Plataforma para la búsqueda por contenido visual y semántico de imágenes médicas

Abstract

This paper present a framework ontology-independent for the automatic semantic annotation of medical images. The automatic annotation is done by (a) semantifying of DICOM medical images (RDF-ization) automatically; (b) Integration of different biomedical ontologies, through the matching process between ontologies and DICOM metadata, making this approach ontology-independent; (c) segmentation and visualization of annotated data which is further used to generate new annotations according to expert knowledge, and validation. Additionally applying context based image retrieval, make possible the retrieval of medical imaging by similarity of images features. This platform is being built on a distributed architecture, which improve the way of classification, distribution and searching on image repository.