Dear SPM experts,
May I ask how can I get the intrinsic connection of DCM for M/EEG data. According to my understanding, the intrinsic connection is in the Ep.G. However, the number of intrinsic connections seems less than that in the spm function. For example, according to my understanding and what I found in spm_fx_cmc.m, there should be around 10 intrinsic connections for each ROI in the DCM for M/EEG data. But I just found 4 intrinsic connections for each ROI in DCM for M/EEG data. May I ask why and what usually should be used in the Ep of DCM for M/EEG, apart from the A, B and C matrices? Would this problem be caused by the function spm_dcm_erp I used to estimated the DCM model? Please see below for the comments I found in spm_fx_cmc.
%--------------------------------------------------------------------------
% free parameters on time constants and intrinsic connections
%--------------------------------------------------------------------------
% G(:,1) ss -> ss (-ve self) 4
% G(:,2) sp -> ss (-ve rec ) 4
% G(:,3) ii -> ss (-ve rec ) 4
% G(:,4) ii -> ii (-ve self) 4
% G(:,5) ss -> ii (+ve rec ) 4
% G(:,6) dp -> ii (+ve rec ) 2
% G(:,7) sp -> sp (-ve self) 4
% G(:,8) ss -> sp (+ve rec ) 4
% G(:,9) ii -> dp (-ve rec ) 2
% G(:,10) dp -> dp (-ve self) 1
Another question is how to apply the function spm_fx_cmm_nmda properly? Where I can found the NMDA, AMPA and GABA parameters in the DCM for M/EEG when I use the NMDA model, as I read in the paper "Ion channels in EEG: isolating channel dysfunction in NMDA receptor antibody encephalitis"?
Looking forward to your kind suggestions. Many thanks.
Best,
Zhi Li
|