Dear Michiel,
To explain how the two-state DCM works, I have written a Wiki page on the topic - https://en.wikibooks.org/wiki/SPM/Two_State_DCM . Thanks to Andre Marreiros for reviewing this. Please let me know if anything isn't clear.
Best,
Peter.
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Michiel Dirkx
Sent: 02 March 2015 17:37
To: [log in to unmask]
Subject: [SPM] Self-connection two-state DCM
Dear SPM'ers,
Hopefully one of you can help me out with my question regarding two-state DCM. In Marreiros paper about two-state DCM they speak of multiple intrinsic (i.e. within one node) connections: E->E (excitatory to excitatory), E->I (excitatory to inhibitory), I->E (inhibitory to excitatory) and I->I (inhibitory to inhibitory) in the Jacobian matrix. However, if you estimate a DCM, the saved DCM.mat will only contain one (log-scaled) value per self connection in the DCM.A, DCM.B and DCM.D expected posterior (Ep.A, Ep.B and Ep.D). My question: which of the above mentioned intrincic connections do these expected posteriors reflect? And let's say there would be a positive Ep.B parameter for that self-connection, would this mean an increase of the excitatory or inhibitory coupling of that self-connection?
Thanks for your help!
With very best wishes,
Michiel Dirkx
|