Building a genomic and clinical NER application

In Chapter 4, you saw how you can extract clinical entities from audio transcripts of a patient visit summary using Comprehend Medical. Sometimes, we may have the need to extend the detection of named entities beyond the set of clinical entities that Comprehend Medical detects out of the box. For example, genomic entities can be detected from genomic testing reports and be combined with clinical entities from Comprehend Medical to get a better understanding of both categories of entities in the report, resulting in better interpretation. This also helps automate the clinical genomic reporting pipeline by automating the process of extracting meaningful information hidden in the testing report. To better understand the application, look at the following diagram.

Figure 7.1 – Workflow for the NER application

The purpose of the application is to create an automated workflow to extract genetic and clinical entities...