An experimental framework for learning the medical image diagnosis

In this paper we propose a framework in which some methods are compared to achieve the medical diagnosis based on image analysis. The proposed framework includes besides components for the extraction of low-level features, methods for the integration of semantic knowledge about the medical diagnosis into the retrieval process. So, the paper approaches modalities for learning the medical diagnosis using low-level characteristics automatically extracted from the visual content to generate high-level concepts by means of semantic association rules. The experiments through this experimental framework were realized on medical collections of images.