Dear SPm users,
I am planning to perform a DCM on EEG data using SPM12 and have a few questions:
I understand from the manual that it is possible to perform DCM on the channel level (by using the LFP option), but it does not seem advisable/valid to me for channel-level EEG data. What is your recommendation for EEG data in general – should one always perform source localization before DCM or are there valid ways to work on channel level?
From the number of publications I’ve come across so far, it seems to me that DCM for EEG is still relatively uncommon – which I thought could be related to the general low quality of EEG data. Given the low SNR of EEG signals, how useful in general do you assess DCM for EEG to be?
Since I am planning to analyze oscillatory processes (rather than ERPs), I understand that I am restricted to the use of a phenomenological (rather than a physiological) model. This made me wonder whether DCM is indeed the most suitable method for this type of data or whether you may recommend other better methods to assess effective connectivity.
I would be grateful for your inputs – any insights from your side would be highly appreciated!
Thanks
Giovanni
|