_COVARIATE/lh.AD_CN_NO_COVARIATE.thickness.15.mgh --avgwf lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.y.ocn.dat --sum /tmp/mri_glmfit-sim.junk.2688
$Id: mri_segstats.c,v 1.75.2.9 2013/02/16 00:09:33 greve Exp $ cwd cmdline mri_segstats --seg lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.ocn.mgh --exclude 0 --i /aramis/dataARAMIS/users/junhao.wen/Back_up/Data_AD_CN/ADNI_60_AD_CN_object_recon_all_output/glm/AD_CN_NO_COVARIATE/lh.AD_CN_NO_COVARIATE.thickness.15.mgh --avgwf lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.y.ocn.dat --sum /tmp/mri_glmfit-sim.junk.2688 sysname Linux hostname HP1973 machine x86_64 user wen UseRobust 0 Loading lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.ocn.mgh Loading /aramis/dataARAMIS/users/junhao.wen/Back_up/Data_AD_CN/ADNI_60_AD_CN_object_recon_all_output/glm/AD_CN_NO_COVARIATE/lh.AD_CN_NO_COVARIATE.thickness.15.mgh Voxel Volume is 1 mm^3 Generating list of segmentation ids Found 1 segmentations Computing statistics for each segmentation
Reporting on 0 segmentations Computing spatial average of each frame
Writing to lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.y.ocn.dat mri_segstats done mri_convert lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.mgh lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.mgh --frame 0 mri_convert lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.mgh lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.mgh --frame 0 $Id: mri_convert.c,v 1.179.2.7 2012/09/05 21:55:16 mreuter Exp $ reading from lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.mgh... TR=0.00, TE=0.00, TI=0.00, flip angle=0.00 i_ras = (-1, 0, 0) j_ras = (0, 0, -1) k_ras = (0, 1, 0) keeping frame 0 writing to lh.AD_CN_NO_COVARIATE.glmdir/group.diff/cache.th13.neg.sig.cluster.mgh... Thu May 26 16:20:39 CEST 2016 Thu May 26 16:20:45 CEST 2016 mri_glmfit-sim done ./cluster_wise_correction_for_multiple_comparisons.sh: line 23: --2spaces: command not found Did not find any volume geometry information in the surface p–Breading colortable from annotation file... colortable with 1 entries read (originally none) colortable with 1 entries read (originally none) CTABisEntryValid: index -1 was OOB Resource temporarily unavailable Resource temporarily unavailable B ginally none) CTABisEntryValid: index -1 was OOB Resource temporarily unavailable Resource temporarily unavailable B ginally none) CTABisEntryValid: index -1 was OOB Resource temporarily unavailable Resource temporarily unavailable B ginally none) CTABisEntryValid: index -1 was OOB Resource temporarily unavailable Resource temporarily unavailable B ginally none)
Can you give me some answer? Thank you very much!
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
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/
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 mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/
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
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/
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