Project Description

Clinical Text Annotator: Heart attack risk factor identification from clinical text

i2B2 is an NIH-funded National Center for Biomedical Computing. It has periodically staged NLP challenges to extract fine -grained information from clinical records. One such challenge was to identify the information that is medically relevant to identifying heart disease risk over sets of longitudinal patient records. The relevant information may include high blood pressure and cholesterol levels, obesity, smoking status, etc. In this demonstration, we show how our NLP and AI capabilities can identify and extract the necessary information from free text clinical record.

Key Features

  • Parse and extract medical concepts and entities from free text clinical records
  • Map the medical concepts and entities to relevant risk factors
  • Identify medications, measurements, and test results from clinical records
  • Apply heuristics to infer risk factors from medications, measurements, and test results.
  • Apply negation to correctly identify whether a risk-factor is present.