Document Detail


Hybrid methods for improving information access in clinical documents: concept, assertion, and relation identification.
MedLine Citation:
PMID:  21597105     Owner:  NLM     Status:  MEDLINE    
Abstract/OtherAbstract:
OBJECTIVE: This paper describes the approaches the authors developed while participating in the i2b2/VA 2010 challenge to automatically extract medical concepts and annotate assertions on concepts and relations between concepts.
DESIGN: The authors'approaches rely on both rule-based and machine-learning methods. Natural language processing is used to extract features from the input texts; these features are then used in the authors' machine-learning approaches. The authors used Conditional Random Fields for concept extraction, and Support Vector Machines for assertion and relation annotation. Depending on the task, the authors tested various combinations of rule-based and machine-learning methods.
RESULTS: The authors'assertion annotation system obtained an F-measure of 0.931, ranking fifth out of 21 participants at the i2b2/VA 2010 challenge. The authors' relation annotation system ranked third out of 16 participants with a 0.709 F-measure. The 0.773 F-measure the authors obtained on concept extraction did not make it to the top 10.
CONCLUSION: On the one hand, the authors confirm that the use of only machine-learning methods is highly dependent on the annotated training data, and thus obtained better results for well-represented classes. On the other hand, the use of only a rule-based method was not sufficient to deal with new types of data. Finally, the use of hybrid approaches combining machine-learning and rule-based approaches yielded higher scores.
Authors:
Anne-Lyse Minard; Anne-Laure Ligozat; Asma Ben Abacha; Delphine Bernhard; Bruno Cartoni; Louise Deléger; Brigitte Grau; Sophie Rosset; Pierre Zweigenbaum; Cyril Grouin
Related Documents :
20502755 - Ultrasound submucosal inferior nasal turbinate reduction technique: histological study ...
16904885 - An inductive power link for a wireless endoscope.
9339955 - Predictions of mathematical models of tissue oxygenation and generation of singlet oxyg...
23366195 - Random location of multiple sparse priors for solving the meg/eeg inverse problem.
18218385 - Pixel-based reconstruction (pbr) promising simultaneous techniques for ct reconstructions.
12720335 - Robust classifier for the automated detection of ammonia in heated plumes by passive fo...
Publication Detail:
Type:  Journal Article; Research Support, N.I.H., Extramural; Research Support, Non-U.S. Gov't; Research Support, U.S. Gov't, Non-P.H.S.     Date:  2011-05-19
Journal Detail:
Title:  Journal of the American Medical Informatics Association : JAMIA     Volume:  18     ISSN:  1527-974X     ISO Abbreviation:  J Am Med Inform Assoc     Publication Date:    2011 Sep-Oct
Date Detail:
Created Date:  2011-08-17     Completed Date:  2012-01-20     Revised Date:  2013-06-28    
Medline Journal Info:
Nlm Unique ID:  9430800     Medline TA:  J Am Med Inform Assoc     Country:  United States    
Other Details:
Languages:  eng     Pagination:  588-93     Citation Subset:  IM    
Affiliation:
LIMSI-CNRS, Orsay Cedex, France.
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Descriptor/Qualifier:
Data Mining*
Decision Support Systems, Clinical*
Electronic Health Records*
Expert Systems
Humans
Natural Language Processing*
Semantics
Support Vector Machines*
Unified Medical Language System
Grant Support
ID/Acronym/Agency:
U54-LM008748/LM/NLM NIH HHS
Comments/Corrections

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine


Previous Document:  Mapping clinical phenotype data elements to standardized metadata repositories and controlled termin...
Next Document:  Patient safety factors in children dying in a paediatric intensive care unit (PICU): a case notes re...