Ongoing Work on Deep Learning for Lung Cancer Prediction
Deep learning is one of the breakthrough technologies that have emergent in the last few years. It has been applied to a wide variety of problems, most of them related with image processing. It is also being considered for 3D data in medical image processing. This paper is a report of ongoing work about the development of deep learning architectures for lung cancer prediction. Data has been extracted from an ongoing Kaggel challenge, involving multi-center CTA data. First we have normalized in intensity the images. Then we have devised an auto encoder architecture with convolutional layers to obtain a compressed representation of the lung images. These representations are fed as features to a random forest classifier.
KeywordsGraphic Processing Unit Deep Learning Convolutional Neural Network Minority Class Imbalanced Dataset
- 1.Ciresan, D.C., Giusti, A., Gambardella, L.M., Schmidhuber, J.: Deep neural networks segment neuronal membranes in electron microscopy images. In: Advances in neural information processing systems, pp. 2843–2851 (2012)Google Scholar
- 2.Donahue, J., Jia, Y., Vinyals, O., Hoffman, J., Zhang, N., Tzeng, E., Darrell, T.: Decaf: a deep convolutional activation feature for generic visual recognition. CoRR, abs/1310.1531 (2013)Google Scholar
- 4.Oquab, M., Bottou, L., Laptev, I., Sivic, J.: Learning and transferring mid-level image representations using convolutional neural networks. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2014, pp. 1717–1724. IEEE Computer Society Washington, DC, USA (2014)Google Scholar
- 6.Vincent, P., Larochelle, H., Bengio, Y., Manzagol, P.-A.: Extracting and composing robust features with denoising autoencoders. In: Proceedings of the 25th International Conference on Machine Learning, ICML 2008, pp. 1096–1103, New York, NY, USA, ACM (2008)Google Scholar