Abstract
As long ago as 1878 attempts have been made to define photographically the typical facial characteristics of persons exhibiting particular appearances or afflictions. State-of-the-art numerical computing techniques facilitate definition of highly accurate facial composites. Such images are devoid of ghosting caused by nonalignment of facial features. Composites can help to define the prototypical visual appearance of facial types and as such have applications in visual cognitive psychology and clinical imaging. This article describes the imaging process and discusses how clinical presentation of different diseases can be visualized using physiognomic information and how this has the potential to be developed as a diagnostic tool.