Dear freesurfer experts,
I started using qdec for comparing cortical measures in adolescent patients with ADHD versus typically developing adolescents and I have a few questions concerning the ROI analysis as well as the Desikan-Killiany versus Destrieux atlas.
English is not my native language so please excuse typing errors or wrong expressions. 1) Is it somehow possible to have the significant clusters mapped onto the Destrieux atlas in qdec instead of the Desikan-Killiany one?
2) I would like to export the mean thickness/surface area from significant clusters to SPSS to correlate them with clinical symptoms. I understand you can either use means from the stats tables (aparc_thikness_lh.txt etc.) or manually draw a ROI and then map the lable back to other subjects. But is there also a way to simply export the measures from the significant clusters that show up without drawing around them or export measures from the peak vertices?
3) If I use methods as described in 2), do I have to be concerned about ‘double dipping’ or ‘circular analysis’?
4) Does the measure ‘area.pial’ you can select during design creation correspond to ‘cortical surface area’ and the measure ‘area’ to ‘WM surface area’? Which measure do the aparc_area_lh.txt files that I get when generating stats data tables correspond to? I do not see aparc_area.pial.txt files, should I be getting them?
5) After reading a lot about correcting for global brain measures I am still confused about the best way to do it. I am quite sure that for the volume-based measures it is best to control for ICV/eTIV but for thickness and surface area I read different approaches from not controling to using total brain volume, global thickness or ICV. Is it safe not to control for anything for thickness measures as I do not expect any kind of atrophy in my sample? What should my decision be based on?
Sorry, these are many questions – I appreciate any thoughts on these problems, thank you so much!
Lea Backhausen, B.Sc.
Research Assistant Department of Child and Adolescent Psychiatry, Faculty of Medicine of the Technische Universität Dresden, Dresden, Germany
On 2/2/18 8:01 AM, Backhausen, Lea wrote:
Dear freesurfer experts, I started using qdec for comparing cortical measures in adolescent patients with ADHD versus typically developing adolescents and I have a few questions concerning the ROI analysis as well as the Desikan-Killiany versus Destrieux atlas. English is not my native language so please excuse typing errors or wrong expressions. 1)Is it somehow possible to have the significant clusters mapped onto the Destrieux atlas in qdec instead of the Desikan-Killiany one?
Sorry, I don't know. No one is currently "in charge" of qdec, so we don't support it all that well. I think you can select the annotation though.
2)I would like to export the mean thickness/surface area from significant clusters to SPSS to correlate them with clinical symptoms. I understand you can either use means from the stats tables (aparc_thikness_lh.txt etc.) or manually draw a ROI and then map the lable back to other subjects. But is there also a way to simply export the measures from the significant clusters that show up without drawing around them or export measures from the peak vertices?
No, I would do your analysis using the "command-line stream".
3)If I use methods as described in 2), do I have to be concerned about ‘double dipping’ or ‘circular analysis’?
Possibly so, unless your post hoc tests are completely independent of the test used to generate the clusters.
4)Does the measure ‘area.pial’ you can select during design creation correspond to ‘cortical surface area’ and the measure ‘area’ to ‘WM surface area’? W
Yes
hich measure do the aparc_area_lh.txt files that I get when generating stats data tables correspond to? I do not see aparc_area.pial.txt files, should I be getting them?
It refers to the white surface. Not sure where you are getting these files from, but you can generate the pial area.
5)After reading a lot about correcting for global brain measures I am still confused about the best way to do it. I am quite sure that for the volume-based measures it is best to control for ICV/eTIV but for thickness and surface area I read different approaches from not controling to using total brain volume, global thickness or ICV. Is it safe not to control for anything for thickness measures as I do not expect any kind of atrophy in my sample? What should my decision be based on?
In general, you do not need to control for any global measure with thickness since thickness does not change with head or brain size.
Sorry, these are many questions – I appreciate any thoughts on these problems, thank you so much!
Lea Backhausen, B.Sc.
Research Assistant Department of Child and Adolescent Psychiatry, Faculty of Medicine of the Technische Universität Dresden, Dresden, Germany
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Backhausen, Lea -
Douglas Greve