Scaling of Texture in Training Autoencoders for Classification of Histological Images of Colorectal Cancer
Autoencoding in deep learning has been known as a useful tool for extracting image features in multiple layers, which are subsequently configured for classification by deep neural networks. A practical burden for the implementation of autoencoders is the time required for training a large number of artificial neurons. This paper shows the effects of scaling of texture in the histology of colorectal cancer, which can result in significant training time reduction being approximately to an exponential function, with improved classification rates.
KeywordsDeep neural networks Image classification Digital pathology Colorectal cancer Tissue types
- 4.The Cancer Genome Atlas Research Network: Comprehensive molecular characterization of human colon and rectal cancer. Nature 487, 330–337 (2012)Google Scholar
- 5.Aerts, H.J., et al.: Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat. Commun. 5, Article no. 4006 (2014)Google Scholar
- 9.Goodfellow, I., Bengio, Y., Courville, A.: Deep Learning. MIT Press (2016, in preparation). http://www.deeplearningbook.org
- 10.Pham, T.D.: Enhancing texture characteristics with synthesis and noise for image retrieval. In: Proceedings of the 8th IEEE International Conference on Intelligent Systems, pp. 433–437. IEEE Press, New York (2016)Google Scholar
- 14.Nielsen, M.A.: Neural Networks and Deep Learning. Determination Press (2015). http://neuralnetworksanddeeplearning.com