Dear everyone,
someone might have mentioned this and I missed it, but if not.
I think the reason why event related studies have been suggested to be possibly
less sensitive to motion has do do with the hemodynamic delay.
Given that you have movement that are related to events (e.g. movements due to
button presses in response to phasic stimuli) one would expect the movement
effects to manifest immediately, i.e. in the present and possibly next EPI
volume. The neural effects on the other hand will (thanks to the delay) manifest
themselves in subsequent volumes. Hence, for very short TRs one might hope for a
reasonably good separation between movement and neural effects "automatically",
and for longer TRs one might consider using a modified HRF to model mainly later
effects where movement effects are no longer expected.
I think there was a presentation along these lines at HBM in Montreal. I am not
really sure I believe it works, but I think the general idea is as above.
Jesper
Andersson
[log in to unmask] wrote:
> Hi,
>
> I think one possibility is that averaging multiple trials diminishes the
> contributions of any one trial in which motion was particularly large, just
> like spatial averaging would do.
>
> -jim
>
> >Dear James,
> >
> >>
> >> I my experience event-related designs are only "less susceptible" to
> >> artefactual activation if you average multiple trials into a single
> >> event type. Without averaging multiple trials the degree of artefactual
> >> activation appears to scale with the size of movements just like in
> >> block design experiments.
> >
> >
> >Do you have any ideas about why this should be the case (i.e., why averaging
> >should reduce the effect of motion?).
> >
> >Thanks for any reply,
> >Eric
>
> -----------------------------------
> Jim Eliassen, Ph.D.
> Box 1953
> Dept. of Neuroscience
> Brown University
> Providence, RI 02912
> USA
>
> For courier service--
> ADD : 190 Thayer Street, Metcalf Research Laboratory, Room 337
> DELETE : Box 1953
>
> office 401-863-9984
> lab 401-863-1874
> fax 401-863-1074
> -----------------------------------
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|