The 3D Reconstruction of the cerebral vascular anatomy can greatly improve the diagnosis of vascular pathologies such as embolisms and hemorrhages. This paper presents a new method for the automatic extraction of the blood vessels surface and the labeling of its features, specifically branching, aneurysms and stenosis. The method works with segmented Magnetic Resonance Angiography images. It extracts a discrete medial axis transform from the stacked images and it vectorizes it. During the vectorization process, the features are detected and labeled automatically. The final output is a labeled continuous medial axis transforms representation of the vascular tree suitable for interactive visualizations and navigations, along with 3D geometrical and topological queries.