Your search term is “Dole”. Do you mean Bob Dole? Dole Pineapple? Or being “on the dole?” How do search engines recognize the meaning of terms and how can that capability be leveraged in medical research and practice? Complex algorithms analyze words and their relationships with other words in text, performing what is called Natural Language Processing (NPL). This is the backbone of many search engines we use everyday such as Google and Bing. Among the pioneers of NPL-based technology is Linguamatics, a company based in the UK. Their flagship product, I2E, searches and mines text in knowledge repositories and documents, extracting new insights and information based on the hidden relationships it uncovers in text and data content.
Text mining is an exciting frontier in medical informatics and as the Georgetown University Medical Center Hospital Informationist, I am proud to be leading the introduction of this new technology into patient care, hospital programs and medical research. Currently I am testing the Linguamatics text mining web tool using my iPad during patient rounds to identify and locate relevant information that assists with clinical decisions.
In October I will be presenting the results of my experience with I2E at the Linguamatics Text Mining Summit conference. As an invited speaker I will be joined by other presenters talking about their experiences using I2E with clinical trials; surgical pathology, radiology, and physician notes; and healthcare analytics which can build predictive models.
I am the only Hospital Informationist in the MedStar family of healthcare facilities who uses a text mining tool on rounds — in fact, I’m the only one in the world to be doing this right now! I am excited to be contributing to the transformation of medical practice and improvement of patient care here at Georgetown.