SPM12 includes a helper function to adjust the design matrix filtering when concatenating runs. You need to specify the design as a single run and include regressors for each run except the last (i.e. 1 during the run, 0 elsewhere), but don’t estimate the model. Then run spm_fmri_concatenate and pass it the path of your SPM.mat and a vector of number of scans per run, then estimate the adjusted model.
-Mike
--
Mike Angstadt
Research Computer Specialist / PANLab Lab Manager
Department of Psychiatry / University of Michigan
(734) 936-8229
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Colin Sauder
Sent: Wednesday, January 11, 2017 3:38 PM
To: [log in to unmask]
Subject: [SPM] concatenating multiple runs
I had a specific question about concatenating runs at the first level analysis. We have three separate runs of the same task. Typically I of course analyze each run as separate session during first level specification. However, I'm interested in a low probability event that occurs unevenly across runs (errors). I'm hesitant to create a contrast per run, as some runs won't have any errors, while others may only have 1-2 errors in a given run. However, if I were to concatenate across all three runs, my contrast would have at least 10 errors, which I'm more comfortable with.
Can someone speak to the feasibility of treating all three runs as one session for this specific analysis only? Obviously this is not ideal, but is it possible? If so, are there any ways to account for between-run motion?
Thanks,
Colin
--
Colin L. Sauder, Ph.D.
Assistant Professor, Child and Adolescent Psychiatry
Department of Psychiatry
University of Texas Health Science Center at San Antonio
7703 Floyd Curl Drive MC 7719
San Antonio, TX 78229
V: 210-567-5417
F: 210-567-5677
mailto:[log in to unmask]
**********************************************************
Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues
|