Dear all, I am constructing my own VOI structures using the time series for inputting in to DCM. That means I already have the time series which I want to input in the DCM. I found that I only need 2 parameters xY.u = eigen variate time series (mean time series from the ROI/VOI) xY.name = name of the VOI However this way I will get one xY for each subject and each VOI. I think there should be only one xY per VOI, but I will get one for each subject which is wrong. Can somebody please suggest how to combine the time series extracted from the VOIs of different subjects in to one ? I know one simple way might be to just take a mean, but I m not sure if thats the correct approach. Is that what SPM uses ? Best regards, Rabia