Print

Print
Dear Robin,

Firstly, we think that DCM can be performed on resting-state as well as task-related functional data. Are there particular constraints or interpretation difficulties associated with doing DCM on resting-state data?

You can interprete resting state (spectral) DCMs as how you would do for task related DCM. You can see examples here:

http://www.sciencedirect.com/science/article/pii/S1053811915000531

http://www.sciencedirect.com/science/article/pii/S1053811914009446

Secondly, with regard to the sample size, is 40 participant enough for doing DCM analysis?

Probably yes, depends on several factor, foremost the numbers within each group.

Thirdly, we think that DCM requires relatively short TRs. Is a TR of 2 seconds short enough?

You would be fine with TR = 2sec (infact this is what we used in most simulations for spectral DCMs).

I hope this helps.

Adeel

On Mon, May 11, 2015 at 12:58 PM, Robin Shao <[log in to unmask]> wrote:
Dear experts:

We are currently analyzing fMRI resting-state and task-related data during affective processing, focusing on the region of raphe nucleus. We are interested in the (temporal) causal relationship between signals in the RN and those in the rest of the prefrontal, limbic and perceptual networks, so we think maybe the dynamic causal modelling approach can be utilized to answer this question. However, we have a few questions with which you might be able to help:

Firstly, we think that DCM can be performed on resting-state as well as task-related functional data. Are there particular constraints or interpretation difficulties associated with doing DCM on resting-state data?

Secondly, with regard to the sample size, is 40 participant enough for doing DCM analysis?

Thirdly, we think that DCM requires relatively short TRs. Is a TR of 2 seconds short enough?

Many thanks in advance for your help.

Best regards,
Robin