OvaHybrid: Channel Fusion Representation and Explainable Classification of Ovarian Cancer Subtypes in Histopathological Images

• Raphael S. R. Rates UFCA
• Luana B. da Cruz UFCA
• João O. B. Diniz IFMA

DOI: https://doi.org/10.5753/sbcas.2026.21372

Abstract

Ovarian cancer is characterized by high mortality and heterogeneity of histological subtypes, which challenges early diagnosis. This work presents OvaHybrid, a hybrid approach that integrates image processing and Deep Feature extraction to classify ovarian cancer subtypes in histopathological images. The method uses channel fusion combined with Deep Features to enrich the representation of features. Hyperparameter optimization was performed using the Optuna framework. The best configuration (ResNet50 + SVM) achieved an accuracy of 97.96%, precision of 98.13%, sensitivity of 97.96%, specificity of 98.95% and an F1-score of 97.95%, indicating high effectiveness in diagnosis.

References

Alahmadi, A. (2024). Towards ovarian cancer diagnostics: A vision transformer-based computer-aided diagnosis framework with enhanced interpretability.

Alvino, A., Alves, E., Brito, S., Nascimento, V., Cruz, L., Diniz, J., Jr, L. S., Silva, J., and Jr, D. G. (2025). Abordagem baseada em deep features para diagnóstico de câncer seroso de ovário em imagens histopatológicas. In Anais do XXV Simpósio Brasileiro de Computação Aplicada à Saúde, pages 401–412, Porto Alegre, RS, Brasil. SBC.

Beaufort, C. M. et al. (2014). Ovarian cancer cell line panel (occp): clinical importance of in vitro morphological subtypes. PloS one, 9(9):e103988.

Bikku, T., Pujari, J. J., Satyasree, K., Thota, S., and Joseph, S. (2025). Deep learning-based ovarian cancer detection from histopathology images. Quality & Quantity, pages 1–15.

El-Latif, E. I. A., El-Dosuky, M., Darwish, A., and Hassanien, A. E. (2024). A deep learning approach for ovarian cancer detection and classification based on fuzzy deep learning.

Gonzalez, R. C. and Woods, R. E. (2010). Modelos de Cores. Pearson Prentice Hall, 3 edition.

Instituto Nacional de Câncer (INCA) (2026). Estimativa 2026: Incidência de Câncer no Brasil. Instituto Nacional de Câncer, Rio de Janeiro, RJ, Brasil. Relatório técnico sobre estimativa de incidência de câncer no Brasil.

Kasture, K. (2021). Ovariancancer & subtypes dataset histopathology. Mendeley Data, V1. Disponível em: [link]. Acesso em: 18 Fev. 2026.

Mallya, M., Mirabadi, A. K., Farnell, D., Farahani, H., and Bashashati, A. (2025). Benchmarking histopathology foundation models for ovarian cancer bevacizumab treatment response prediction from whole slide images. Discover Oncology, 16(1):196.

Pavel, C. D., Moldovanu, S., Pavel, I. A., Dragostin, O.-M., Chit,escu, C. L., and Zamfir, C. L. (2025). Hybrid deep learning models for analyzing histological images of the zebrafish intestine under oxidative stress. Diagnostics, 15(22):2859.

Peneira, C., Maciel, M. G., and de Lima, Á. M. (2025). O aumento de câncer em pacientes jovens. UniLS Acadêmica, 3(1):14–14.

Rahman, M. M., Reem, S. Z., Rahaman, M. A., Bepery, C., Ashiqussalehin, M., Hossain, M. J., et al. (2025). Explainable transfer learning for histopathological image classification of ovarian cancer. Journal of Technological Enquiry and Computer Miscellaneous, 1(1).

Rawi, A. A., Elbashir, M. K., and Ahmed, A. M. (2023). Deep learning models for multilabel ecg abnormalities classification: A comparative study using tpe optimization. Journal of Intelligent Systems, 32(1):20230002.

Ribeiro, M. T., Singh, S., and Guestrin, C. (2016). ”why should i trust you?”explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, pages 1135–1144.

Saha, A. K. et al. (2025). An enhanced deep learning model for accurate classification of ovarian cancer detection from histopathological images. Scientific Reports, 15(1):152180.

Soslow, R. A. (2008). Histologic subtypes of ovarian carcinoma: an overview. International Journal of Gynecological Pathology, 27(2):161–174.

Sung, H. et al. (2021). Global cancer statistics 2020: Globocan estimates of incidence and mortality worldwide for 36 cancers in 185 countries. 71(3):209–249.

Tamanini, B. A. et al. (2025). Classificaçao de carcinoma endometrioide de ovário por transformaçao de esquema de cor e radiomics em imagens histopatológicas. In Simpósio Brasileiro de Computação Aplicada à Saúde (SBCAS), pages 68–79. SBC.

World Ovarian Cancer Coalition (2022). Ovarian cancer – key stats. [link]. Acessado: 18 Fev. 2026.