Abstract
This paper presents and evaluates a deep learning architecture for automated basal cell carcinoma cancer detection that integrates (1) image representation learning, (2) image classification and (3) result interpretability. A novel characteristic of this approach is that it extends the deep learning architecture to also include an interpretable layer that highlights the visual patterns that contribute to discriminate between cancerous and normal tissues patterns, working akin to a digital staining which spotlights image regions important for diagnostic decisions. Experimental evaluation was performed on set of 1,417 images from 308 regions of interest of skin histopathology slides, where the presence of absence of basal cell carcinoma needs to be determined. Different image representation strategies, including bag of features (BOF), canonical (discrete cosine transform (DCT) and Haar-based wavelet transform (Haar)) and proposed learned-from-data representations, were evaluated for comparison. Experimental results show that the representation learned from a large histology image data set has the best overall performance (89.4% in F-measure and 91.4% in balanced accuracy), which represents an improvement of around 7% over canonical representations and 3% over the best equivalent BOF representation.
Chapter PDF
Similar content being viewed by others
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
References
Díaz, G., Romero, E.: Micro-structural tissue analysis for automatic histopathological image annotation. Microsc. Res. Tech. 75(3), 343–358 (2012)
Krizhevsky, A., et al.: Imagenet classification with deep convolutional neural networks. In: NIPS, pp. 1106–1114 (2012)
Cruz-Roa, A., et al.: Automatic Annotation of Histopathological Images Using a Latent Topic Model Based On Non-negative Matrix Factorization. J. Pathol. Inform. 2(1), 4 (2011)
Cruz-Roa, A., et al.: Visual pattern mining in histology image collections using bag of features. Artif. Intell. Med. 52(2), 91–106 (2011)
Cruz-Roa, A., González, F., Galaro, J., Judkins, A.R., Ellison, D., Baccon, J., Madabhushi, A., Romero, E.: A visual latent semantic approach for automatic analysis and interpretation of anaplastic medulloblastoma virtual slides. In: Ayache, N., Delingette, H., Golland, P., Mori, K. (eds.) MICCAI 2012, Part I. LNCS, vol. 7510, pp. 157–164. Springer, Heidelberg (2012)
Pang, B., et al.: Cell nucleus segmentation in color histopathological imagery using convolutional networks. In: CCPR, pp. 1–5. IEEE (2010)
He, L., et al.: Histology image analysis for carcinoma detection and grading. Comput. Meth. Prog. Bio. (2012)
Le, Q.V., et al.: Building high-level features using large scale unsupervised learning. In: ICML (2011)
Raina, R., et al.: Self-taught learning: transfer learning from unlabeled data. In: ICML 2007, pp. 759–766 (2007)
Bengio, Y., et al.: Representation learning: A review and new perspectives. Arxiv (2012)
Montavon, G.: A machine learning approach to classification of low resolution histological samples. Master’s thesis (2009)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Cruz-Roa, A.A., Arevalo Ovalle, J.E., Madabhushi, A., González Osorio, F.A. (2013). A Deep Learning Architecture for Image Representation, Visual Interpretability and Automated Basal-Cell Carcinoma Cancer Detection. In: Mori, K., Sakuma, I., Sato, Y., Barillot, C., Navab, N. (eds) Medical Image Computing and Computer-Assisted Intervention – MICCAI 2013. MICCAI 2013. Lecture Notes in Computer Science, vol 8150. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-40763-5_50
Download citation
DOI: https://doi.org/10.1007/978-3-642-40763-5_50
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-40762-8
Online ISBN: 978-3-642-40763-5
eBook Packages: Computer ScienceComputer Science (R0)