Dear all,
I have resting state eeg-fMRI data. I already ran correlation analysis between the eeg signal (from certain electrodes) and the BOLD signal. I would like to test a certain hypothesis about how different parts of the brain (that correlate positively or negatively with the eeg signal) could explain the EEG findings. I am thinking of using the eeg signal (which can be convolved with the haemodynamic response, but from my understanding this can be done part of the DCM) as an input in the model, much like the way as I used it as an input in the GLM analysis. Does that sound like a reasonable approach or should the input always be a stimulus or an experimental task?
Best regards,
George
|