A global description of medical image with high precision

Abstract

Medical Imaging suffers from different problems. This paper explores the authors' solution that aims to provide efficient retrieval of medical imaging. Depending on the user, the same image can be described through different views. In essence, an image can be described on the basis of either low-level properties, such as texture or color; context, such as date of acquisition or author; or semantic content, such as real-world objects and relations. The authors' approach consists of providing a global description solution capable of integrating different dimensions (or views) of a medical image. The description problem of medical images during both storage and retrieval processes is studied. Few proposed solutions take into consideration the heterogeneity of user competence (physician, researcher, student, etc.) and the necessity of a high expressive power for medical imaging description. For example, spatial content in terms of relationships in surgical or radiation therapy of brain tumors is very decisive because the location of a tumor has profound implications on a therapeutic decision. Visual solutions are recommended and are the most appropriated for non computer-scientist users. However, current visual languages suffer from several problems, especially ambiguities generated by the user and/or the system at different levels of image description, imprecision and no respect of the integrity of spatial relations. This framework exposes the authors' solution showing how this problematic can be resolved. An implementation has been realized to prove their proposition.