> It should be difficult to discern in case of a design for which we expect slow frequency
> signal. Which brings us back to the question why we go with designs like that in the first place

Right, it's challenging to handle a block design with a long duration indeed because slow drift might be confounded with the true signal. And another complication is that the slow drift is caused by a mixture of multiple sources such as scanner thermal fluctuations as well as physiological confounds such as cardiac and breathing effects. I'm not so sure whether the following approach would work (without using polynomials): 1) obtain cardiac and breathing signal, and remove them from the EPI data; and 2) use the EPI signal in white matter as a regressor.

Gang



Edited 1 time(s). Last edit at 06/18/2015 11:02AM by Gang.