System for Medical Image Retrieval The MIMS Model

Abstract

The multifaceted description of image data presents a couple of problems for traditional information systems designed for textual data. Image databases require efficient and direct spatial search based on image objects and their relationships instead of some cumbersome alphanumeric encodings oft he images. Current systems generally approach the problem from different views. Thus, they describe one image facet dependently on application domain. This paper presents a model for Medical Image Management System (MIMS) that allows physicians to retrieve images and required information by combining the strengths of several approaches. Via a convivial iconic interface, our system assign to each image object a graphical representation to describe image objects and their attributes. It allows also to automatically calculate their interrelations.