Hydrocephalus is a neurological disease which causes ventricular dilation due to abnormalities in the cerebrospinal fluid (CSF) circulation. Although treatment via a CSF shunt in the brain ventricles has been performed, poor rates of patient responses continue. Thus, to aid surgeons in hydrocephalus treatment planning, we propose a geometric computational approach for tracking hydrocephalus ventricular boundary evolution via the level set method and a mesh warping technique. In our previous work , we evolved the ventricular boundary in 2D CT images which required a backtracking line search for obtaining valid intermediate meshes. In this paper, we automatically detect the ventricular boundary evolution for 2D CT images. To help surgeons determine where to implant the shunt, we also compute the brain ventricle volume evolution for 3D MR images using our approach.