Unfortunately, there isn't a rule-of-thumb really. You can find this out in the list archives.
It looks like 10-15% of your scans are bad (due to fast motion), so you may want to see what happens when you repair all those slices.
Also note that the 0.5mm / TR isn't a mandatory cutoff -- you can adjust the thresholds a bit if you like.
________________________________________
From: SPM (Statistical Parametric Mapping) [[log in to unmask]] on behalf of Manuela Sibold [[log in to unmask]]
Sent: Monday, January 28, 2013 5:43 AM
To: [log in to unmask]
Subject: [SPM] ArtRepair: bad volumes
Dear all,
I am currently using the ArtRepair toolbox to detect volume artifacts in
fMRI data of healthy subjects. With the default thresholds, there were
bad volumes in several subjects some of which seem clearly associated
with rapid motion (> 0.5 mm/TR; see attached files). According to
recommendations posted previously on the SPM mailing list, I´d prefer to
exclude subjects with too many scans of rapid motion (instead of looking
at maximum motion or repairing single volumes). However, I´m not sure
which criterion to apply and some of the bad volumes also seem pretty
normal in the contrast movie slider. Is there any rule-of-thumb
regarding the proportion of rapid motion scans?
Thanks for your help!
Manuela
|