Webinar: Using Linguamatics for Clinical Support - Johnathan Hartman & Linda Van Keuren
Jonathan Hartman and Linda Van Keuren described how the Linguamatics NLP platform is used to provide real-time support for clinical decision making. Jonathan focused on the use of NLP to text-mine and abstract key information from published sources, and how the information is searched and displayed to clinicians; and Linda discussed how publishers were persuaded to allow text mining of their various publications.
Jonathan accompanies clinicians as they make patient rounds, and has an iPad with which he can immediately search the text-mined information to answer clinicians’ questions about diseases and treatments. Initially GU used Linguamatics I2E OnDemand to quickly retrieve specific relevant information from Medline abstracts during patient rounds. They then added I2E Enterprise to the GU server to text-mine downloaded full-text articles, and developed a web portal that worked on an iPad. This provides a menu of smart query options to quickly address clinical questions and link to relevant literature abstracts and full text articles.
Linda outlined how the GU clinical text mining service has expanded over time and the licensing work needed to make that possible. After text mining Medline abstracts and then full text, GU worked with Elsevier to access a subset of the ScienceDirect journal collection and agreed a suitable licensing agreement and data access via an API that worked with Linguamatics NLP software. Next, they negotiated similar text-mining access to the New England Journal of Medicine and the British Medical Journal collection. And recently they have added PubMed Central open access content.