On 04/03/2017 01:45 PM, Veggeberg, Rosanna wrote:
Hello, I am using QDEC to analyze cortical structural data and am looking for some clarification on some questions that have come up.
- _Design tab_: In the "Measure" drop-down selection, I understand
what we are getting when looking at thickness but the other measurements I am unclear on. Is there information or publications on the other 12 possible measures (how they are measured, what they mean, applications for using them). In particular, I am interested in learning more about area, area.pial, volume, sulc, and curv but would like to understand what the others mean as well to determine if they would be applicable to our study questions.
There is not a specific pub on this. The area is the area of a vertex (the average of the triangles around it), same interpreation for the volume. These are kind of like VBM analysis. area.pial is the area of the pial surface whereas simple area is the area of the white. curv is the curvature; sulc is the sulcal depth. These are used for registration and generally not that interesting.
- _Design tab_: I am a little unclear on why the model factors are
labeled the way they are and wanted to see if I am using them correctly. -Discrete (Fixed Factors): this is pretty straight forward to be used as independent categorical variables, correct?
yes
-Continuous (Covariate): even though it says covariate, should these variables just be treated as independent continuous varaibles?
correct
-Nuisance Factors: should these just be treated as the covariates we wish to control for?
yes. it allows you to have more continuous variable, but you won't get tests that include them
- _Results_: When I am doing a group comparison (i.e. disease vs.
control), does QDEC run the comparison based off of which order the labels appear in the table: disease vs. control, or does it run the comparison based off of which label comes first alphabetically: control vs. disease?
The order in your levels file
- _Results_: When viewing the results in freeview, if I click on a
significant cluster, in the bottom panel, does a negative value next to the mc-z.abs.th13.sig.cluster.mgh label indicate a decrease in cortical thickness (i.e. disease vs. control: negative value indicates control is greater than disease)?
It depends on the order above. It will be contrast = first - second.
- Finally, once you have closed QDEC, is it possible to reopen the
files again in the Gui? Or would you have to rerun the analysis?
sorry, not sure
Thank you for help and please let me know if you need any further clarification.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer