Hi Haewon
The problem is that you will never be able to distinguish a scanner effect from a real biological effect since they are in the same subspace. At the very least you should find some control data on both scanners
to show that there is no bias in the direction of the effect you are finding.
Cheers
Bruce
From: freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu>
On Behalf Of Haewon Roh
Sent: Thursday, August 6, 2020 12:16 AM
To: freesurfer@nmr.mgh.harvard.edu
Subject: [Freesurfer] Needing advises for results in my study.
External Email - Use Caution
Hi.
Recently, I analyzed patients who have taken 1.5T MRI ; (Because they had to go in MRI room with some metal materials) and normal controls who took 3.0T MRI. And, after recon and preprocessing of each group of patients and controls in freesurfer
v6, I've found that many cortices of patients (1.5T MRI) were significantly reduced, compared with normal age-sex matched controls (3.0T MRI). When I searched in order to
know if there is difference between 1.5T vs 3.0T MRI and magnetic field can make a huge difference in calculating volumes or thickness of cortices in Freesurfer, I can find only one article by Heinen and his colleague, reporting there is no big difference
and different(1.5 vs 3.0T) magnetic field strength cannot make a significant difference. (Plos ONE, 2016)
Furthermore, given that data which is taken in 1.5T MRI is more likely to have more increased volumes or thickness of cerebral cortices than those of 3.0T MRI, I think that reduced cortical volumes which are calculated using 1.5T MRI have
more significance than controls of 3.0T MRI...
Is it true that different magnetic strengths of MRI cannot change the result significantly in Freesurfer?
Thank you always!