dear fsl experts.
we would like to perform T-ICA on n number of subjects with 2 separate
sessions. as long as i understand former mails about this topic the
concatenated approach would be the one to follow. does this perform
something like a session specific scaling to control for session (or
timeseries interruption) specific artefacts? how could i prevent components
which definetely show artefacts due to interruption of the timeseries?
any advice appreciated
thanks markus
|