During the last decade, the convergence of medical imaging and computational modeling technologies has enabled tremendous progress in the development and application of image-based computational fluid dynamics modeling of patient-specific blood flows. These techniques have been used for studying the basic mechanisms involved in the initiation and progression of vascular diseases, for studying possible ways to improve the diagnosis and evaluation of patients by incorporating hemodynamics information to the anatomical data typically available, and for the development of computational tools that can be used to improve surgical and endovascular treatment planning. However, before these technologies can have a significant impact on the routine clinical practice, it is still necessary to demonstrate the connection between the extra information provided by the models and the natural progression of vascular diseases and the outcome of interventions. This paper summarizes some of our contributions in this direction, focusing in particular on cerebral aneurysms.