Some of this Martin will have to comment on, but my comments below On 03/03/2015 12:35 PM, Janosch Linkersdörfer wrote: > Hey Doug, > > thanks again! That means, I would do something like: > > 1.) construct longitudinal qdec table > - only use fsid, fsid-base, and years/age columns > - substract the average age of the one-time point subjects from years/age > value for every subject at every time point > > 2.) run long_mris_slopes > - use the --do-avg flag? Yes, I think so > > 3.) concatenate the output files from 2) and the maps from the one-time point > subjects with mris_preproc > - use --is flag to select maps by specifying full path What command are you talking about to do the concatenation? Otherwise sounds right. Make sure you use the average of the time points from long_mris_slopes and not the slope. > > 4.) run mri_glmfit > - use --osgm flag for one-sample group mean Yes > > 5.) replace sig.mgz file with the one from LME analysis and run mri_glmfit-sim Yes > > Is that correct? (especially, is using --do-avg what you meant by taking the > offset) > > Thanks, > > Janosch > > > Am 03.03.2015 um 08:32 schrieb Douglas N Greve <gr...@nmr.mgh.harvard.edu>: > >> I don't know, esp since it is an approximation to begin with. An alternative >> is to take the offsets from your multi-time-point subjects and the single >> maps from your subjects with one time point and run that through the >> one-sample-group-mean (--osgm in mri_glmfit). If you go this route, then you >> should subtract the mean age of the one time-point subjects from the age of >> the multi-time point subjects before the 1st stage of analysis. >> >> doug >> >> >> On 03/02/2015 07:25 PM, Janosch Linkersdörfer wrote: >>> Hi Doug, >>> >>> thank you very much for your answer! >>> >>> Am 02.03.2015 um 11:30 schrieb Douglas N Greve <gr...@nmr.mgh.harvard.edu>: >>> >>>> You would do the long analysis using a random effects analysis. >>> OK, so basically do 2-stage modeling >>> (https://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalTwoStageModel), >>> right? >>> >>>> For each subject you can get a slope (this won't work if the subject only >>>> has 1 time point), then concatenate the slopes into a file and run >>>> mri_glmfit, then follow the procedures from the archive email you >>>> reference. >>> One quarter of my subjects only has one measurement time point. Can I still >>> use this method (which would only consider the 3/4 of the subjects) to >>> correct clusters found in the whole group? Or would it only be valid for >>> correcting clusters from an LME model for the reduced group? >>> >>> Thanks, >>> >>> Janosch >>> >>>> doug >>>> >>>> On 02/24/2015 08:35 PM, Janosch Linkersdörfer wrote: >>>>> Hi Doug and others, >>>>> >>>>> I would like apply (Monte Carlo simulation) cluster correction (as >>>>> opposed to the implemented vertex-wise FDR correction) on the results >>>>> from a longitudinal study I analyzed using the LME toolbox. The design is >>>>> unbalanced (different number of time points, from 1 to 4, per subject). >>>>> >>>>> In this thread >>>>> (http://www.mail-archive.com/freesurfer%40nmr.mgh.harvard.edu/msg35367.html) >>>>> you, Doug, suggested, if I understand correctly: >>>>> >>>>> - concatenating the images using `mris_preproc --paired-diff` >>>>> - smooth with the same kernel size as used in the lme analysis >>>>> - running `mri_glmfit` on them with an fsgd file that uses the same >>>>> covariates and the same contrast (excluding the interaction term with >>>>> time) as used in the lme analysis >>>>> - overwriting sig.mgh with the one from the lme analysis >>>>> - running `mri_glmfit-sim --cache` >>>>> >>>>> How would I extend this to my case where I don't have pairwise images, >>>>> but 1 image for some participants, for others up to 4? >>>>> >>>>> Thanks a lot, >>>>> >>>>> Janosch >>>>> >>>>> >>>>> >>>>> >>>> -- >>>> 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: 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/ >>>> >>>> >>>> >>>> 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. >>>> >>> >> -- >> 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: 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/ >> > >
-- 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: 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
