To develop and validate a quantitative and observer-independent method to evaluate pial collateral circulation by DSC-perfusion MRI and test whether this novel method delivers diagnostic information which is redundant to or independent from conventional penumbra imaging by the mismatch approach.