You are here

NLM's I2B2 tool system description.

Printer-friendly versionPrinter-friendly version
Mork JG, Bodenreider O, Demner-Fushman D, Doagan RI, Lang F-M, Lu Z, Névéol A, Peters L, Shooshan SE, Aronson AR
Third i2b2 Workshop on Natural Language Processing Challenges for Clinical Records, November, San Francisco, 2009
Abstract: 

The i2b2 medication extraction challenge provided us with an opportunity to assess the usability of publicly available drug-related resources on clinical text and to contribute to the generation of a publicly available collection of annotated clinical notes. The challenge also presented us with a chance to evaluate how MetaMap, our UMLS concept recognition tool, would work on discharge summaries and to roll the knowledge gained back into MetaMap development. Our approach to identify drug-related entities within the scope of this challenge relied on the use of lookup lists and rules built solely with publicly available resources. Preliminary results show promise with the clinical drug information specific entity lists. However, more sophisticated methods will be needed to improve the identification of the reason and duration elements of drug mentions.

Mork JG, Bodenreider O, Demner-Fushman D, Doagan RI, Lang F-M, Lu Z, Névéol A, Peters L, Shooshan SE, Aronson AR. NLM's I2B2 tool system description. Third i2b2 Workshop on Natural Language Processing Challenges for Clinical Records, November, San Francisco, 2009