Automated search tools to retrieve published rare disease cases

Details

Last Updated on Wednesday, 05 November 2014 06:28

Automated search tools to retrieve published rare disease cases would improve patient identification and diagnosis
 
As the number of published clinical case reports and literature on rare diseases increases, so does the need to develop automated tools to identify and analyse the literature. This is particularly relevant in the field of rare diseases, where practitioners rely frequently on manually retrieved case reports to support their diagnosis of patients with rare conditions. In an article published in Database, Taboada et al. propose their methods of identifying and annotating relevant case reports in efforts to improve phenotypic descriptions for rare disease diagnosis.

Currently available tools, such as PubMed’s Medical Subject Headings (MeSH) standard terminology thesaurus, GoPubMed search engine based on the Gene Ontology (GO) and MeSH and SEGOPubmed, offer various methods of retrieving abstracts using key term searches. While MeSH and GO are the most frequently used tools, others such as the Open Biological and Biomedical Ontologies (OBO), the National Center of Biomedical Ontology (NCBO) bioportal and the Human Phenotype Ontolgy (HPO) are more suitably adapted to certain diseases.

Taboada et al. suggest developing methods, using the various available tools, to facilitate automated and exhaustive extraction of reports on patients with phenotypic similarities. The authors propose techniques to generate semantic patient data indexing using linguistic patterns which can be further analysed. They believe their approach would contribute towards data curation for rare disease indexing and analysis.

Read the open access article