Biomedical Literature Retrieval Based on Patient Information

Abstract

Information and Communication Technologies has led to a biomedical data explosion. A proportional growth has been produced regarding the amount of scientific literature, but information retrieval methods did not follow the same pattern. By using specialized clinical search engines such as {PubMed}, Medscape and Cochrane, biomedical publications has became instantly available for clinical users. However, additional parameters, such as user context, are not taken into account yet. Initial queries still retrieve too many results without a relevance-based ranking. The objective of this work was to develop a new method to enhance scientific literature searches from various sources, by including patient information in the retrieval process. Two pathologies have been used to test the proposed method: diabetes and arterial hypertension. Results obtained suggest the suitability of the approach, highlighting the publications related to patient characteristics.