Abstract
U-Net architectures are an extremely powerful tool for segmenting 3D volumes, and the recently proposed multi-planar U-Net has reduced the computational requirement for using the U-Net architecture on three-dimensional isotropic data to a subset of two-dimensional planes. While multi-planar sampling considerably reduces the amount of training data needed, providing the required manually annotated data can still be a daunting task. In this article, we investigate the multi-planar U-Net’s ability to learn three-dimensional structures in isotropic sampled images from sparsely annotated training samples. We extend the multi-planar U-Net with random annotations, and we present our empirical findings on two public domains, fully annotated by an expert. Surprisingly we find that the multi-planar U-Net on average outperforms the 3D U-Net in most cases in terms of dice, sensitivity, and specificity and that similar performance from the multi-planar unit can be obtained from half the number of annotations by doubling the number of automatically generated training planes. Thus, sometimes less is more!
Keywords
- 3D imaging
- Segmentation
- Deep learning
- U-Net
- Sparse annotations
This is a preview of subscription content, access via your institution.
Buying options
References
Ioffe, S., Szegedy, C.: Batch normalization: accelerating deep network training by reducing internal covariate shift. In: International Conference on Machine Learning (ICML), pp. 448–456. PMLR (2015)
Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. In: International Conference on Learning Representations (ICLR) (2015)
Lucchi, A., Li, Y., Fua, P.: Learning for structured prediction using approximate subgradient descent with working sets. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (2013). https://doi.org/10.1109/CVPR.2013.259
Odena, A., Dumoulin, V., Olah, C.: Deconvolution and checkerboard artifacts. Distill (2016). https://doi.org/10.23915/distill.00003. http://distill.pub/2016/deconv-checkerboard
Perslev, M., Dam, E.B., Pai, A., Igel, C.: One network to segment them all: a general, lightweight system for accurate 3D medical image segmentation. In: Shen, D., et al. (eds.) MICCAI 2019. LNCS, vol. 11765, pp. 30–38. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-32245-8_4
Ronneberger, O., Fischer, P., Brox, T.: U-net: convolutional networks for biomedical image segmentation. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015. LNCS, vol. 9351, pp. 234–241. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-24574-4_28
Simpson, A.L., et al.: A large annotated medical image dataset for the development and evaluation of segmentation algorithms, February 2019
Çiçek, Ö., Abdulkadir, A., Lienkamp, S.S., Brox, T., Ronneberger, O.: 3D U-net: learning dense volumetric segmentation from sparse annotation. In: Ourselin, S., Joskowicz, L., Sabuncu, M.R., Unal, G., Wells, W. (eds.) MICCAI 2016. LNCS, vol. 9901, pp. 424–432. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-46723-8_49
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Laprade, W.M., Perslev, M., Sporring, J. (2021). How Few Annotations are Needed for Segmentation Using a Multi-planar U-Net?. In: , et al. Deep Generative Models, and Data Augmentation, Labelling, and Imperfections. DGM4MICCAI DALI 2021 2021. Lecture Notes in Computer Science(), vol 13003. Springer, Cham. https://doi.org/10.1007/978-3-030-88210-5_20
Download citation
DOI: https://doi.org/10.1007/978-3-030-88210-5_20
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-88209-9
Online ISBN: 978-3-030-88210-5
eBook Packages: Computer ScienceComputer Science (R0)
-
Published in cooperation with
http://miccai.org/
