Hi Pia,
you can use DCM with your cap coordinates and without registering with
individual MRIs. That's actually what we're currently doing. The EEG is
spatially quite smooth so that mis-registration of sensors, between
subjects, shouldn't matter too much. The assumption you're making is
that the sensors are in the same scalp coordinates, over all subjects.
This should be approximately true with caps.
Stefan
> We are trying to run a DCM on ERP data that was collected using 128
> channels (biosemi).
> We successfully did the whole preprocessing and design specification.
>
> However, for the DCM analysis:
> we realize that we do not have an MRI structural scan for each
> subject, neither do we have an exact coordinates for the sensors, but
> we were using a conventional 10-20 cap in which Cz was located in the
> center point, equally distance from the two ear lobes and from the
> nasion-inion.
>
> Is there still a hope for us to be able to run a DCM model, or are we
> doomed?
>
> thanks
>
> Pia and Sanjay
>
>
|