Dear DCM experts...
Some of our recent research has led us to hypothesise
about how a given brain region (A) might 'mediate' connectivity between 2
other brain regioons (B and C) - ie how A might determine to what extent B
and C exert influence over one another, depending on the presence of certain
form of stimulation or a certain task context (this is something like a
'psycho-physio-physiological interaction').
Whilst it's clearly possible to use correlation-based approaches to
connectivity to address this problem, I have been trying
to think about how the problem might be addressed with model-based
approaches such as DCM. Could one, for example, create a 'virtual' VOI
consisting of the
interaction between B and C, and enter this into the DCM? Other than just
multiplying the timeseries in Y, how else would one need to adapt the
parameters in xY? Are there other pitfalls I should be aware
of?
many thanks,
Chris
Christopher Summerfield
[log in to unmask]
|