Qualifying Certainty in Radiology Reports through Deep Learning–Based Natural Language Processing

Abstract

A deep transfer learning model can be developed to reliably assess the level of uncertainty communicated in a radiology report.