Hi Brianna,

Yes, that is the best option in my opinion. You should include field strength as a covariate, and maybe other acquisition parameters.

Best Martin


Sent via my smartphone, please excuse brevity.

-------- Original message --------
From: Brianna Damadian
Date:07/02/2014 10:04 PM (GMT+01:00)
To: Freesurfer@nmr.mgh.harvard.edu
Subject: [Freesurfer] Longitudinal pipeline + LME question

Hello experts,

We are setting out to look at a sample of 10 (subset of eventual larger sample) who have at least 2 sets of MRI. Some have many more than 2. However, some of the scans were done on different scanners (a few on 1.5T versus most on 3T) and with slightly different parameters (namely, different TR) - not exactly ideal.

We are looking for changes in total ventricular volume as well as total gray matter and white matter (among others). I am new to this software, but from what I have read in the literature the best approach would be to use the longitudinal pipeline in FreeSurfer and then use the Linear Mixed Effect model (LME) for statistical analysis. In your opinion, is this the correct approach based on our sample/goals?

Thank you.