>In my event-related design I am able to record the subjects' actions to
>0.1 seconds with video taping. Since the event time is known
>accurately, is using the temporal derivative recommended?
It still may be useful to include the temporal derivatives in your design
matrix because they permit regional (voxel-by-voxel) variation in the onset
of the HRF over a small temporal window. So, for instance, if your
stimulation produces a very quick response in early visual areas which
peaks before 6s, a more canonical response in temporal areas, and then a
slower response peaking at 7-8s in frontal areas, without TDs you will have
less sensitivity in visual and frontal areas because your model won't
permit deviation from the canonical HRF. For a typical first level
analysis, the hit in terms of degrees of freedom is normally
negligible. This, BTW, assumes you're doing continuous sampling (as
opposed to some kind of sparse sampling design).
Someone once told me that the 6s-peak in the HRF was based primarily on
work in primary visual cortex. Anyone know if that's true and whether it's
been tested empirically for other areas?
Hope this helps.
Joe
Joseph Devlin, Ph. D.
FMRIB, Dept. of Clinical Neurology
University of Oxford
John Radcliffe Hospital
Headley Way, Headington
Oxford OX3 9DU, U.K.
Phone: +44 (0)1865 222 738
Fax: +44 (0)1865 222 717
Email: [log in to unmask]
|