The AAPM grand auto-segmentation challenge: Deep Learning Vs Atlas

The Thoracic Auto‐Segmentation Challenge was organized at the 2017 Annual Meeting of the American Association of Physicists in Medicine. The purpose of the challenge was to provide a benchmark dataset and platform for evaluating the performance of auto segmentation methods of organs at risk (OARs) in thoracic CT images. Sixty thoracic CT scans provided by three different institutions were separated into 36 training, 12 offline tastings, and 12 online testing scans. The top three participants using Deep Learning produced the best segmentation for all structures, but there was no significant difference in the performance among them. Read more here: 
https://aapm.onlinelibrary.wiley.com/doi/abs/10.1002/mp.13141