Mitosis domain generalization in histopathology images - The MIDOG challenge
Aubreville, Marc; Stathonikos, Nikolas; Bertram, Christof A; Klopfleisch, Robert; Ter Hoeve, Natalie; Ciompi, Francesco; Wilm, Frauke; Marzahl, Christian; Donovan, Taryn A; Maier, Andreas; Breen, Jack; Ravikumar, Nishant; Chung, Youjin; Park, Jinah; Nateghi, Ramin; Pourakpour, Fattaneh; Fick, Rutger H J; Ben Hadj, Saima; Jahanifar, Mostafa; Shephard, Adam; Dexl, Jakob; Wittenberg, Thomas; Kondo, Satoshi; Lafarge, Maxime W; Koelzer, Viktor H; Liang, Jingtang; Wang, Yubo; Long, Xi; Liu, Jingxin; Razavi, Salar; Khademi, April; Yang, Sen; Wang, Xiyue; Erber, Ramona; Klang, Andrea; Lipnik, Karoline; Bolfa, Pompei; Dark, Michael J; Wasinger, Gabriel; Veta, Mitko; Breininger, Katharina
(2023) Medical Image Analysis, volume 84
(Article)
Abstract
The density of mitotic figures (MF) within tumor tissue is known to be highly correlated with tumor proliferation and thus is an important marker in tumor grading. Recognition of MF by pathologists is subject to a strong inter-rater bias, limiting its prognostic value. State-of-the-art deep learning methods can support experts
... read more
but have been observed to strongly deteriorate when applied in a different clinical environment. The variability caused by using different whole slide scanners has been identified as one decisive component in the underlying domain shift. The goal of the MICCAI MIDOG 2021 challenge was the creation of scanner-agnostic MF detection algorithms. The challenge used a training set of 200 cases, split across four scanning systems. As test set, an additional 100 cases split across four scanning systems, including two previously unseen scanners, were provided. In this paper, we evaluate and compare the approaches that were submitted to the challenge and identify methodological factors contributing to better performance. The winning algorithm yielded an F 1 score of 0.748 (CI95: 0.704-0.781), exceeding the performance of six experts on the same task.
show less
Download/Full Text
The full text of this publication is not available.
Keywords: Challenge, Deep Learning, Domain generalization, Histopathology, Mitosis, Radiological and Ultrasound Technology, Health Informatics, Radiology Nuclear Medicine and imaging, Computer Vision and Pattern Recognition, Computer Graphics and Computer-Aided Design
ISSN: 1361-8415
Publisher: Elsevier
Note: Publisher Copyright: © 2022 Elsevier B.V.
(Peer reviewed)