Hi
concatenating time series is always tricky: are you sure that voxel
locations match ? (between run motion can be significantly bigger than
within-run motion); are you sure that the scanner isn't calibrating at
the start of a new run (potentially giving a completely different mean
intensity for the first volume of the second run compared to the last
volume of the first run); what about the temporal filter? (filtering
typically introduces end effects - how do you sensibly concatenate
after filtering?) ... just a few issues
ta
christian
On Tuesday, Jun 24, 2003, at 11:28 Europe/London, Andy Myers wrote:
> what are the reasons for not combining several runs worth of data fopr
> one
> subject if you have first preprocessed all the data, i know the
> general rule
> is 'don't' but why is this?
>
> andy
|