I'd say it depends on what the study is about, but in principle, I'd
think that the covariates should remove what would have potential to
be confounding more what the main effect would be. In the example of
schizophrenia, I'd choose BrainSegVolNotVent, and run both models,
with and without it, and interpret the results combined.
Also, unless if the samples are too small or unless one wants to
explicitly test the effect of ICV and/or BV, I don't see any problem
in including both ICV and BV in the model, even considering they are
well correlated to some extent. The shared variance between them
interferes on the beta-estimates for them, but not on the betas for
the other regressors that you would be testing, and the part of the
variance that is not shared between them will absorb the effects for
which they were chosen to be covaryied out.
This is my opinion. Others may think differently.
All the best!
Anderson
On 24/08/11 12:51, Alan Francis wrote:
Hi Anderson:
Your explanation is very well put. I have a question. Suppose
one is looking at High Risk datasets (for example Schizophrenia)
where the brain morphological alterations are subtle but spread
across the brain, which covariate would you use?
A side note: If you are
using FS 4.5.0 or earlier, there is also BrainSegVolNotVent,
which discount ventricles. If you want to capture effects of
aging or atrophy, perhaps this could be more sensitive than
BrainSegVol.
On 24/08/11 12:32, Anderson Winkler wrote:
Hi Zuzana,
Yes, you can use BrainSegVol as a measurement of brain
volume (I'd assume you are using FS <=4.5.0). Note
that there is another measurement that you might be
interested in, the IntraCranialVol. These measurements
tell different things and not necessarily correlate
well one with another depending on the sample. BSV
considers the voxels labelled as GM and WM, including
subcortical structures and cerebellum and so, it
correlates better with amount of GM and WM and tends
to be more sensitive to pathology, atrophy and aging.
ICV is more robust to these effects. You may want to
choose the one that is more appropriate to your study.
Also, if you are considering brain volume as a
covariate for cortical thickness, it's not necessary.
Thickness correlates poorly with brain volume, and
these two things are not correlated genetically.
Hope this helps!
All the best,
Anderson
On 24/08/11 05:39, zuzana nedelska wrote:
Hello,
I have a question about calculating brain
volume (in mm3) when performing recon-all.
Do I take brainseg volume in the analysis
stream line as subject's brain volume (in mm3)?
Thank you for reply.
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.eduhttps://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.
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.eduhttps://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.
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.
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.eduhttps://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.