In this article the present state of medical image processing is introduced with the stress on two fields - diagnostic aids using image pattern recognition/understanding and aids in treatment using recognition and generation of images. Major topic in the first field is computer aided diagnosis of X-ray images in the screening of cancer. The main goal here is ``marking'', giving a mark on suspicious shadows in a given film. Processing of mammograms is in the most advanced stage. The most important theme in the studies of next generation screening system is the use of chest CT for detection of very early stage of lung cancer. In the latter half of the article, the concept of the virtualized human body and its application to diagnosis and treatment are presented. To search or explore the virtualized body moving around freely inside it is considered as diagnosis, and we call it the navigation diagnosis. As an example, the virtual endoscope system is introduced. Simulation of surgical operation ( pre-operative aids ) and the augmented reality for intra-operative aids are explained briefly.

View full abstract