Medical image interpretation and understanding

A methodological approach is presented to de fine models suitable for the interpretation of complex digital images. The basic entities of this model are the formal descriptions of the textural and structural characteristics peculiar to the images. The structural, or morphometrical, features depend on geometric and topologic properties; the textural properties are related to the intensity distribution of the pixels. The interpretation process is performed on the basis of a knowledge base containing both information extracted from the images under exam and other data belonging fo complementary domains.