Dear all,
I have got several questions about the analysis of my data. Maybe you can help me out:
1) Regarding the different levels of smoothing (FWHM) in Qdec are there already some 'standards' that one should usually stick to? My results look better when I set the smoothing level to 15 or even 20, but is it really appropriate to use such high levels of smoothing?
2) In Qdec is there a possibilty to set something like an 'Extent threshold' in order to display only clusters of a certain size and hide all clusters that are smaller?
3) Am I correct that the best way to get the total gray and white matter volume is a) for the gray matter volume: mris_volume ?h.pial minus mris_volume ?h.white b) for the white matter volume: mris_wm_volume subject ?h
I read that the values for white and gray matter in the aseg.stats are not that accurate, but what about the GrayVol value in the ?h.aparc.stats files? I was a bit surprised that the sum of these GrayVol values is not the same as the total gray matter volume calculated with mris_volume (although it comes quite close)
4) All of recon-all has been run with Freesurfer 4.02 for all subjects - is it OK to use Freesurfer 4.05 or 4.10 for the analysis?
5) I have tried loading my data into the qdec version shipped with Freesurfer 4.05 and noticed that now there is an option to chose between DODS and DOSS. In my study I wish to compare the thickness of a group of patients and a group of controls accounting for gender and age. Did I understand Dougs freshly updated wiki page correctly and can I use DOSS in case I assume that age has the same effect on patients and controls (so same slope) and they are only starting at different thickness levels? And when chosing DOSS is age really still in the calculation as a covariate? I was surprised to see that I can only chose the question 'Does the average thickness, accounting for gender, differ between patient and control' but not 'accounting for gender and age'.
6) I would like to have a look at the local gyrification index aswell but still need to buy Matlab for it. Do I also need some extra toolboxes to run the lgi calculation - or do I just need the core programm?
Any help is greatly appreciated.
Best wishes, Christian Scheel
--- University of Cologne Department of Psychiatry