Dear FreeSurfers,
I am using a Linux computer with FS 5.1 Due to the known "feauture" in version 5.1 where "Control points may adversely affect aseg" but also may alter the " subcortical segmentation", is it acceptable to use the initial subcortical output (before adding control points) for analysis, and then use the cortical volumetric output (after control points have been added) from corrected brains? If this is acceptable, any suggestions on how to phrase this for a paper without it seeming like FreeSurfer has a bug?
Sincerely,
Mark
Hi Mark
that's the way we always processed things before 5.1 and will do after 5.1, so just state what you did Bruce On Thu, 15 Nov 2012, Mark Fletcher wrote:
Dear FreeSurfers,
I am using a Linux computer with FS 5.1 Due to the known "feauture" in version 5.1 where "Control points may adversely affect aseg" but also may alter the " subcortical segmentation", is it acceptable to use the initial subcortical output (before adding control points) for analysis, and then use the cortical volumetric output (after control points have been added) from corrected brains? If this is acceptable, any suggestions on how to phrase this for a paper without it seeming like FreeSurfer has a bug?
Sincerely,
Mark
Just wanted to mention that, in my experience at least, it isn't just that CPs in 5.1 may adversely affect the aseg, but the inclusion of CPs can actually impact the surfaces in 5.1 in places distant from the CPs themselves. So, if you are using surfaces derived from the default 5.1 recon-all, you may potentially have issues in which the CPs are altering the surfaces in a manner that wasn't present prior to 5.1 as well. (Bruce: please correct if I'm wrong about that). Rather, my understanding is that if one wanted to emulate the approach prior to 5.1 for the surfaces as well, one would need to use the backwards compatability flag that Nick introduced in relation to the CP issue.
cheers, -MH
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Mark Fletcher -
Michael Harms