NLP for Clinical IE

AIME 2017 Tutorial: Natural Language Processing for Clinical Information Extraction

21 June 2017

This tutorial will focus on a high-level introduction to NLP applied to clinical text. It will include a presentation of NLP and information extraction, characteristics of clinical text, existing NLP tools applied to clinical text, an overview of clinical NLP projects preparation and tools evaluation, and hands-on exercises with NLP tools to illustrate the theory presented. This tutorial addresses the need in the community to learn through hands-on experience how to apply and experiment with NLP tools that are publicly available.

A set of de-identified clinical documents will be used to illustrate the process of implementing an NLP tool. The documents will be described along with the target concepts that will be extracted.

Aims: At the end of this tutorial, participants will be able to

  • Identify and explain the characteristics of clinical text
  • Distinguish the various methods used to extract information from clinical text
  • Compare these information extraction methods, existing tools, and terminological resources
  • Benefit from hands-on experience with a selection of NLP tools adapted to clinical text


Content details

Presenters:

Stéphane Meystre

Medical University of South Carolina, Charleston, SC, USA

meystre@musc.edu

Meliha Yetisgen

University of Washington, Seattle, WA, USA

Scott DuVall

University of Utah & Department of Veterans Affairs Salt Lake City Health Care System, Salt Lake City, UT, USA

Hua Xu

University of Texas Health Science Center, Houston, TX, USA