Proof of subjective analysis of a Deep-Learning model for autocontouring in clinical practice

Deep-learning contouring (DLC) has emerged as a possible framework for the next generation of autocontouring tools. The overall rate of clinical acceptability for manually delineated contours was 74% as compared to DLC contours which were rated as clinically acceptable 89% of the time. The aim of this study was to explore subjective impressions of the performance of parotid OAR autocontouring using DLC compared to previously defined manual segmentations. Read more here: