This paper describes how patient specific volumetric data are managed from image acquisition through final processing for the purposes of creating a 3D VR rendering of user selected and manipulated 3D models. The system described here allows for the development of quick, inexpensive, and clinician manipulated patient-specific models. The utility of this process is demonstrated by being able to move VRML models to desktop or immersive environments for both pre-operative planning and patient-specific surgical and anatomical training.