Abstract
The benefits of deep neural networks can be hard to realise in medical imaging tasks because training sample sizes are often modest. Pre-training on large data sets and subsequent transfer learning to specific tasks with limited labelled training data has proved a successful strategy in other domains. Here, we implement and test this idea for detecting and classifying nuclei in histology, important tasks that enable quantifiable characterisation of prostate cancer. We pre-train a convolutional neural network for nucleus detection on a large colon histology dataset, and examine the effects of fine-tuning this network with different amounts of prostate histology data. Results show promise for clinical translation. However, we find that transfer learning is not always a viable option when training deep neural networks for nucleus classification. As such, we also demonstrate that semi-supervised ladder networks are a suitable alternative for learning a nucleus classifier with limited data.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
It takes approximately 80 min to process a \(107~250\times 103~168\) whole-slide prostatectomy image on an NVIDIA GTX980 (incl. disk I/O), comparable to [16].
- 2.
The supervised objective is the cross entropy cost at the top of the encoder while the unsupervised objectives are the denoising mean squared errors at each decoder layer. We refer the reader to [11] for a more detailed description of ladder networks.
- 3.
We note that since the FCN performs dense prediction on an input image, training/testing with a single \(250\times 250\)px image patch is equivalent to training/testing with 62, 500 neighbouring patches using a patch-based method.
References
Arteta, C., Lempitsky, V., Noble, J.A., Zisserman, A.: Learning to detect cells using non-overlapping extremal regions. In: Ayache, N., Delingette, H., Golland, P., Mori, K. (eds.) MICCAI 2012. LNCS, vol. 7510, pp. 348–356. Springer, Heidelberg (2012). doi:10.1007/978-3-642-33415-3_43
Bayramoglu, N., Heikkilä, J.: Transfer learning for cell nuclei classification in histopathology images. In: Hua, G., Jégou, H. (eds.) ECCV 2016. LNCS, vol. 9915, pp. 532–539. Springer, Cham (2016). doi:10.1007/978-3-319-49409-8_46
Chatterjee, A., Watson, G., Myint, E., Sved, P., McEntee, M., Bourne, R.M.: Changes in epithelium, stroma, and lumen space correlate more strongly with gleason pattern and are stronger predictors of prostate ADC changes than cellularity metrics. Radiology 277(3), 751–762 (2015)
Gal, Y., Islam, R., Ghahramani, Z.: Deep bayesian active learning with image data. In: NIPS Bayesian Deep Learning Workshop (2016)
Gorelick, L., Veksler, O., Gaed, M., Gómez, J.A., Moussa, M., Bauman, G.S., Fenster, A., Ward, A.D.: Prostate histopathology: learning tissue component histograms for cancer detection and classification. IEEE Trans. Med. Imaging 32(10), 1804–1818 (2013)
Kainz, P., Urschler, M., Schulter, S., Wohlhart, P., Lepetit, V.: You should use regression to detect cells. In: Navab et al. [9], pp. 276–283
Lee, G., Ali, S., Veltri, R., Epstein, J.I., Christudass, C., Madabhushi, A.: Cell orientation entropy (COrE): predicting biochemical recurrence from prostate cancer tissue microarrays. In: Mori, K., Sakuma, I., Sato, Y., Barillot, C., Navab, N. (eds.) MICCAI 2013. LNCS, vol. 8151, pp. 396–403. Springer, Heidelberg (2013). doi:10.1007/978-3-642-40760-4_50
Maaløe, L., Sønderby, C.K., Sønderby, S.K., Winther, O.: Auxiliary deep generative models. In: Balcan, M., Weinberger, K.Q. (eds.) ICML 2016, Proceedings of Machine Learning Research, PMLR, pp. 1445–1453 (2016)
Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.): MICCAI 2015. LNCS, vol. 9351. Springer, Cham (2015). doi:10.1007/978-3-319-24553-9
Ourselin, S., Joskowicz, L., Sabuncu, M.R., Unal, G., Wells, W. (eds.): MICCAI 2016. LNCS, vol. 9901. Springer, Cham (2016)
Rasmus, A., Berglund, M., Honkala, M., Valpola, H., Raiko, T.: Semi-supervised learning with ladder networks. In: Cortes, C., Lawrence, N.D., Lee, D.D., Sugiyama, M., Garnett, R. (eds.) NIPS 2015, pp. 3546–3554. Curran Associates, Inc. (2015)
Sirinukunwattana, K., Raza, S.E.A., Tsang, Y.W., Snead, D.R.J., Cree, I.A., Rajpoot, N.M.: Locality sensitive deep learning for detection and classification of nuclei in routing colon cancer histology images. IEEE Trans. Med. Imaging 35(5), 1196–1206 (2016)
Sutskever, I., Martens, J., Dahl, G., Hinton, G.: On the importance of initialization and momentum in deep learning. In: Dasgupta, S., McAllester, D. (eds.) ICML 2013, Proceedings of Machine Learning Research, PMLR, pp. 1139–1147 (2013)
Wang, S., Yao, J., Xu, Z., Huang, J.: Subtype cell detection with an accelerated deep convolution neural network. In: Ourselin et al. [10], pp. 640–648
Xie, Y., Xing, F., Kong, X., Su, H., Yang, L.: Beyond classification: structured regression for robust cell detection using convolutional neural network. In: Navab et al. [9], pp. 358–365
Xu, Z., Huang, J.: Detecting 10,000 cells in one second. In: Ourselin et al. [10], pp. 676–684
Acknowledgements
We thank the EPSRC for funding EP’s (EP/N021967/1), DA’s (EP/M020533) and GB’s (EP/K015664/1, EP/K503745/1) work on this topic, the UCL Department of Computer Science for JJ’s studentship and the UCL Computer Science Cluster team.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Jacobs, J.G., Brostow, G.J., Freeman, A., Alexander, D.C., Panagiotaki, E. (2017). Detecting and Classifying Nuclei on a Budget. In: Cardoso, M., et al. Intravascular Imaging and Computer Assisted Stenting, and Large-Scale Annotation of Biomedical Data and Expert Label Synthesis. LABELS STENT CVII 2017 2017 2017. Lecture Notes in Computer Science(), vol 10552. Springer, Cham. https://doi.org/10.1007/978-3-319-67534-3_9
Download citation
DOI: https://doi.org/10.1007/978-3-319-67534-3_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-67533-6
Online ISBN: 978-3-319-67534-3
eBook Packages: Computer ScienceComputer Science (R0)