Uma Abordagem Sistemática para Seleção de Encoders aplicada à Segmentação de Tumores com TransUNet
Resumo
A escolha do encoder em arquiteturas híbridas de segmentação impacta diretamente o desempenho do modelo. Este estudo avalia cinco encoders integrados à uma TransUNet para segmentação de tumores ósseos, além de compará-la com arquiteturas de referência. Os resultados mostram diferenças relevantes entre os modelos, com melhor desempenho global para o ConvNeXt-Tiny, maior precisão para a ResNet34 e maior sensibilidade para a EfficientNet-B0. A análise qualitativa evidencia padrões distintos de segmentação, fornecendo diretrizes práticas para a escolha do encoder conforme o objetivo da aplicação.Referências
Bădilă, A. E., Rădulescu, D. M., Niculescu, A. G., Grumezescu, A. M., Rădulescu, M., and Rădulescu, A. R. (2021). Recent advances in the treatment of bone metastases and primary bone tumors: an up-to-date review. Cancers, 13(16):4229.
Bloier, M., Hinterwimmer, F., Breden, S., Consalvo, S., Neumann, J., Wilhelm, N., and Burgkart, R. (2022). Detection and segmentation of heterogeneous bone tumours in limited radiographs. Current Directions in Biomedical Engineering, 8(2):69–72.
Das, S. and Majumder, S. (2020). Lung cancer detection using deep learning network: A comparative analysis. In Proceedings of the Fifth International Conference on Research in Computational Intelligence and Communication Networks (ICRCICN), pages 30–35. IEEE.
He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778.
Huang, G., Liu, Z., Maaten, L. V. D., and Weinberger, K. Q. (2017). Densely connected convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 4700–4708.
Jégou, S., Drozdzal, M., Vazquez, D., Romero, A., and Bengio, Y. (2017). The one hundred layers tiramisu: Fully convolutional densenets for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pages 11–19.
Kalejahi, B. K., Khan, S., and Zakirov, R. (2026). Bone-cnn: A lightweight deep learning architecture for multi-class classification of primary bone tumours in radiographs. Biomedicines, 14(2):299.
Kaliraj, V., Mangaleswaran, M., Usha, N. S., and Dharaniya, R. (2026). Automated artificial intelligent approach for enhancing bone cancer detection through hybrid feature extraction and adaptive elman recurrent neural network. Radiation Physics and Chemistry, page 113732.
Li, J., Li, S., Li, X., Miao, S., Dong, C., Gao, C., and Cui, J. (2023). Primary bone tumor detection and classification in full-field bone radiographs via yolo deep learning model. European Radiology, 33(6):4237–4248.
Li, L., Tao, W., Eid, J. E., Gupta, S., Zheng, S., Taswell, C. S., and Trent, J. C. (2026). Advances in radiation therapy for primary bone malignancies. Technology in Cancer Research & Treatment, 25:15330338261425345.
Liu, Z., Mao, H., Wu, C. Y., Feichtenhofer, C., Darrell, T., and Xie, S. (2022). A convnet for the 2020s. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 11976–11986.
Loshchilov, I. and Hutter, F. (2017). Sgdr: Stochastic gradient descent with warm restarts. In International Conference on Learning Representations (ICLR).
Naseer, M. M., Veerabhadraiah, V., Ramji, B. R., Lokeshappa, S. K. K., Nagendra, S. C., and Ramesh, Y. G. (2025). An optimized efficientnet-b0 framework for multi-class brain tumour detection and classification from mri images. Biomedical and Pharmacology Journal, 18. October Special Edition.
Oktay, O., Schlemper, J., Folgoc, L. L., Lee, M., Heinrich, M., Misawa, K., Mori, K., Mc-Donagh, S., Hammerla, N. Y., Kainz, B., Glocker, B., and Rueckert, D. (2018). Attention u-net: Learning where to look for the pancreas. arXiv preprint arXiv:1804.03999.
Rhyou, S. Y., Bang, C., Cho, Y. J., Bae, H., Ha, Y. J., Baek, S. Y., and Moon, J. E. (2025). Tracking temporal progression of benign bone tumors through x-ray based detection and segmentation. Scientific Reports, 15(1):39491.
Tan, M. and Le, Q. (2019). Efficientnet: Rethinking model scaling for convolutional neural networks. In Proceedings of the International Conference on Machine Learning (ICML), pages 6105–6114. PMLR.
Woo, S., Park, J., Lee, J., and Kweon, I. S. (2018). Cbam: Convolutional block attention module. In Proceedings of the European Conference on Computer Vision (ECCV), pages 3–19.
Xie, Z., Zhao, K., Yan, X., Wu, S., Mei, J., and Lu, H. (2022). Merged u-net for bone tumors x-ray images segmentation. In 2022 IEEE International Conference on Image Processing (ICIP), pages 1276–1280. IEEE.
Yao, S., Huang, Y., Wang, X., Zhang, Y., Paixao, I. C., Wang, Z., and Song, J. (2025). A radiograph dataset for the classification, localization, and segmentation of primary bone tumors. Scientific Data, 12(1):88.
Zuiderveld, K. J. (1994). Contrast limited adaptive histogram equalization. In Heckbert, P. S., editor, Graphics Gems, pages 474–485. Elsevier.
Bloier, M., Hinterwimmer, F., Breden, S., Consalvo, S., Neumann, J., Wilhelm, N., and Burgkart, R. (2022). Detection and segmentation of heterogeneous bone tumours in limited radiographs. Current Directions in Biomedical Engineering, 8(2):69–72.
Das, S. and Majumder, S. (2020). Lung cancer detection using deep learning network: A comparative analysis. In Proceedings of the Fifth International Conference on Research in Computational Intelligence and Communication Networks (ICRCICN), pages 30–35. IEEE.
He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778.
Huang, G., Liu, Z., Maaten, L. V. D., and Weinberger, K. Q. (2017). Densely connected convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 4700–4708.
Jégou, S., Drozdzal, M., Vazquez, D., Romero, A., and Bengio, Y. (2017). The one hundred layers tiramisu: Fully convolutional densenets for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pages 11–19.
Kalejahi, B. K., Khan, S., and Zakirov, R. (2026). Bone-cnn: A lightweight deep learning architecture for multi-class classification of primary bone tumours in radiographs. Biomedicines, 14(2):299.
Kaliraj, V., Mangaleswaran, M., Usha, N. S., and Dharaniya, R. (2026). Automated artificial intelligent approach for enhancing bone cancer detection through hybrid feature extraction and adaptive elman recurrent neural network. Radiation Physics and Chemistry, page 113732.
Li, J., Li, S., Li, X., Miao, S., Dong, C., Gao, C., and Cui, J. (2023). Primary bone tumor detection and classification in full-field bone radiographs via yolo deep learning model. European Radiology, 33(6):4237–4248.
Li, L., Tao, W., Eid, J. E., Gupta, S., Zheng, S., Taswell, C. S., and Trent, J. C. (2026). Advances in radiation therapy for primary bone malignancies. Technology in Cancer Research & Treatment, 25:15330338261425345.
Liu, Z., Mao, H., Wu, C. Y., Feichtenhofer, C., Darrell, T., and Xie, S. (2022). A convnet for the 2020s. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 11976–11986.
Loshchilov, I. and Hutter, F. (2017). Sgdr: Stochastic gradient descent with warm restarts. In International Conference on Learning Representations (ICLR).
Naseer, M. M., Veerabhadraiah, V., Ramji, B. R., Lokeshappa, S. K. K., Nagendra, S. C., and Ramesh, Y. G. (2025). An optimized efficientnet-b0 framework for multi-class brain tumour detection and classification from mri images. Biomedical and Pharmacology Journal, 18. October Special Edition.
Oktay, O., Schlemper, J., Folgoc, L. L., Lee, M., Heinrich, M., Misawa, K., Mori, K., Mc-Donagh, S., Hammerla, N. Y., Kainz, B., Glocker, B., and Rueckert, D. (2018). Attention u-net: Learning where to look for the pancreas. arXiv preprint arXiv:1804.03999.
Rhyou, S. Y., Bang, C., Cho, Y. J., Bae, H., Ha, Y. J., Baek, S. Y., and Moon, J. E. (2025). Tracking temporal progression of benign bone tumors through x-ray based detection and segmentation. Scientific Reports, 15(1):39491.
Tan, M. and Le, Q. (2019). Efficientnet: Rethinking model scaling for convolutional neural networks. In Proceedings of the International Conference on Machine Learning (ICML), pages 6105–6114. PMLR.
Woo, S., Park, J., Lee, J., and Kweon, I. S. (2018). Cbam: Convolutional block attention module. In Proceedings of the European Conference on Computer Vision (ECCV), pages 3–19.
Xie, Z., Zhao, K., Yan, X., Wu, S., Mei, J., and Lu, H. (2022). Merged u-net for bone tumors x-ray images segmentation. In 2022 IEEE International Conference on Image Processing (ICIP), pages 1276–1280. IEEE.
Yao, S., Huang, Y., Wang, X., Zhang, Y., Paixao, I. C., Wang, Z., and Song, J. (2025). A radiograph dataset for the classification, localization, and segmentation of primary bone tumors. Scientific Data, 12(1):88.
Zuiderveld, K. J. (1994). Contrast limited adaptive histogram equalization. In Heckbert, P. S., editor, Graphics Gems, pages 474–485. Elsevier.
Publicado
19/07/2026
Como Citar
NEVES, Iulle M. G.; CLARK, Mauro A. G.; ARAÚJO, José D. L.; CARVALHO FILHO, Antonio O. de.
Uma Abordagem Sistemática para Seleção de Encoders aplicada à Segmentação de Tumores com TransUNet. In: SEMINÁRIO INTEGRADO DE SOFTWARE E HARDWARE (SEMISH), 53. , 2026, Gramado/RS.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2026
.
p. 782-793.
ISSN 2595-6205.
DOI: https://doi.org/10.5753/semish.2026.23389.
