期刊
MEDICAL IMAGE ANALYSIS
卷 84, 期 -, 页码 -出版社
ELSEVIER
DOI: 10.1016/j.media.2022.102699
关键词
Domain generalization; Histopathology; Challenge; Deep Learning; Mitosis
The density of mitotic figures (MF) in tumor tissue is an important marker for tumor grading, but its recognition by pathologists is biased and limited. Deep learning methods can support the recognition, but their performance deteriorates in different clinical environments due to variability caused by using different whole slide scanners. The MICCAI MIDOG 2021 challenge aimed to develop scanner-agnostic MF detection algorithms and the winning algorithm outperformed six experts on the same task.
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 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 1 score of 0.748 (CI95: 0.704-0.781), exceeding the performance of six experts on the same task.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据