The development of three-dimensional (3D) digital modalities, such as computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET) and single photon emission computed tomography (SPECT), and the replacement of film in 2D imaging techniques by very high resolution detectors, such as those found in digital mammography and digital chest X-ray systems, have led to an explosion of the quantity of digital data that must be stored, transmitted, processed and analyzed in a modern radiology department. That the data are in digital form has distinct advantages. The sheer management of this immense volume of information is best accomplished by computerized systems. For example the images, once in digital form, may be manipulated by elaborate computer graphics and display systems to screen 3D renderings such as those in CT and MR images for reconstructive surgery of the skull and others, as well as for 3D anthropometry and virtual navigation of organs (e.g., Virtual Endoscopy). A more general term can be visual computing.

In one aspect, these digital images can be transmitted at very high speeds to faraway sites for remote diagnosis or online digital collaboration by physicians separated by a great distance. This capability, made possible by recent advances in high-speed digital communication technology, is in the process of revolutionizing the way that clinical radiology will be conducted, named as telemedicine and teleradiology.

In another aspect, perhaps most importantly, the useful information or features in these digital images can be analyzed and extracted to achieve a specific task. The features can be transmitted at an ultra fast speed (due to the smaller size) to remote sites. The features can be very efficiently rendered at a PC computer for visualization. Especially those features, which may not be visualizable by the human eyes, could be converted into visualizable ones. This is a research area of feature-based visualization or information communication between human and computer.

Even more importantly, sophisticatedly-powered computer can process and quantify the image features in a way mimicking the human decision process and, therefore, assist the diagnosis in medicine. This is sometimes called computer aided diagnosis (CADx).