Dear Paul,
> I have a basic question concerning event-related fMRI. Is there a rule
> concerning the duration of the TR and SOA? In other words, does the SOA
> have to be longer than the TR and if so by how much?
No - assuming that you're using multi-slice EPI then the relevant variable
is the slice acquisition time (of the order of ~100ms), not the time to
acquire an entire volume. The relevant issues are (1) linearity of the
evoked response as SOA decreases and (2) ensuring adequate sampling of the
haemodynamic response by each slice.
Existing evidence suggests that (at least for some cortical areas) linear
superposition of haemodynamic responses to successive events starts to break
down at <1.5s or so (Friston et al 1993 MRM). So using an SOA of >1.5s would
probably be prudent.
The second issue regards ensuring each slice of the volume adequately
samples the haemodynamic response. This entails avoiding a constant phase
relationship between the timing of events and the timing of successive
volume acquisitions (because then each slice of the volume would sample only
a fixed portion of the haemodynamic response). In practice you can do this
by either ensuring that the SOA is a non-integer multiple of the TR (thus
making the event onsets progressively 'walk through' the acquired volume) or
by adding a 'jitter' to the SOA.
These aren't the only timing issues - the temporal realignment issue is
distinct and has been dealt with in other recent postings. Josephs & Henson
(1999) also has lots of useful information regarding optimisation of
event-related designs.
> I have modeled a study with a TR of 2.1 and an SOA of 1.9. Will I have
> problems with this?
Sounds fine - I've just had excellent results with a visual motion paradigm
using an SOA of 2.1s and a TR of 2.8s. Mind you, you never can tell...good
luck!
best wishes,
Geraint
--
Dr. Geraint Rees
Wellcome Advanced Fellow, Lecturer,
Division of Biology 139-74, Institute of Neurology,
California Institute of Technology, University College London,
Pasadena CA 91125 London WC1N 3BG
voice 626-395-2880 020-7833-7472
fax 626-796-8876 020-7813-1420
web http://www.klab.caltech.edu/~geraint
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|