Hi Freesurfer Experts,
Just to make it sure i have a question about correcting for cranial size.
I compared two groups, 100 patients versus 50 Controls. I entered a discrete variable 'groups' (Patients: 1 and controls:2) in qdec. For cortical thickness difference between these groups, do i have to control for total intracranial volume? ICV (intra cranial volume is output from aseg2tables) or controlling for another variable which estimate total brain volume, like total gray matter or supra tentorial volume?
In my correlation analyses, relating cognitive peformance and cortical thickness within the patient group. Do i have to control for ICV or Total gray matter? besides the variable age, gender and education.
I think only volume differences between groups need controlling for ICV, is that correct?
And in my correlation analyses, relating cognitive performance ato subcortical volume within the patient group. Do i have to control for ICV or Total gray matter? besides the variables age, gender and education.
Do you have a reference to a methodological paper, which descibes different (anatomical) nuisance factors and their influence on the data? Is there a automated correction for brain size in the cortical and subcortical freesurfer stream (step of recon -all command)?
Many Thanks,
Stan
Hi Stan, I don't think you have to correct thickness for head size or grey matter volume, but you should do so for subcortical volumes. Whether you use ICV or GM volume or brain volume depends on what you believe is happening and what hypotheses you want to test. Check out the paper by Randy Buckner on estimating ICV from the talairach transform:
http://surfer.nmr.mgh.harvard.edu/fswiki/eTIV?action=AttachFile&do=view&...
doug
On 12/10/2012 11:34 AM, Berg, S.F. van den wrote:
Hi Freesurfer Experts,
Just to make it sure i have a question about correcting for cranial size.
I compared two groups, 100 patients versus 50 Controls. I entered a discrete variable 'groups' (Patients: 1 and controls:2) in qdec. For cortical thickness difference between these groups, do i have to control for total intracranial volume? ICV (intra cranial volume is output from aseg2tables) or controlling for another variable which estimate total brain volume, like total gray matter or supra tentorial volume?
In my correlation analyses, relating cognitive peformance and cortical thickness within the patient group. Do i have to control for ICV or Total gray matter? besides the variable age, gender and education.
I think only volume differences between groups need controlling for ICV, is that correct?
And in my correlation analyses, relating cognitive performance ato subcortical volume within the patient group. Do i have to control for ICV or Total gray matter? besides the variables age, gender and education.
Do you have a reference to a methodological paper, which descibes different (anatomical) nuisance factors and their influence on the data? Is there a automated correction for brain size in the cortical and subcortical freesurfer stream (step of recon -all command)?
Many Thanks,
Stan
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
In addition to the Buckner article, for other literature on the topic of adjusting for volumes you can see: O'Brien et al., Psychiatry Research: Neuroimaging 193:113-122 (2011) Sanfilipo et al., Neuroimage 22:1732-1743 (2004)
I'll also mention that people sometimes use global global thickness as a covariate for studies involving thickness as the dependent variable.
cheers, -MH
freesurfer@nmr.mgh.harvard.edu