Medical 3D visualization needs image understanding

This contribution focusses on two problems: one is the 3D visualization of 3D data from sources like e.g. computer tomographies (CT), magnetic resonance tomography (MR) and comparable devices like PET, SPECT or even ultrasound (US). The other one describes the necessity to firstly segment or classify objects in the 3D data cube before you can illuminate or display them. There are promising approaches to the visualization part. The real problem is the primary segmentation which is not solved at all, there are no robust algorithms.