PALM isn’t intended to replace the level 1 (timeseries) analysis. Permutation testing doesn’t handle the autocorrelated timeseries appropriately, because time points are not truly exchangeable.
It should be possible to do a repeated measures analysis in PALM (treating acitvation estimates from each run as repeated measures). However, the easiest use would be to use the output of the level 2 analyses as inputs to a simple group-level random effects analysis. As an aside, if you’re attending the 2016 HCP Course in Boston (https://www.humanconnectome.org/courses/2016/exploring-the-human-connectome.php), you will be able to go through a practical session on this! (We hope to release the practicals to the public after the course, but that will likely take some time to implement.) --Greg ____________________________________________________________________ Greg Burgess, Ph.D. Staff Scientist, Human Connectome Project Washington University School of Medicine Department of Psychiatry Phone: 314-362-7864 Email: [email protected] > On Aug 4, 2016, at 8:47 AM, Michael F.W. Dreyfuss <[email protected]> > wrote: > > Hi, I am trying to adapt palm into my analysis scripts upon your > recommendation and I have a few questions: > > 1) Does this function essentially like FEAT/film_gls? That is, should I be > running PALM for level 1 (run), level 2 (subject) and level 3 (group) > analyses, or just at one of these levels? > > 2) Are there any examples available of how I can work palm in place of > film_gls in existing HCP processing scripts? I'm thinking of lines like the > following: > > #Run film_gls on subcortical volume data > film_gls --rn=${FEATDir}/SubcorticalVolumeStats --sa --ms=5 > -in=${FEATDir}/${LevelOnefMRIName}_AtlasSubcortical"$TemporalFilterString""$SmoothingString".nii.gz > --pd="$DesignMatrix" --con=${DesignContrasts} --fcon=${DesignfContrasts} > --thr=1 --mode=volumetric > rm > ${FEATDir}/${LevelOnefMRIName}_AtlasSubcortical"$TemporalFilterString""$SmoothingString".nii.gz > > 3) Can I simply use the same design files produces by feat_model from my fsf > files? Such as done at the line: > > feat_model ${FEATDir}/design ${ResultsFolder}/${LevelOnefMRIName}/${Confound} > > 4) For motion, is it recommended to you detrended regressors on non-detrended > regressors? > > Sorry for all the questions but I am finding few examples available of how > this is implemented practically. > > Thank you, > Michael Dreyfuss > MD-PhD Student > Weill Cornell Medical College > _______________________________________________ > HCP-Users mailing list > [email protected] > http://lists.humanconnectome.org/mailman/listinfo/hcp-users ________________________________ The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail. _______________________________________________ HCP-Users mailing list [email protected] http://lists.humanconnectome.org/mailman/listinfo/hcp-users
