A large annotated medical image dataset for the development and evaluation of segmentation algorithms

• DOI: 10.48550/arxiv.1902.09063
• Publication Date: 2019-01-01
• Authors: Amber L. Simpson; Michela Antonelli; Spyridon Bakas; Michel Bilello; Keyvan Farahani; Bram van Ginneken; Annette Kopp‐Schn...; Bennett A. Landman; Geert Litjens; Bjoern Menze; Olaf Ronneberger; Ronald M. Summers; Patrick Bilic; Patrick Ferdinand...; Richard Kinh Gian; Marc J. Gollub; Jennifer Golia-Pe...; Stephan Heckers; William R. Jarnagin; Maureen McHugo; Sandy Napel; Eugene Vorontsov; Lena Maier‐Hein; M. Jorge Cardoso
• Keywords: Benchmarking; Initialization; Segmentation-based object categorization

Semantic segmentation of medical images aims to associate a pixel with label in image without human initialization. The success semantic algorithms is contingent on the availability high-quality imaging data corresponding labels provided by experts. We sought create large collection annotated datasets various clinically relevant anatomies available under open source license facilitate development algorithms. Such resource would allow: 1) objective assessment general-purpose methods through comprehensive benchmarking and 2) free access for any researcher interested problem domain. Through multi-institutional effort, we generated large, curated dataset representative several highly variable tasks that was used crowd-sourced challenge - Medical Segmentation Decathlon held during 2018 Image Computing Computer Aided Interventions Conference Granada, Spain. Here, describe these ten labeled so may be effectively reused research community.