Explainable Transfer Learning for Histopathological Image Classification of Ovarian Cancer
DOI:
https://doi.org/10.65496/jcste.2025.44Keywords:
Ovarian Cancer, Early Detection, XAI, Transfer Learning, DenseNetAbstract
Timely diagnosis and effective treatment of ovarian cancer depend heavily on accurate subtype classification. This study presents a deep learning model that utilizes transfer learning with a customized DenseNet201 architecture. Several subtypes of ovarian cancer are classified by the proposed model using histopathology images. Two balanced, openly accessible datasets are used for training and validation. In order to enhance generalization and avoid overfitting, data augmentation approaches are used during preprocessing. Accuracy and AUC are two of the measures used to assess the performance of the model. Interpretability is improved through the use of Local Interpretable Model-agnostic Explanations (LIME). LIME offers graphical representations of the model's decision-making procedure. For clinical applications where transparency is required, this capability is essential. On each dataset, experimental results demonstrate that the proposed model performs better than the state-of-the-art approaches currently in use. Every metric show that it performs better in classification. Even with variable input, the model exhibits good robustness and steady accuracy. The efficacy of transfer learning and the prospect of explainable AI in medical image classification are demonstrated by these findings. In particular, it emphasizes how useful it is for diagnosing ovarian cancer. The results highlight how combining explainable AI and deep learning methods can yield dependable, scalable solutions for precise and early ovarian cancer subtype diagnosis. This may increase the likelihood of timely diagnosis and individualized treatment regimens. Additionally, the finding opens the door for further research into healthcare technology and automated diagnostic tools for oncology.
Downloads
References
1. H. Sung et al., “Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries,” CA: A Cancer Journal for Clinicians, vol. 71, no. 3, pp. 209–249, 2021, doi: 10.3322/caac.21660. DOI: https://doi.org/10.3322/caac.21660
2. F. Bray et al., “Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries,” CA: A Cancer Journal for Clinicians, vol. 74, no. 3, pp. 229–263, 2024, doi: 10.3322/caac.21834. DOI: https://doi.org/10.3322/caac.21834
3. U. Menon et al., “Ovarian cancer population screening and mortality after long-term follow-up in the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS): a randomised controlled trial,” The Lancet, vol. 397, no. 10290, pp. 2182–2193, 2021, doi: 10.1016/S0140-6736(21)00877-2. DOI: https://doi.org/10.1016/S0140-6736(21)00731-5
4. J. S. Berek, M. Renz, S. Kehoe, L. Kumar, and M. Friedlander, “Cancer of the ovary, fallopian tube, and peritoneum: 2021 update,” International Journal of Gynecology and Obstetrics, vol. 155, Suppl. 1, pp. 61–85, 2021, doi: 10.1002/ijgo.13878. DOI: https://doi.org/10.1002/ijgo.13878
5. M. Köbel et al., “Ovarian Carcinoma Subtypes Are Different Diseases: Implications for Biomarker Studies,” PLOS Medicine, vol. 5, no. 12, p. e232, 2008, doi: 10.1371/journal.pmed.0050232. DOI: https://doi.org/10.1371/journal.pmed.0050232
6. M. L. Wilson, K. A. Fleming, M. A. Kuti, L. M. Looi, N. Lago, and K. Ru, “Access to pathology and laboratory medicine services: a crucial gap,” The Lancet, vol. 391, no. 10133, pp. 1927–1938, May 2018, doi: 10.1016/S0140-6736(18)30458-6. DOI: https://doi.org/10.1016/S0140-6736(18)30458-6
7. J. Zhou, W. Cao, L. Wang, Z. Pan, and Y. Fu, “Application of artificial intelligence in the diagnosis and prognostic prediction of ovarian cancer,” Computers in Biology and Medicine, vol. 146, p. 105608, 2022, doi: 10.1016/j.compbiomed.2022.105608. DOI: https://doi.org/10.1016/j.compbiomed.2022.105608
8. M. Talo, “Automated classification of histopathology images using transfer learning,” Artificial Intelligence in Medicine, vol. 101, p. 101743, 2019, doi: 10.1016/j.artmed.2019.101743. DOI: https://doi.org/10.1016/j.artmed.2019.101743
9. M. Wu, C. Yan, H. Liu, and Q. Liu, “Automatic classification of ovarian cancer types from cytological images using deep convolutional neural networks,” Bioscience Reports, vol. 38, no. 3, pp. BSR20180289, 2018, doi: 10.1042/BSR20180289. DOI: https://doi.org/10.1042/BSR20180289
10. L. Guo, A. Wu, Y. Wang et al., “Deep learning-based ovarian cancer subtypes identification using multi-omics data,” BioData Mining, vol. 13, no. 1, p. 10, 2020, doi: 10.1186/s13040-020-00222-x. DOI: https://doi.org/10.1186/s13040-020-00222-x
11. M. Wu et al., “Deep convolutional neural networks for multiple histologic types of ovarian tumors classification in ultrasound images,” Frontiers in Oncology, vol. 13, p. 1154200, 2023, doi: 10.3389/fonc.2023.1154200. DOI: https://doi.org/10.3389/fonc.2023.1154200
12. J. Breen, K. Allen, K. Zucker, N. M. Orsi, and N. Ravikumar, “Multi-Resolution Histopathology Patch Graphs for Ovarian Cancer Subtyping,” arXiv preprint, arXiv:2407.18105, 2024. [Online]. Available: https://arxiv.org/abs/2407.18105
13. H. Farahani et al., “Deep learning-based histotype diagnosis of ovarian carcinoma whole-slide pathology images,” Modern Pathology, vol. 35, pp. 1983–1990, 2022, doi: 10.1038/s41379-022-01146-z. DOI: https://doi.org/10.1038/s41379-022-01146-z
14. 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,” arXiv preprint, arXiv:2405.09990, 2024. [Online]. Available: https://arxiv.org/abs/2405.09990 DOI: https://doi.org/10.1038/s41698-025-00799-8
15. 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, p. 361, 2024, doi: 10.1007/s00432-024-05879-z. DOI: https://doi.org/10.1007/s00432-024-05879-z
16. J. Breen, K. Allen, K. Zucker, G. Hall, N. M. Orsi, and N. Ravikumar, “Efficient subtyping of ovarian cancer histopathology whole slide images using active sampling in multiple instance learning,” arXiv preprint, arXiv:2302.08867, 2023. [Online]. Available: https://arxiv.org/abs/2302.08867 DOI: https://doi.org/10.1117/12.2653869
17. G. Huang, Z. Liu, L. van der Maaten, and K. Q. Weinberger, “Densely Connected Convolutional Networks,” in Proc. IEEE Conf. Computer Vision and Pattern Recognition (CVPR), 2017, pp. 4700–4708, doi: 10.1109/CVPR.2017.243. DOI: https://doi.org/10.1109/CVPR.2017.243
18. S. Thite, “Ovarian Cancer Subtype Classification. Kaggle,” 2023. [Online]. Available: https://www.kaggle.com/datasets/sunilthite/ovarian-cancer-classification-dataset
19. K. Kasture, OvarianCancer & Subtypes Dataset Histopathology. Mendeley Data, V1, 2021, doi: 10.17632/kztymsrjx9.1.
20. J. Huang and C. Ling, “Using AUC and Accuracy in Evaluating Learning Algorithms,” IEEE Transactions on Knowledge and Data Engineering, vol. 17, pp. 299–310, Apr. 2005, doi: 10.1109/TKDE.2005.50. DOI: https://doi.org/10.1109/TKDE.2005.50
21. A. Kodipalli, V. S. Devi, S. Guruvare, and T. Ismail, “Explainable AI-based feature importance analysis for ovarian cancer classification with ensemble methods,” Frontiers in Public Health, vol. 13, p. 1479095, Mar. 2025, doi: 10.3389/fpubh.2025.1479095. DOI: https://doi.org/10.3389/fpubh.2025.1479095
