Capsule Networks Against Medical Imaging Data Challenges
A key component to the success of deep learning is the availability of massive amounts of training data. Building and annotating large datasets for solving medical image classification problems is today a bottleneck for many applications. Recently, capsule networks were proposed to deal with shortcomings of Convolutional Neural Networks (ConvNets). In this work, we compare the behavior of capsule networks against ConvNets under typical datasets constraints of medical image analysis, namely, small amounts of annotated data and class-imbalance. We evaluate our experiments on MNIST, Fashion-MNIST and medical (histological and retina images) publicly available datasets. Our results suggest that capsule networks can be trained with less amount of data for the same or better performance and are more robust to an imbalanced class distribution, which makes our approach very promising for the medical imaging community.
KeywordsCapsule networks Small datasets Class imbalance
This work has received funding from the European Unions Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No. 713673. Amelia Jiménez-Sánchez has received financial support through the “la Caixa” INPhINIT Fellowship Grant for Doctoral studies at Spanish Research Centres of Excellence, “la Caixa” Banking Foundation, Barcelona, Spain. The authors would like to thank Nvidia for the GPU donation and Aurélien Geron for his tutorial and code on Capsule Networks.
- 1.MICCAI Grand Challenge Tumor Proliferation Assessment Challenge (TUPAC16). http://tupac.tue-image.nl/. Accessed 18 Jan 2018
- 3.Cardoso, M.J., et al. (eds.): Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis: 6th Joint International Workshops, CVII-STENT and Second International Workshop, LABELS (2017), held in Conjunction with MICCAI 2017 (2017)Google Scholar
- 4.Glorot, X., Bengio, Y.: Understanding the difficulty of training deep feedforward neural networks. In: International Conference on Artificial Intelligence and Statistics, vol. 9, pp. 249–256. PMLR, 13–15 May 2010Google Scholar
- 5.Kalesnykiene, V., Kamarainen, J.-K., Voutilainen, R., Pietil, J., Kälviäinen, H., Uusitalo, H.: Diaretdb1 diabetic retinopathy database and evaluation protocolGoogle Scholar
- 6.Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. CoRR abs/1412.6980 (2014). http://arxiv.org/abs/1412.6980
- 8.LeCun, Y., Cortes, C.: MNIST handwritten digit database (2010). http://yann.lecun.com/exdb/mnist/
- 9.Litjens, G.J.S., et al.: A survey on deep learning in medical image analysis. CoRR abs/1702.05747 (2017). http://arxiv.org/abs/1702.05747
- 10.Sabour, S., Frosst, N., Hinton, G.E.: Dynamic routing between capsules. In: Guyon, I., et al. (eds.) Advances in Neural Information Processing Systems, vol. 30, pp. 3856–3866. Curran Associates, Inc. (2017). http://papers.nips.cc/paper/6975-dynamic-routing-between-capsules.pdf
- 12.Vasconcelos, C.N., Vasconcelos, B.N.: Increasing deep learning melanoma classification by classical and expert knowledge based image transforms. CoRR abs/1702.07025 (2017). http://arxiv.org/abs/1702.07025
- 13.Xiao, H., Rasul, K., Vollgraf, R.: Fashion-MNIST: a novel image dataset for benchmarking machine learning algorithms (2017)Google Scholar
- 14.Zhou, J., Li, Z., Zhi, W., Liang, B., Moses, D., Dawes, L.: Using convolutional neural networks and transfer learning for bone age classification. In: International Conference on Digital Image Computing: Techniques and Applications (DICTA), pp. 1–6 (2017)Google Scholar