Dynamic functional connectivity (dFC) analysis has gained considerable interest in the past years. The goal of this technique is to estimate the temporal changes of resting state functional connectivity networks and get insights into brain pathologies by analyzing these dynamic patterns. dFC uses functional connectivity correlations as a means to understand the brain functional principle. Dynamic Causal Modeling (DCM) has been widely used in the neuroimaging community to estimate the effective connectivity by fitting a neuronal model to the observed fMRI data. Stochastic DCM together with Bayesian Model Comparison applied to resting state fMRI data results in the selection of the most plausible neuronal model explaining the observed data. The input to these model estimation methods are the full length time series extracted from the regions of interest of mouse resting state fMRI data, neglecting the temporal evolution of the model parameters. In this work we combine the two approaches by estimating the temporal changes in the effective connectivity as derived from DCM.