(sorry for re-post, I forgot the subject field this afternoon)
Dear SPM-list,
I have a question regarding onsets of events occurring at times <0
(before scanning started). For events with considerable duration, such
as blocks, neural activation should be present around t=0, that was
initiated before t=0. In SPM this is problematic because one cannot
enter onset times <0. When convolving with HRF, it might make sense to
allow this, to have partially evolved hemodynamic responses at t=0, it
would make sense for things one measured (RTs, task, respiration,
whatever) starting before scanning. One would have to scan dummy scans
(not recorded but RF pulses administred, we usually do that to 'heat up'
the signal before starting) before t=0 when allowing onset times<0.
otherwise the signal wouldn't be at baseline yet when starting.
I am asking this on behalf of colleages that have datasets on a task
that started at dummy scan 1, 3 scans before actual MR data acquisition.
We thought of the following options:
1. add 3 scantimes to all onsets, create SPM.mat, and then chop of the
first 3 rows of the design matrix SPM.xX.X using matlab scripts, and
save SPM.mat again. I guess that is a problem because filter vectors,
whitening matrices and possibly other settings would be off as well.
2. Add 3 scans before the first one containing rest scans or the average
of all scans, and add 3 to all onset times. Add a regressor with 1s at
scan 1-3, zero otherwise, to remove a possible effect of these
artificial volumes. This would cost one a degree of freedom, and
erroneously win 3 again by adding scans (on the first level).
3. Tweak the spm code in such a way that onsets<0 are allowed, so that
the convolved model is already evolving around t=0. Did anyone try this?
I will try it when no-one did it before.
Did anyone encounter this problem, or has any advice on how to tackle
this correctly?
Thank you for any help,
Bas
--------------------------------------------
Dr. S.F.W. Neggers
dept. of Psychonomics,Helmholtz Institute
Utrecht University
Heidelberglaan 2
3584 CS, Utrecht, room 17.09
the Netherlands
Tel: (+31) 30 253 4582 Fax: (+31) 30 2534511
E-mail: [log in to unmask]
Web: http://www.fss.uu.nl/psn/pionier
--------------------------------------------
|