International Journal of Engineering and Manufacturing (IJEM)
IJEM Vol. 16, No. 5, 8 Oct. 2026
Cover page and Table of Contents: PDF (size: 1064KB)
Attention-Guided Deep Learning Framework for Ovarian Cancer Subtype Classification
PDF (1064KB), PP.158-178
Views: 0 Downloads: 0
Author(s)
Vijay H. Kalmani
1,*
Nagaraj V. Dharwadkar
2,*
Amol C. Adamuthe
3
Altaf Husain
4
Index Terms
Ovarian cancer, Computational pathology, Whole-slide images, Pathology foundation model, Multiple-instance learning, Gated attention, Histopathology classification
Abstract
Ovarian cancer histotype classification is challenging because of substantial morphological heterogeneity and subtle subtype-specific features. This work presents a controlled evaluation of high-resolution pathology representations and slide-level aggregation methods for automated classification of five ovarian cancer subtypes from whole-slide histopathology images. Precomputed CONCH patch embeddings were aggregated using mean pooling, max pooling, gated attention-based multiple-instance learning, and a max-pooling cascade. The models were evaluated on 513 non-TMA whole-slide images from UBC-OCEAN using leakage-controlled five-fold cross-validation, with each slide receiving exactly one out-of-fold prediction. Gated ABMIL achieved a balanced accuracy of 81.39% and a macro-F1 score of 81.96%. Max pooling produced the highest numerical performance, with a balanced accuracy of 81.44%, macro-F1 of 82.59% (95% CI: 78.84-86.26%), macro-AUROC of 96.99%, and macro-AUPRC of 90.79%. However, paired slide-level bootstrap comparisons found no statistically significant differences among the CONCH aggregation strategies after Holm correction. Compared with the EfficientNet-B0 thumbnail baseline, CONCH max pooling improved macro-F1 by 30.30 percentage points and balanced accuracy by 25.70 percentage points, with both paired bootstrap confidence intervals excluding zero. Attention weights enabled visualization of influential patches, although these regions were not independently validated by pathologists. The findings show that high-resolution pathology foundation-model representations support ovarian cancer subtyping, while greater aggregation complexity does not necessarily improve performance. External multi-institutional validation is required before clinical generalizability can be established.
Cite This Paper
Vijay H. Kalmani, Nagaraj V. Dharwadkar, Amol C. Adamuthe, Altaf Husain, "Attention-Guided Deep Learning Framework for Ovarian Cancer Subtype Classification", International Journal of Engineering and Manufacturing (IJEM), Vol.16, No.5, pp. 158-178, 2026. DOI:10.5815/ijem.2026.05.09
Reference
[1]I. Tsibulak and C. Fotopoulou, “Tumor biology and impact on timing of surgery in advanced epithelial ovarian cancer,” International Journal of Gynecological Cancer, vol. 33, no. 10, pp. 1627–1632, Aug. 2023.
[2]Z. Liu, C. Jing, and F. Kong, “From clinical management to personalized medicine: Novel therapeutic approaches for ovarian clear cell cancer,” Journal of Ovarian Research, vol. 17, no. 1, Feb. 2024.
[3]J. Xu et al., “A hierarchical integration deep flexible neural forest framework for cancer subtype classification by integrating multi-omics data,” BMC Bioinformatics, vol. 20, no. 1, Oct. 2019.
[4]D. Zhang, P. Chen, C.-H. Zheng, and J. Xia, “Identification of ovarian cancer subtype-specific network modules and candidate drivers through an integrative genomics approach,” Oncotarget, vol. 7, no. 4, pp. 4298–4309, Dec. 2015.
[5]D. P. Cook and B. C. Vanderhyden, “Ovarian cancer and the evolution of subtype classifications using transcriptional profiling,” Biology of Reproduction, vol. 101, no. 3, pp. 645–658, Jun. 2019.
[6]G. Wang et al., “Machine learning-based rapid diagnosis of human borderline ovarian cancer on second-harmonic generation images,” Biomedical Optics Express, vol. 12, no. 9, p. 5658, Sep. 2021.
[7]B. Ziyambe et al., “A deep learning framework for the prediction and diagnosis of ovarian cancer in pre and post-menopausal women,” Diagnostics, vol. 13, no. 10, p. 1703, May 2023.
[8]M. L. D´oria et al., “Epithelial ovarian carcinoma diagnosis by desorption electrospray ionization mass spectrometry imaging,” Scientific Reports, vol. 6, no. 1, Dec. 2016.
[9]A. Titoriya and S. Sachdeva, “Breast cancer histopathology image classification using AlexNet,” in Proceedings of the 4th International Conference on Information Systems and Computer Networks (ISCON), 2019, pp. 708–712.
[10]J. Sun and A. Binder, “Comparison of deep learning architectures for H&E histopathology images,” in Proceedings of the International Conference on Big Data Analytics and Applications (ICBDAA), 2017, pp. 43–48.
[11]M. M. Srikantamurthy, V. P. S. Rallabandi, D. B. Dudekula, S. Natarajan, and J. Park, “Classification of benign and malignant subtypes of breast cancer histopathology imaging using hybrid CNN-LSTM-based transfer learning,” BMC Medical Imaging, vol. 23, no. 1, Jan. 2023.
[12]S. Tabibu, P. K. Vinod, and C. V. Jawahar, “Pan-renal cell carcinoma classification and survival prediction from histopathology images using deep learning,” Scientific Reports, vol. 9, no. 1, Jul. 2019.
[13]Y. Guo, S. Liu, Z. Li, and X. Shang, “Towards the classification of cancer subtypes by using cascade deep forest model in gene expression data,” in Proceedings of the IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2017, pp. 1664–1669.
[14]“BCDForest: A boosting cascade deep forest model towards the classification of cancer subtypes based on gene expression data,” BMC Bioinformatics, vol. 19, no. S5, Apr. 2018.
[15]S. P. Blagden, “Harnessing pandemonium: The clinical implications of tumor heterogeneity in ovarian cancer,” Frontiers in Oncology, vol. 5, Jun. 2015.
[16]R. Chen, L. Yang, S. Goodison, and Y. Sun, “Deep-learning approach to identifying cancer subtypes using high-dimensional genomic data,” Bioinformatics, vol. 36, no. 5, pp. 1476–1483, 2020.
[17]J. Wang, Z. Zhang, and Y. Wang, “Utilizing feature selection techniques for AI-driven tumor subtype classification: Enhancing precision in cancer diagnostics,” Biomolecules, vol. 15, no. 1, p. 81, Jan. 2025.
[18]Y. Shen, Y. Luo, D. Shen, and J. Ke, “RandStainNA: Learning stain-agnostic features from histology slides by bridging stain augmentation and normalization,” in Medical Image Computing and Computer Assisted Intervention – MICCAI 2022, ser. Lecture Notes in Computer Science. Springer, 2022, pp. 212–221.
[19]L. Aldakhil, H. Alhasson, and S. Alharbi, “Attention-based deep learning approach for breast cancer histopathological image multi-classification,” Diagnostics, vol. 14, no. 13, p. 1402, Jul. 2024.
[20]A. Shrivastava et al., “Self-attentive adversarial stain normalization,” in Domain Adaptation and Representation Transfer, and Distributed and Collaborative Learning, ser. Lecture Notes in Computer Science. Springer, 2021, pp. 120–140.
[21]S. K. Behera, A. Das, and P. K. Sethy, “Deep fine-KNN classification of ovarian cancer subtypes using EfficientNet-B0 extracted features: A comprehensive analysis,” Journal of Cancer Research and Clinical Oncology, vol. 150, no. 7, p. 361, Jul. 2024.
[22]T. A. Fahim, F. B. Alam, and K. T. Ahmmed, “OVANet: Dual attention mechanism based new deep learning framework for diagnosis and classification of ovarian cancer subtypes from histopathological images,” IEEE Access, vol. 12, pp. 131 942–131 953, 2024.
[23]Z. Tarek and E. Hassan, “Rehearsal-based continual learning for robust ovarian cancer subtype classification under catastrophic forgetting,” Journal of Engineering and Applied Science, vol. 73, no. 1, Apr. 2026.
[24]H. Al-Asi et al., “Pathology foundation models: Evolution, current landscape, challenges and opportunities from a technical and clinical perspective,” Bioengineering, vol. 13, no. 5, p. 577, May 2026.
[25]B. Zhang et al., “Foundation models in cancer pathology: Techniques, applications, and future directions,” Research, vol. 9, 2026.
[26]H. Xu et al., “A whole-slide foundation model for digital pathology from real-world data,” Nature, vol. 630, no. 8015, pp. 181–188, May 2024.
[27]J. I. Pisula and K. Bozek, “Efficient WSI classification with sequence reduction and transformers pretrained on text,” Scientific Reports, vol. 15, no. 1, p. 5612, Feb. 2025.
[28]S. Orsulic, J. John, A. E. Walts, and A. Gertych, “Computational pathology in ovarian cancer,” Frontiers in Oncology, vol. 12, p. 924945, Jul. 2022.
[29]J. Breen, K. Allen, K. Zucker, L. Godson, N. M. Orsi, and N. Ravikumar, “A comprehensive evaluation of histopathology foundation models for ovarian cancer subtype classification,” npj Precision Oncology, vol. 9, no. 1, p. 33, Jan. 2025.
[30]J. Chun et al., “Deep learning-based diagnosis of epithelial ovarian cancer from whole-slide histopathology images,” Diagnostics, vol. 16, no. 10, p. 1470, May 2026.
[31]H. J. Song, Y. S. Cho, and Y. S. Kim, “Comparative evaluation of feature extractors, aggregation strategies, and classification hierarchies for ovarian cancer subtype classification in whole slide images,” Diagnostics, vol. 16, no. 10, p. 1570, May 2026.
[32]M. J. Sai and N. S. Punn, “KS-TMIL: A K-stage transformer approach with multiple instance learning model for ovarian cancer subtype classification,” Computers in Biology and Medicine, p. 110726, 2025.
[33]A. Bashashati and H. Farahani, “UBC ovarian cancer subtype classification and outlier detection (UBC-OCEAN),” Kaggle Competition, 2023, online; accessed: 20 August 2026. [Online]. Available: https://www.kaggle.com/competitions/UBC-OCEAN
[34]K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 770–778.
[35]S. Xie, R. Girshick, P. Doll´ar, Z. Tu, and K. He, “Aggregated residual transformations for deep neural networks,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, 5987–5995.