Hi Bing,
the basic answer is: yes, you can use SPM5 for that.
Simplest way to do that is to set up a box-car design that models the "ON"
periods (with the OFF periods being what is left between the ON
periods...). Apply that on your individual time series that you have
preprocessed adequately, of course. A contrast of +1 set on you box car
will then give you the activations, and if sufficienty demanding in terms
of cognitive load, the negative contrast will give you the DMN.
ICA: yes, can be applied too. Totally different approach, because here the
DMN is not extracted by its activation/deactivaion characteristic, but
based on the synchronicity of the participating anatomical structures.
There are several tools out there for ICA, such as Matlab based GIFT, ot
FSL.
In any case, it seems that you should read some of papers that deal with
the "deactivation" DMN pattern or its network features (e.g . Fox et
al.)...
hth,
Philipp
> Dear SPMers:
>
> I have a question of long resting period. I am currently have a new
> project that involves long resting period to investigate the connectivity
> of the DMN (default mode network) and its deactivation. There are two long
> rest periods each for 3 minutes.
>
> There are four runs. The first 3-min. resting period is at the end of the
> second run and the second 3-min. resting period is at the end of the last
> run. I wonder if SPM5 is suitable to analyze these resting-periods with
> the task blocks to see the activation/deactivation of the brain ( I mean
> using the contrast between long resting-period and task blocks). If it is
> not, any suggestions about what program I could use. Is ICA possible to do
> it?
>
> Thank you in advance!
>
> Bing
>
Max-Planck-Institut für Psychiatrie
AG Kernspintomographie
Kraepelinstr. 2-10
80804 München
Tel. 089-30622-413
Fax 089-30622-520
|