AUTOMATIC INDEXER
AutoIndexer: Research and Development of Methodologies and Terminological Resources to Support the Indexing Processes of Clinical Reports in Biomedicine and Health Areas
REFERENCE
TSI-020100-2009-252
DURATION
24 meses (2009-2010)
FUNDING RECEIVED
Presupuesto Financiable:390.882€ Subvención propuesta: 205.745€
ROLE OF THE COMPANY
Coordinador/a – socio/a
PROGRAMME
Avanza R+D. Strategic Action for Telecommunications and the Information Society 2009-2010 The main purpose of the AutoIndexer project is to establish an infrastructure of algorithms and web services for the automatic indexing of electronic health documents written in a natural language. AutoIndexer tries to research and develop a variety of natural language analysis techniques and algorithms applied to the specificities of the medical domain, for example: · Identifying negation and doubt · Expanding acronyms based on the context and disambiguation · Correcting spelling errors · Identifying, indexing and labelling named entities: diseases, medical procedures, causes, symptoms, signs etc. AutoIndexer is a project designed to solve current indexing needs of semi-structured complex documents such as those used in electronic health records and simplify later coding processes of information based on the standards of the international classification of diseases ICD-9-CM.

SUMMARY

The project’s scope of application covers all health centres equipped with structured or semi-structured electronic information. AutoIndexer is complementary and provides support for coding and automatic classification systems, clinical alert and expert review systems, decision-making, coupling of knowledge and automated administrative support for certain healthcare support tasks. The results of the project help to make the most of the information stored in EHRs and ensure it’s useful for coding and classification tasks, speeding up selective document retrieval and recognising predominant structures in clinical information to make informed decisions on models and archetypes as an essential requirement during the entire adaptation period.
COMPANIES OR COLLABORATING INSTITUTIONS
School of Knowledge Engineering of Complutense University of Madrid