Three-dimensional processing of medical images

Abstract

There are currently in use several medical imaging modalities such as: computed tomography, magnetic resonance imaging, positron
emission tomography, single photon emission computed tomography etc. which are capable of generating information of the human body in the form of slices. On the other hand the study of the brain electrical activity generates functional images which are also widely used nowadays. However the quality of the diagnosis sometimes depends on the tools available for the visualization of the image information. In the normal clinical practiee the image data usually are evaluated by visual investigation of 2D gray scale images. In this way the information contained in the original data cannot be fully orploited since it requires both extensive training and naturd skill to be able to visudize 3D phenomena based on slice representations alone. \[hen using topographic maps 2D visualization doesn't allow carrying out a more precise localization of a given activity. During the last years several techniques for computer aided 3D visualization have been developed in order to overcome the above mentioned difficulties. These techniques al. low the generation of 3D views of the studied volume and its usefulness has been already demonstrated in planning of several kinds of surgery, but they have found limited use in strictly diagnostic applications. This paper presents some considerations for processing in three dimensions tomographic image volumes.