3 D Display with Minimal Predefinition

The portrayal of medical images in three dimensions can be strikingly improved in presentational usefulness and in user time requirements with methods based on volume rendering rather than surface rendering . We describe methods that lead to such improvement. These involve, in order of application, region-of-interest definition, as opposed to object definition; fuzzy surface specification; rendering not only fuzzy surfaces but together with them, surface models and grey-level slices; and fast display. These techniques lead to improved usefulness in many areas of medicine, including radiation oncology, surgery, multimodality image display, and diagnostic imaging from CT, MRI, SPECT, PET, and ultrasound.