VA Research holds conference on natural language processing
Natural language processing (NLP) is the ability of software programs to recognize and make sense of human speech as it is spoken, or everyday language as it is typed into a computer. It is a component of artificial intelligence, which is the simulation of human intelligence by machines.
VA Research has supported more than 20 NLP studies since 2009. The growth of electronic heath records (EHRs) has spurred the need for such studies. NLP can be used to quickly extract information from text notes and other unstructured elements of these records, on a large scale. The data can then be analyzed to help answer medical questions, solve problems, and improve care.
On Sept. 9 and 10, 2015, VA's Health Services Research and Development Service (HSR&D) convened a "state of the science" conference on NLP at VHA's National Conference Center in Washington, D.C. More than 20 VA researchers attended the conference. In addition, 17 other leading NLP researchers attended, including participants from IBM, 3M, Kaiser Permanente, the Mayo Clinic, and Microsoft Research. Other VA offices, including the Office of Patient Care Services and the Office of Information and Analytics, were also represented.
In his opening presentation to the group, Dr. David Atkins, HSR&D's director, named a number of research priorities that could benefit from NLP, including creating and refining disease phenotypes (the traits of an organism that can be observed, such as appearance, behavior, and metabolism) for genomic research.
Atkins suggested NLP could improve researchers' ability to control for confounding variables in comparative effectiveness research and also improve the ability to identify outcomes in research on quality and safety issues. He told attendees VA hopes to move from research that uses NLP as a tool to solve narrow problems, to research that is more innovative and advances more generalizable solutions.
Discussions at the conference also focused on how NLP might be used to better understand patients' experiences of care and coordination of care; to monitor for changes in function and disability relevant to Veterans' benefits; and to improve surveillance of new symptom clusters and emerging diseases.
Other topics of interest were methodological changes associated with NLP research, and opportunities for future research, such as speech recognition and synthesis, data visualization, and social media discourse. Finally, the attendees also discussed collaborations with academic institutions, which could provide training for post-doctoral students specializing in the field.
More information about the conference, including its agenda, participant list, and information about several presentations and discussions, can be found here.