Dear Frederik,
The mean-centring option in DCM for fMRI (DCM.options.centre) was introduced as part of stochastic DCM - an extended model which enables endogenous noise in the brain to drive the dynamics. So for a standard DCM analysis, leaving DCM.options.centre=0 gives results compatible with earlier versions of DCM.
If you turn this on - DCM.options.centre=1 - then it mean-centres the input regressors. This changes the interpretation of your a-matrix. Rather than being the baseline in the absence of input, it becomes the average connectivity across inputs.
Best
P.
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Frederik Van de Steen
Sent: 20 August 2015 08:49
To: [log in to unmask]
Subject: [SPM] DCM and centre input
Dear list,
Could someone explain when (and why) it is appropriate to centre the input when setting up DCM's? Many thanks.
Kind regards,
Frederik Van de Steen
|