Project

Goal

Participants

Publications

Resources and Tools

References

intranet

 

    Goal

    1. Provide solutions to the problem of accessing big volumes of information and knowledge by the healthcare systems, using techniques from Natural Language Processing (NLP) and Information Retrieval (IR).

    2. Creating efficient filters so that people working at healthcare will have the possibility of accessing and retrieving information adequate to their queries over millions of documents written in English and Spanish, by means of information technologies.

    Application to specific cases:
    • Electronic Health Records (EHRs) from the hospitals of Galdakao and Basurto, for the identification of Adverse Drug Reactions (ADRs)
    • Orphanet, to identify disabilities associated to rare diseases

    Two subprojects:

    Subproject DETECT:

    Objectives:
    • Detection of ADRs in EHRs using technologies from NLP (UPV-EHU).
    • Adapting NLP resources for English, Basque and Spanish to the medical domain, so that the linguistic processors will be available for new applications in medicine.

    Subproject MED-RECORD (UNED):

    Objectives:
    • Aplication of NLP and IR techniques to the problem of identifying disabilities associated to rare diseases from heterogeneous information sources (scientific articles, Tweets, ?)
    • Generalization of the techniques developed for the identification of disabilities associated to rare diseases to ADRs, evaluating the accuracy of ranking algorithms based on heuristics on the medical domain.

Highlights



IXA Group - EHU-UPV

Itzulbide ikerketa-proiektua 2019-Oct-01

Visitor: Andrea Horbach, Automatic scoring 2019-Sep-20

One of the best three papers on Clinical NLP in 2017 was published by Ixa Group 2019-Jun-28

Mitxelena Award for PhD theses 2018 to Olatz Perez-de-Viņaspre: Automatic medical term generation 2019-May-24

Read more IXA