One method to check if motion between sessions has caused an artifact is run art_movie on the normalized images across the session boundary, e.g. start near the end of one session and continue through the beginning of the next session. If there is a sudden change in timeseries intensities across the session boundary, then linking timeseries across sessions may have introduced an error.
The SPM option of having a separate constant covariate per session can model this intensity jump (at a tiny cost of one degree of freedom). The art_movie program is part of the ArtRepair toolbox.
-Paul
----- Original Message -----
From: "Jonathan Peelle" <[log in to unmask]>
To: [log in to unmask]
Sent: Friday, June 18, 2010 3:15:38 PM
Subject: Re: [SPM] Inter-scan movement
hi Eve
> If the human subject has big movement between the consecutive
> functional runs, can we still go ahead to realign all the functional
> runs to the first volume of the first run? And can we still use the
> same anatomical image for coregistration? Thank you!
In general between-session movement isn't such a big concern (because
it's not associated with spikes in the data); so yes, you can align
everything to the first volume of the first run, and use the same
anatomical for coregistration.
Hope this helps,
Jonathan
|