Hi John - You can read more on how to use the motion measures in our paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4029882/
The gist of it is that it's not easy to set a hard threshold on the motion measures so that everything above the threshold is good data and everything below is bad. Ultimately it's always a good idea to look at your images before deciding to toss or keep them.
One thing you can do is test for significant differences in motion between the groups that you plan to compare (or significant correlation of motion with any disease or behavioral variables that you plan to correlate to the diffusion data). You can also add motion as a nuisance regressor to your analyses, which we found to reduce spurious findings (see the paper).
Best, a.y
________________________________ From: freesurfer-bounces@nmr.mgh.harvard.edu [freesurfer-bounces@nmr.mgh.harvard.edu] on behalf of John Anderson [John.anderso@protonmail.com] Sent: Monday, June 05, 2017 7:38 AM To: Freesurfer support list Subject: [Freesurfer] TRACULA (dmri_motion)
Dear TRACULA experts, I plan to use TRACULA in freesurfer 6 to analyze DTI data. I want to kindly to inquire about the command "dmri_motion". This script,nicely, generates: the SNR, drop out score, # drop out slices and motion parameters.
I highly appreciate if you can guide me to a method, so I can benefit from these numbers in my evaluation to motion in the data. I tried to plot SNR and Fa, or SNR and motion parameters and I found outliers consistent with more motion and less SNR. In this case is it healthy procedure just to remove the outliers?! Are there any ranges of safety for SNR and motion parameters that can be used to include/exclude the images.
Thanks Jon