“LeukNet” - Um Modelo de Rede Neural Convolucional para o Diagnóstico de Leucemia
Resumo
A leucemia é um tipo de câncer que afeta a produção de células sanguíneas na medula óssea o que dificulta a coagulação do sangue e o combate a infecções. Nesse trabalho propomos um método para o diagnóstico automático de leucemia utilizando Redes Neurais Convolucionais (CNNs). Nós utilizamos CNNs pré-treinadas e técnicas de transferência de aprendizagem na construção do método proposto. Empregamos a técnica modified Deeply Fine Tuning (mDFT) combinada com operações de aumento de dados para refinar um modelo pré-treinado. Para treinar e testar o método proposto, utilizamos um conjunto de 3.536 imagens de 18 bases diferentes. A técnica de validação Leave-One-Dataset-Out Cross-Validation (LODOCV) foi proposta para avaliar a capacidade de generalização do modelo. Os principais resultados obtidos utilizando o LODOCV em três bases de dados foram 97,04, 82,46 e 70,24% de acurácia.Referências
Chollet, F. (2017). Xception: Deep learning with depthwise separable conIn 2017 IEEE Conference on Computer Vision and Pattern Recognition, volutions. CVPR 2017, Honolulu, HI, USA, July 21-26, 2017, pages 1800–1807.
He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning In 2016 IEEE Conference on Computer Vision and Pattern for image recognition. Recognition (CVPR), pages 770–778. IEEE Computer Society.
Khosravan, N., Celik, H., Turkbey, B., Jones, E. C., Wood, B., and Bagci, U. (2019). A collaborative computer aided diagnosis (c-cad) system with eye-tracking, sparse attentional model, and deep learning. Medical Image Analysis, 51:101–115.
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., Berg, A. C., and Fei-Fei, L. (2015). ImageNet Large Scale Visual Recognition Challenge. International Journal of Computer Vision (IJCV), 115(3):211–252.
Shanthi, T. and Sabeenian, R. (2019). Modied alexnet architecture for classication of diabetic retinopathy images. Computers and Electrical Engineering, 76:56–64.
Simonyan, K. and Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. CoRR, abs/1409.1556.
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., and Wojna, Z. In 2016 IEEE (2016). Rethinking the inception architecture for computer vision. Conference on Computer Vision and Pattern Recognition CVPR 2016, Las Vegas, NV, USA, June 27-30, 2016, pages 2818–2826.
Vogado, L. H. S., Veras, R. M. S., Araújo, F. H. D., e Silva, R. R. V., and Aires, K. R. T. (2018). Leukemia diagnosis in blood slides using transfer learning in cnns and SVM for classication. Engineering Applications of Articial Intelligence, 72:415–422.
Yosinski, J., Clune, J., Bengio, Y., and Lipson, H. (2014). How transferable are features in deep neural networks? In Proceedings of the 27th International Conference on Neural Information Processing Systems Volume 2, NIPS’14, pages 3320–3328, Cambridge, MA, USA.
He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning In 2016 IEEE Conference on Computer Vision and Pattern for image recognition. Recognition (CVPR), pages 770–778. IEEE Computer Society.
Khosravan, N., Celik, H., Turkbey, B., Jones, E. C., Wood, B., and Bagci, U. (2019). A collaborative computer aided diagnosis (c-cad) system with eye-tracking, sparse attentional model, and deep learning. Medical Image Analysis, 51:101–115.
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., Berg, A. C., and Fei-Fei, L. (2015). ImageNet Large Scale Visual Recognition Challenge. International Journal of Computer Vision (IJCV), 115(3):211–252.
Shanthi, T. and Sabeenian, R. (2019). Modied alexnet architecture for classication of diabetic retinopathy images. Computers and Electrical Engineering, 76:56–64.
Simonyan, K. and Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. CoRR, abs/1409.1556.
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., and Wojna, Z. In 2016 IEEE (2016). Rethinking the inception architecture for computer vision. Conference on Computer Vision and Pattern Recognition CVPR 2016, Las Vegas, NV, USA, June 27-30, 2016, pages 2818–2826.
Vogado, L. H. S., Veras, R. M. S., Araújo, F. H. D., e Silva, R. R. V., and Aires, K. R. T. (2018). Leukemia diagnosis in blood slides using transfer learning in cnns and SVM for classication. Engineering Applications of Articial Intelligence, 72:415–422.
Yosinski, J., Clune, J., Bengio, Y., and Lipson, H. (2014). How transferable are features in deep neural networks? In Proceedings of the 27th International Conference on Neural Information Processing Systems Volume 2, NIPS’14, pages 3320–3328, Cambridge, MA, USA.
Publicado
15/06/2021
Como Citar
VOGADO, Luis H. S.; VERAS, Rodrigo M. S.; AIRES, Kelson R. T..
“LeukNet” - Um Modelo de Rede Neural Convolucional para o Diagnóstico de Leucemia. In: CONCURSO DE TESES E DISSERTAÇÕES - SIMPÓSIO BRASILEIRO DE COMPUTAÇÃO APLICADA À SAÚDE (SBCAS), 21. , 2021, Evento Online.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2021
.
p. 85-90.
ISSN 2763-8987.
DOI: https://doi.org/10.5753/sbcas.2021.16106.
