Our understanding of the working brain is partly based on observations through the "windows of the brain", electric (EEG/MEG), metabolic (PET), or hemodynamic (fMRI,OT). To shed light on the intricacies of the neurovascular coupling, and in order to model phenomena such as CSD, it is important to work towards an integrated model of the three key players, electrophysiology, metabolism, and blood flow. Finding a functioning detailed model of each one of these phenomena separately is already a challenge; integrating them requires that one addresses the multiple spatial and temporal scales characteristic to each one of them. This talk describes the ongoing work towards a computational predictive integrated model, addressing some of the challenges due to the multiscale nature of the problem.