I want to figure out the casual relationship among three brain regions and DCM is widely known to solve this problem. Does AFNI have a way to calculate DCM or some alternative methods? Do I have to transform the .brick + .head format to a format that SPM can read ?
We do not have an implementation of DCM. We do, however, have other programs that compute some sort of connectivity including SEM (Structural Equation Modeling), Granger causality, SVAR (Structural Vector Autoregression), PPI (psychophysiological interaction), as well as a variety of ways to compute correlations. See this document for an overview of each method: