Dear Mobin,
I don't know if I understand the problem, but the intrinsic
self-connections are always negative, preventing the system suffering
from an "epileptic seizure". In other words, if the self-connections
would be non-negative, activation in such one node would likely
propagate exponentially (? at least quite fast...) and the system would
leave a reasonable steady-state. Such model parameter(s) are not likely
to describe any real in-vivo data well (at least this is my guess).
See also the basic papers on DCM (e.g. Friston et al. Neuroimage 2003
(?)).
Cheers,
Thilo
On Tue, 2011-11-08 at 12:51 -0500, Mobin Anandwala wrote:
> Dear SPM List
>
>
> We would like to know get negative in self connection in Ep.A.
>
>
> Mobin
|