How many subjects are in each group, and how large is the "subsample" of patients which you scanned on both scanner on the same day?  The issue at hand is whether that subsample is large enough to get an accurate and meaningful estimate of the "scanner" effect to use as a covariate.  Unless that subsample is reasonably large, you are stuck with a situation where "scanner" is rather hopelessly confounded with "group".  

Stated differently, as you know, you ideally want scanner and group to be orthogonal (i.e., balanced design).  Depending on the N of the subsample, I suspect that they are probably actually highly co-linear in your case, which means that inclusion of a scanner covariate is probably going to account for most of whatever variability you have that may be an actual group effect (if one exists), which means that your odds of seeing any group difference are rather low.

cheers,
-MH

-- 
Michael Harms, Ph.D.
-----------------------------------------------------------
Conte Center for the Neuroscience of Mental Disorders
Washington University School of Medicine
Department of Psychiatry, Box 8134
660 South Euclid Ave. Tel: 314-747-6173
St. Louis, MO  63110 Email: mharms@wustl.edu

From: Chris Watson <Christopher.Watson@childrens.harvard.edu>
Reply-To: Freesurfer support list <freesurfer@nmr.mgh.harvard.edu>
Date: Friday, August 8, 2014 1:21 PM
To: "freesurfer@nmr.mgh.harvard.edu" <freesurfer@nmr.mgh.harvard.edu>
Subject: Re: [Freesurfer] Cortical Thickness Correction Factor

In your FSGD you'll have to add a covariate for "scanner"
On 08/08/2014 12:34 PM, David Tate wrote:
FreeSurfer Gurus,
I would like to pose this question again to the group if possible.  We are working with data that comes from two different scanners.  Our experimental group was collected on one scanner and then control data was acquired on another scanner.  Not the optimal situation I realize but unavoidable.  We were fortunate to be able to acquire imaging for a subsample of patients on both scanners in the same day.  Preliminary results suggest that one site measures cortical thickness and subcortical volumes in a non-significant (statistically) but consistent direction with one scanner producing larger volume and thickness measures.  

My question is whether or not it is possible to mathematically account for these differences in any analysis that we are doing using FreeSurfer and if so where can I find those procedures for practically implementing the correction?

Thanks in advance for any help with this.

David


On Apr 17, 2014, at 11:07 AM, David Tate <dftatephd@mac.com> wrote:

We are working on a study that includes data from multiple scanners and initial indication suggest that there may be small differences in cortical thickness measures that are scanner related (human and other phantom data were acquired at both sites).

Is it possible to include a correction factor adjusting for the scanner differences when using glmfit or qdec?

Thanks,

David

David F. Tate, Ph.D.
Research Neuropsychologist
Defense and Veterans Brain Injury Center
Contractor, General Dynamics Information Technology
Brooke Army Medical Center MCHE MDU (DVBIC)
3551 Roger Brooke Drive
JSBA Ft. Sam Houston, TX 78234-4504

Office (210) 916-8090






_______________________________________________
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 materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.