Introduction to Clinical Natural Language Processing with Python

Background: Many of the most valuable insights in medicine are contained in written patient records. While some of these are coded into structured data as part of the record entry, many exist only as text. Although a complete understanding of this text is beyond current technology, a surprising amount of insight can be gained from relatively simple natural language processing. Learning objectives: This chapter introduces the basics of text processing with Python, such as name-entity recognition, regular expressions, text tokenization and negation detection. By working through the four structured NLP tutorials in this chapter, the reader will learn these NLP techniques to extract valuable clinical insights from text. Limitations: The field of Natural Language Processing is as broad and varied as human communication. The techniques we will discuss in this chapter are but a sampling of what the field has to offer. That said, we will provide enough basic techniques to allow the reader to start to unlock the potential of textual clinical notes.