Generative adversarial networks: a renewal for data augmentation in lung nodule classification

• Bruno H. L. dos Anjos Universidade Federal de Alagoas
• Anthony E. A. Jatobá Universidade Federal de Alagoas
• Marcelo C. Oliveira Universidade Federal de Alagoas

DOI: https://doi.org/10.5753/wvc.2019.7629

Resumo

Obtaining medical images is an ethically restrictive process and still difficult to validate, depending on well-trained professional, being a laborious and time-consuming activity. Therefore, the construction of large databases of structured medical images is one of the major challenges of the deep learning applications in the computerized aid to the diagnosis in medical images. GAN presents itself as an adequate solution to supply the small number of pathological exams that compose the most diverse medical images banks. In this work we intend to develop a method using GAN to balance the data set of Computed Tomography images and improve the performance of an arbitrary classifier of pulmonary nodules. For this, two GAN architectures with the capacity to generate synthetic images of nodal cuts were trained and in a second moment, a Convolutional Neural Network was trained in rooms of different data set. The training of the different data sets were evaluated by AUC-ROC.

Palavras-chave: Generative Adversarial Networks, lung classification, data augmentation

Referências

P. Azevedo-Marques, Diagnóstico auxiliado por computador na radiologia, Radiologia Brasileira, vol. 34, 10 DOI: 10.1590/s0100-39842001000500008

J. R. Ferreira, M. C. Oliveira, and A. L. Freitas, Performance evaluation of medical image similarity analysis in a heterogeneous architecture, in 2014 IEEE 27th International Symposium on Computer-Based Medical Systems, May 2014, pp. 159 -

K. Doi, Computer-aided diagnosis in medical imaging: Historical review, current status and future potential, Computerized Medical Imaging And Graphics, vol. 31, no. 4-5, pp. 198 - 211, DOI: 10.1016/j.compmedimag.2007.02.002

M. P. McBee, O. A. Awan, A. T. Colucci, C. W. Ghobadi, N. Kadom, A. P. Kansagra, S. Tridandapani, and W. F. Auffermann, Deep learning in radiology, Academic Radiology, vol. 25, no. 11, pp. 1472 - 1480, Nov [Online]. Available: https://doi.org/10.1016/j.acra. 2018. 02.018 DOI: 10.1016/j.acra.2018.02.018

Y. LeCun, Y. Bengio, and G. Hinton, Deep learning, Nature, vol. 521, pp. 436 EP -, May [Online]. Available: https://doi.org/10.1038/nature14539 DOI: 10.1038/nature14539

G. Litjens, T. Kooi, B. E. Bejnordi, A. A. A. Setio, F. Ciompi, M. Ghafoorian, J. A. van der Laak, B. van Ginneken, and C. I. Sa ´nchez, A survey on deep learning in medical image analysis, Medical Image Analysis, vol. 42, pp. 60 - 88, [Online]. Available: http://www.sciencedirect.com/science/article/pii/S1361841517301135

P. Meyer, V. Noblet, C. Mazzara, and A. Lallement, Survey on deep learning for radiotherapy, Computers in Biology and Medicine, vol. 98, 05 DOI: 10.1016/j.compbiomed.2018.05.018

G. Chartrand, P. M. Cheng, E. Vorontsov, M. Drozdzal, S. Turcotte, C. J. Pal, S. Kadoury, and A. Tang, Deep learning: A primer for radiologists, RadioGraphics, vol. 37, no. 7, pp. 2113 - 2131, 2017, pMID: [Online]. Available: https://doi.org/10.1148/rg.2017170077