Dear Nima,
You can of course get better labels from a 1mm iso T1w scan. If your registration is good, then this is the way to go.
Analyzing the diffusion data will give you less accurate labels, but eliminates registration errors. You can try both and keep the best looking one.
Having said that: I would use SynthSeg rather than SAMSEG, which will give you a crisper 1mm isotropic segmentation.
Cheers,
Eugenio
http://www.jeiglesias.com
From:
freesurfer-bounces@nmr.mgh.harvard.edu <freesurfer-bounces@nmr.mgh.harvard.edu> on behalf of Taghizadeh Mortezaei, Nima <ntaghizadehmortezaei@mgh.harvard.edu>
Date: Friday, April 11, 2025 at 12:58 PM
To: Freesurfer support list <freesurfer@nmr.mgh.harvard.edu>
Subject: [Freesurfer] Feasibility and Accuracy of Using SAMSEG with ADC Maps
Dear FreeSurfer team,
I am currently using FreeSurfer v7.4.1 and I want to ask is it feasible to run SAMSEG directly on ADC maps derived from diffusion MRI for segmenting normal brain anatomy? Alternatively, would it
be more accurate to perform segmentation on a T1-weighted image and then transfer the labels to the co-registered ADC image?
Thank you very much for your time and support.