eduzhai > Applied Sciences > Engineering >

Mitosis Detection Under Limited Annotation A Joint Learning Approach

  • Save

... pages left unread,continue reading

Document pages: 6 pages

Abstract: Mitotic counting is a vital prognostic marker of tumor proliferation inbreast cancer. Deep learning-based mitotic detection is on par withpathologists, but it requires large labeled data for training. We propose adeep classification framework for enhancing mitosis detection by leveragingclass label information, via softmax loss, and spatial distribution informationamong samples, via distance metric learning. We also investigate strategiestowards steadily providing informative samples to boost the learning. Theefficacy of the proposed framework is established through evaluation on ICPR2012 and AMIDA 2013 mitotic data. Our framework significantly improves thedetection with small training data and achieves on par or superior performancecompared to state-of-the-art methods for using the entire training data.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...