|
Hi,
Couple things:
1) Values for each structure are NOT already adjusted for total brain size.
2) I would use a regression (covariate) approach for controlling for brain size (not taking the ratio for each individual structure).
3) I would NOT use FS_InterCranial_Vol, as it is an estimate that is based entirely on the determinant of the FS talairach transform, and that transformation was not checked for accuracy.
As to the rest of your email, I couldn't quite follow:
Based on your numbers, FS_TotCort_GM_Vol and its proportionate value (i.e., FS_TotCort_GM_Vol / ICV) are NOT correlated (r=0.03, p=.490). Why does that seem odd?
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: mha...@wustl.edu
From: <Amlung>, Michael <amlu...@mcmaster.ca>
Date: Tuesday, July 14, 2015 11:19 AM To: "hcp-users@humanconnectome.org" <hcp-users@humanconnectome.org> Subject: [HCP-Users] Total brain size correction Hello All, We are analyzing the FreeSurfer volume variables in relation to the delay discounting tasks. We have found statistically significant associations in the expected directions
when we use the raw values for the individual regions and the overall cortical and subcortical grey/white matter values. However, we should ideally be using variables that have corrected for total brain size in some way. We have tried this in two ways: calculating
a proportionate volume using total gray matter (FS_Total_GM_Vol); and calculating a proportionate volume using intercranial volume (FS_InterCranial_Vol). The issue is that our associations completely disappear, or in some cases, even reverse direction. It
seems odd to us that r values would essentially drop to zero when a proportionate volume is used.
We probed this further to see if proportionate volumes were correlated with their corresponding raw value. FS_TotCort_GM_Vol and its proportionate value (i.e., FS_TotCort_GM_Vol
/ ICV) are correlated r = .030, p = .490. This seems odd to us. For subcortical GM, the correlation is even more confusing, r = -.105, p = .02.
When we use total grey matter, the correlations are a bit better, but still not as high as we would expect (.48 for cortical GM and .12 for subcortical GM).
Are we missing something here? Are the raw values already adjusted for total brain size in some way and we’re essentially doing something redundantly? Any thoughts would be helpful. Michael Amlung ******************************************** Michael Amlung, Ph.D. Assistant Professor Department of Psychiatry and Behavioural Neurosciences DeGroote School of Medicine McMaster University Peter Boris Centre for Addictions Research 100 West 5th Street Hamilton, Ontario L8N 3K7 Canada Telephone: (+1) 905-522-1155, ext. 39014 Email:
amlu...@mcmaster.ca _______________________________________________ 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. _______________________________________________ |
- [HCP-Users] Total brain size correction Amlung, Michael
- Re: [HCP-Users] Total brain size correction Harms, Michael
