Training deep learning networks requires large amounts of accurate labels which are usually extracted from reports using natural language processing or by time-consuming manual review. This study aimed to develop and evaluate a workflow for using data from structured reports as labels that can be used in a deep learning application.
Authors: Daniel Pinto dos Santos, Sebastian Brodehl, Bettina Baeßler, Gordon Arnhold, Thomas Dratsch, Seung-Hun Chon, Peter Mildenberger and Florian Jungmann