There is large consent that successful training of deep networks requires many thousand annotated training samples. In this paper, we present a network and training strategy that relies on the strong use of data augmentation to use the available annotated samples more efficiently. The architecture consists of a contracting path to capture context and a symmetric expanding path that enables precise localization. We show that such a network can be trained end-to-end from very few images and outperforms the prior best method (a sliding-window convolutional network) on the ISBI challenge for segmentation of neuronal structures in electron microscopic stacks. Using the same network trained on transmitted light microscopy images (phase contrast and DIC) we won the ISBI cell tracking challenge 2015 in these categories by a large margin. Moreover, the network is fast. Segmentation of a 512x512 image takes less than a second on a recent GPU. The full implementation (based on Caffe) and the trained networks are available at http://lmb.informatik.uni-freiburg.de/people/ronneber/u-net.

scihub/10.1007/978-3-319-24574-4_28.pdf

Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, ... III (Lecture Notes in Computer Science, 9351)

Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015 : 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III

International Conference on Medical Image Computing and Computer-Assisted Intervention

Nassir Navab; Joachim Hornegger; William M Wells; Alejandro F Frangi

Alejandro Frangi, Joachim Hornegger, Nassir Navab, William M. Wells

Author

Springer International Publishing AG

Springer Nature Switzerland AG

Springer London, Limited

Image Processing, Computer Vision, Pattern Recognition, and Graphics, 9351, Cham, 2015

Lecture notes in computer science, 9349-9351, Cham, 2015

LNCS sublibrary, 1st ed. 2015, Cham, 2015

Springer Nature, Cham, 2015

Switzerland, Switzerland

1, 20150928

{"container_title":"Lecture Notes in Computer Science","first_page":234,"issns":["0302-9743","1611-3349"],"last_page":241,"parent_isbns":["9783319245737","9783319245744"]}

Referenced by: doi:10.1109/cvpr.2015.7298761 doi:10.1162/neco.1989.1.4.541 doi:10.1109/cvpr.2015.7298965 doi:10.1109/iccv.2015.123 doi:10.1145/2647868.2654889 doi:10.1371/journal.pbio.1000502 doi:10.1093/bioinformatics/btu080 doi:10.1109/cvpr.2015.7298642 doi:10.1109/iccv.2013.269

The three-volume set LNCS 9349, 9350, and 9351 constitutes the refereed proceedings of the 18th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015, held in Munich, Germany, in October 2015. Based on rigorous peer reviews, the program committee carefully selected 263 revised papers from 810 submissions for presentation in three volumes. The papers have been organized in the following topical sections: quantitative image analysis I: segmentation and measurement; computer-aided diagnosis: machine learning; computer-aided diagnosis: automation; quantitative image analysis II: classification, detection, features, and morphology; advanced MRI: diffusion, fMRI, DCE; quantitative image analysis III: motion, deformation, development and degeneration; quantitative image analysis IV: microscopy, fluorescence and histological imagery; registration: method and advanced applications; reconstruction, image formation, advanced acquisition - computational imaging; modelling and simulation for diagnosis and interventional planning; computer-assisted and image-guided interventions.