I think your approach of using a covariate will work well enough. The one thing you'll want to check is whether your statistical test will be confounded by headcoil/FSversion. Eg, if all your patients came from 8 channel and all your controls came from 32, then you would not know whether your differences are caused by diagnosis or by head coil.
On 9/10/2025 11:21 AM, Stacey Schaefer wrote:
External Email - Use Caution
Hello Freesurfer gurus, We have two samples of approximately n~125-150 whose T1s were collected on the same scanner with the same acquisiton, /except/ for a difference in head coil (8 vs 32 channel) and processed with different Freesurfer versions (v5.3.0 and v. 6.0.0). Both also had the FreeSurfer segmentation of hippocampal subfields v 6.0.0 applied. We would like your guidance about combining their data if possible to maximize sample size.
What would your suggestion be for hypothesis testing? Can we combine the samples - for example to test moderators of hippocampal x age relationships, using simply a covariate to account for the differences in acquisition/Freesurfer processing (in addition to the usual sociodemographic and estimated intracranial volume covariates)?
OR is that problematic and would you instead suggest keeping them separate and using one as a replication sample?
Thank you for your guidance! Stacey
*Stacey M. Schaefer, Ph.D.* Leader of the MIDUS Neuroscience Project &
Co-Leader of the MIDUS Admin Core
MIDUS NEURO
Keystone House
901 University Bay Dr
Madison WI 53705
Freesurfer mailing list --freesurfer@nmr.mgh.harvard.edu To unsubscribe send an email tofreesurfer-leave@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman3/lists/freesurfer@nmr.mgh.harvard.e...