Seleção de Backbone Para Extração de Características com a U-Net na Segmentação de Patologias Renais
Resumo
Este artigo apresenta uma análise comparativa de diferentes backbones em combinação com a arquitetura U-Net para a segmentação de patologias renais, com ênfase em lesões de esclerose glomerular. O objetivo principal do estudo é demonstrar a viabilidade e a eficácia do uso de backbones pré-treinados nessa tarefa. Foram avaliadas cinco redes neurais convolucionais em um conjunto de 271 imagens. Ao final dos experimentos, a VGG19 destacou-se, apresentando o melhor desempenho, com um coeficiente Dice de 35,88% no conjunto de teste e uma acurácia de 89,84%.
Referências
Barros, G. O., Navarro, B., Duarte, A., and Dos-Santos, W. L. (2017). Pathospotter-k: A computational tool for the automatic identification of glomerular lesions in histological images of kidneys. Scientific reports, 7:46769.
Bel, T., Hermsen, M., van der Laak, J., Litjens, G., Smeets, B., and Hilbrands, L. (2018). Automatic segmentation of histopathological slides of renal tissue using deep learning. page 37.
Chaurasia, A. and Culurciello, E. (2017). Linknet: Exploiting encoder representations for efficient semantic segmentation. In 2017 IEEE Visual Communications and Image Processing (VCIP), pages 1–4.
Ciaparrone, G., Bardozzo, F., Priscoli, M. D., Kallewaard, J. L., Zuluaga, M. R., and Tagliaferri, R. (2020). A comparative analysis of multi-backbone mask r-cnn for surgical tools detection. In 2020 International Joint Conference on Neural Networks (IJCNN), pages 1–8.
Claro, M. d. L. (2022). Classificação de leucemias utilizando aumento de dados, transferência de aprendizado e combinação de cnns. PhD thesis, Universidade Federal do Maranhão.
Costa, D. M. N., Valente, L. M., Gouveia, P. A. C., Sarinho, F. W., Fernandes, G. V., Cavalcante, M. A. G. M., Oliveira, C. B. L., Vasconcelos, C. A. J., and Sarinho, E. S. C. (2017). Comparative analysis of primary and secondary glomerulopathies in the northeast of brazil: data from the pernambuco registry of glomerulopathies – repeg. Brazillian Journal of Nephrology, 39(1):29–35.
Doi, K. (2007). Computer-aided diagnosis in medical imaging: historical review, current status and future potential. Computerized Medical Imaging and Graphics, 31(4-5):198–211.
Ferron, M. and Rancano, J. (2007). Grande Atlas do Corpo Humano. MANOLE, [S.l.].
Gadermayr, M., Dombrowski, A. K., Klinkhammer, B. M., Boor, P., and Merhof, D. (2019). Cnn cascades for segmenting sparse objects in gigapixel whole slide images. Comput. Med. Imaging Graph, 71:40–48.
Govind, D., Ginley, B., Lutnick, B., Tomaszewski, J., and Sarder, P. (2018). Glomerular detection and segmentation from multimodal microscopy images using a butterworth band-pass filter. page 39.
Kaur, G., Garg, M., Gupta, S., Juneja, S., Rashid, J., Gupta, D., Shah, A., and Shaikh, A. (2023). Automatic identification of glomerular in whole-slide images using a modified unet model. Diagnostics, 13:3152.
Long, J., Shelhamer, E., and Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3431–3440.
Ronneberger, O., Fischer, P., and Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical Image Computing and Computer-Assisted Intervention, pages 234–241. Springer.
SBN (2023). Glomerulopatias. [link]. Acesso em: 20 fev. 2024.
Tajbakhsh, N., Shin, J., Gurudu, S., Hurst, R. T., Kendall, C. B., Gotway, M. B., and Liang, J. (2016). Convolutional neural networks for medical image analysis: Fine tuning or full training? IEEE Transactions on Medical Imaging, 35:1299–1312.
Bel, T., Hermsen, M., van der Laak, J., Litjens, G., Smeets, B., and Hilbrands, L. (2018). Automatic segmentation of histopathological slides of renal tissue using deep learning. page 37.
Chaurasia, A. and Culurciello, E. (2017). Linknet: Exploiting encoder representations for efficient semantic segmentation. In 2017 IEEE Visual Communications and Image Processing (VCIP), pages 1–4.
Ciaparrone, G., Bardozzo, F., Priscoli, M. D., Kallewaard, J. L., Zuluaga, M. R., and Tagliaferri, R. (2020). A comparative analysis of multi-backbone mask r-cnn for surgical tools detection. In 2020 International Joint Conference on Neural Networks (IJCNN), pages 1–8.
Claro, M. d. L. (2022). Classificação de leucemias utilizando aumento de dados, transferência de aprendizado e combinação de cnns. PhD thesis, Universidade Federal do Maranhão.
Costa, D. M. N., Valente, L. M., Gouveia, P. A. C., Sarinho, F. W., Fernandes, G. V., Cavalcante, M. A. G. M., Oliveira, C. B. L., Vasconcelos, C. A. J., and Sarinho, E. S. C. (2017). Comparative analysis of primary and secondary glomerulopathies in the northeast of brazil: data from the pernambuco registry of glomerulopathies – repeg. Brazillian Journal of Nephrology, 39(1):29–35.
Doi, K. (2007). Computer-aided diagnosis in medical imaging: historical review, current status and future potential. Computerized Medical Imaging and Graphics, 31(4-5):198–211.
Ferron, M. and Rancano, J. (2007). Grande Atlas do Corpo Humano. MANOLE, [S.l.].
Gadermayr, M., Dombrowski, A. K., Klinkhammer, B. M., Boor, P., and Merhof, D. (2019). Cnn cascades for segmenting sparse objects in gigapixel whole slide images. Comput. Med. Imaging Graph, 71:40–48.
Govind, D., Ginley, B., Lutnick, B., Tomaszewski, J., and Sarder, P. (2018). Glomerular detection and segmentation from multimodal microscopy images using a butterworth band-pass filter. page 39.
Kaur, G., Garg, M., Gupta, S., Juneja, S., Rashid, J., Gupta, D., Shah, A., and Shaikh, A. (2023). Automatic identification of glomerular in whole-slide images using a modified unet model. Diagnostics, 13:3152.
Long, J., Shelhamer, E., and Darrell, T. (2015). Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3431–3440.
Ronneberger, O., Fischer, P., and Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical Image Computing and Computer-Assisted Intervention, pages 234–241. Springer.
SBN (2023). Glomerulopatias. [link]. Acesso em: 20 fev. 2024.
Tajbakhsh, N., Shin, J., Gurudu, S., Hurst, R. T., Kendall, C. B., Gotway, M. B., and Liang, J. (2016). Convolutional neural networks for medical image analysis: Fine tuning or full training? IEEE Transactions on Medical Imaging, 35:1299–1312.
Publicado
11/09/2024
Como Citar
ROCHA, Ana A. F.; BORGES, Rodrigo N.; BATISTA, Rodrigo E. C.; NASCIMENTO, Rhaylson S.; MORICONI, Émery F.; SANTOS, Justino D.; VERAS, Rodrigo M. S..
Seleção de Backbone Para Extração de Características com a U-Net na Segmentação de Patologias Renais. In: ESCOLA REGIONAL DE COMPUTAÇÃO DO CEARÁ, MARANHÃO E PIAUÍ (ERCEMAPI), 12. , 2024, Parnaíba/PI.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2024
.
p. 179-188.
DOI: https://doi.org/10.5753/ercemapi.2024.243753.
