Thanks Doug! If I may follow up:
- (I can't see whether there's a tutorial for this, as the main FS
site seems to be down at the moment) How can a group comparison of CT be done inside a given ROI, and in what ways can the ROI be specified? I guess the command line version of QDEC would produce, in this case, the same contrasts as QDEC, but within the ROI as opposed to the whole brain
You mean you want to do an exploratory analysis within a mask of an ROI or that you want to average the CT scan within the ROI and do a group analysis of the ROI values? If the former, you can use mri_glmfit with the -label or -mask option. If the latter, you can create a table with asegstats2table then run mri_glmfit with --table
3') Is it the case that mri_glmfit with mask=brainmask.mgz is equivalent to doing a whole-brain analysis in QDEC?
3'') I didn't find, on the mri_glmfit help page, how masks can be formed out of one of the atlas structures, e.g. "lh_S_intrapariet_and_P_trans "... The --table argument is also not present anywhere on that page..
3''') Would it not be a*parc*stats2table rather than a*seg*stats2table, if the ROI in question was cortical and not subcortical? Is the difference between what FS calls "segmentation" and "parcellation" just terminology, i.e. the former delineates subcortical structures and the latter delineates cortical structures? Or does it imply anything else?
- How do the results of these two different types of ROI group
analyses differ, and is one of them more "correct" than the other: A) running the command prompt version of QDEC within the confines of a certain atlas-defined ROI, and looking at the resulting statistical map (clusters), as per question 3; B) extracting the CT values for that ROI for all subjects using aparcstats2table, and doing t-tests to look for a group difference.
Oops, looks like these are the two I mentioned from #3 above. The first is an exploratory analysis in which the groups are compared on a vertex-by-vertex basis. If there is a subset of vertices that are different between the groups, it may show up in the exploratory analysis. However, the effect may be small at each vertex and averaging over the vertices may improve your power (unless the effect is only at a few vertices). One is not more correct than the other, just testing different hypotheses.
4') Can the second option (averaging the CT values within the ROI) not also be done by taking the values from the table into SPSS and simply doing an independent samples t-test? Is that not equivalent to what mri_glmfit would do in this case?
And one more question: 5) Can the statistical map in QDEC be built using a lower, uncorrected p value (e.g. 0.0001) instead of an FDR-corrected test value?
Thanks!! Tudor
doug
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.