Determining Assertion Status for Medical Problems in Clinical Records

This paper describes the MITRE system entries for the 2010 i2b2/VA community evaluation “Challenges in Natural Language Processing for Clinical Data” for the task of classifying assertions associated with problem concepts extracted from patient records. Our best performing system obtained an overall micro-averaged F-score of 0.9343. The methods employed were a combination of machine learning (Conditional Random Field and Maximum Entropy) and rule-based (pattern matching) techniques.