Ontologia de imagens do domínio da nefrologia a partir dos atributos visuais e verbais dos laudos e prontuários de pacientes

Abstract

The images from health’s field are in high feature, that’s because they become part of human body as extension, including the movements after the capture of a sick organ or not. They make new discovering about the state of the art from a person and, consequently, a major precision on diagnosis investigation and pathologies treatment. It’s a very important information resource for this knowledge field and, therefore, they need information treatment. One of the alternatives is the ontology building, concept extract from platonic and aristotelic philosophy, reappearing, reminding, on Cognitive Science field, as treatment tool, organization, retrieval and information management. The literature shows that ontologies have been build in an vast knowledge fields researches. Nevertheless, we didn’t find any ontology covering nephrology. This fact motivate us to make an research, in mind of Universal/Project, with the basic objective of building an ontology of image from this knowledge domain, with emphasis to nephropathies, leading to mapping, retrieval and image management in electronic environment. Present image ontology of the nephrology domain from discourses analysis from patients records annotations. The empirical study has been made on Medical Record and Statistics Service from Hospital Universitário Walter Cantidio- Universidade Federal do Ceará. The building of this ontology has been made in two moments: we first used the Protégé software from Stanford University, where we builded the text-based ontology. The next step, we imported this ontology for the Active Media Software - Ontology Based Annotation system -University Sheffield- in attempt to build that ontology in nonverbal text (image). The results show that the union of both softwares is extremely important for ontologies building of verbal and non-verbal texts.