A robust AI-pipeline for ovarian cancer classification on histopathology images

Haitham Kussaibi; Elaf Alibrahim; Eman Alamer; Ghadah   Al hajji; Shrooq Alshehab; Zahra Shabib; Noor Alsafwani; Ritesh G. Menezes

doi:10.23750/abm.v95i5.16407

A robust AI-pipeline for ovarian cancer classification on histopathology images

Authors

Haitham Kussaibi Department of Pathology, College of Medicine, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia https://orcid.org/0000-0002-9570-0768
Elaf Alibrahim College of Medicine, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia
Eman Alamer College of Medicine, Imam Abdulrahman bin Faisal University, Dammam, Saudi Arabia
Ghadah Al hajji College of Medicine, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia
Shrooq Alshehab College of Medicine, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia
Zahra Shabib College of Medicine, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia
Noor Alsafwani Department of Pathology, College of Medicine, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia https://orcid.org/0000-0002-9449-9708
Ritesh G. Menezes Department of Pathology, College of Medicine, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia https://orcid.org/0000-0002-2135-4161

DOI: https://doi.org/10.23750/abm.v95i5.16407

Keywords:

Ovarian cancer, artificial intelligence, whole slide images, Deep Learning, Histopathology

Abstract

Background and aim: Ovarian cancer is the leading cause of gynecological cancer deaths due to late diagnosis and high recurrence rates. While histopathological analysis is the gold standard for diagnosis, artificial intelligence (AI) models have shown promise in accurately classifying ovarian cancer subtypes using histopathology images. Herein, we introduce an end-to-end AI pipeline for automated identification of epithelial ovarian cancer (EOC) subtypes based on histopathology images and evaluate its performance compared to pathologists' diagnoses.

Methods: A dataset of over 2 million image tiles from 82 whole slide images (WSIs) of major EOC subtypes (clear cell, endometrioid, mucinous, and serous) was curated from public and institutional sources. A convolutional neural network (ResNet50) was used to extract features, which were then introduced to 2 classifiers (NN and LightGBM) to predict cancer subtypes.

Results: Both AI classifiers achieved patch-level accuracy (97-98%) on a test set. Furthermore, adding a class-weighted cross-entropy loss function to the pipeline showed better discriminative performance among subtypes.

Conclusions: AI models trained on histopathology images can accurately classify EOC subtypes, potentially assisting pathologists and reducing subjectivity in ovarian cancer diagnosis.

Author Biographies

Haitham Kussaibi, Department of Pathology, College of Medicine, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia

Associate Professor, Department of Pathology, College of Medicine

Master BioMedical Informatic, Paris Nord University, Paris, France

Noor Alsafwani, Department of Pathology, College of Medicine, Imam Abdulrahman Bin Faisal University, Dammam, Saudi Arabia

Assistant Professor, Department of Pathology

References

Momenimovahed Z, Tiznobaik A, Taheri S, Salehiniya H. Ovarian cancer in the world: epidemiology and risk factors. Int J Womens Health. 2019; 11:287-99. doi: 10.2147/IJWH.S197604.

Al-Rawaji A. Cancer Incidence Report 2020 - Kindom of Saudi Arabia, Saudi Health Council, National Cancer Center, Saudi Cancer registry, 2020,1-81.

Mikdadi D, O'Connell KA, Meacham PJ, et al. Applications of artificial intelligence (AI) in ovarian cancer, pancreatic cancer, and image biomarker discovery. Cancer Biomark. 2022;33(2):173-84. doi:10.3233/CBM-210301.

Zhou J, Cao W, Wang L, Pan Z, Fu Y. Application of artificial intelligence in the diagnosis and prognostic prediction of ovarian cancer. Comput Biol Med. 2022; 146:105608. doi: 10.1016/j.compbiomed.2022.105608.

Orsulic S, John J, Walts AE, Gertych A. Computational pathology in ovarian cancer. Front Oncol. 2022; 12:924945. doi: 10.3389/fonc.2022.924945.

Devouassoux-Shisheboran M, Genestie C. Pathobiology of ovarian carcinomas. Chin J Cancer. 2015;34(1):50-5. doi: 10.5732/cjc.014.10273.

Ignacio R, Susanna L, Belén Pérez M, Andrés Poveda V, José P. Morphological and molecular heterogeneity of epithelial ovarian cancer: therapeutic implications. Eur J Cancer Suppl. 2020; 15:1-15. doi: 10.1016/j.ejcsup.2020.02.001.

Lalwani N, Prasad SR, Vikram R, Shanbhogue AK, Huettner PC, Fasih N. Histologic, molecular, and cytogenetic features of ovarian cancers: implications for diagnosis and treatment. Radiographics. 2011;31(3):625-46. doi: 10.1148/rg.313105066.

Rahman H, Naik Bukht TF, Ahmad R, Almadhor A, Javed AR. Efficient breast cancer diagnosis from complex mammographic images using deep convolutional neural network. Comput Intell Neurosci. 2023; 2023:7717712. doi: 10.1155/2023/7717712.

Aliza B. Artificial intelligence in medicine: what is it doing for us today? Health Policy Technol. 2019; 8(2):198-205. doi: 10.1016/j.hlpt.2019.03.004.

Myszczynska MA, Ojamies PN, Lacoste AMB, et al. Applications of machine learning to diagnosis and treatment of neurodegenerative diseases. Nat Rev Neurol. 2020; 16(8):440-56. doi: 10.1038/s41582-020-0377-8.

El-Ghany SA, Azad M, Elmogy M. Robustness fine-tuning deep learning model for cancers diagnosis based on histopathology image analysis. Diagnostics (Basel). 2023; 13(4):806. doi: 10.3390/diagnostics13040699.

Dolezal JM, Kochanny S, Dyer E, et al. Slideflow: deep learning for digital histopathology with real-time whole-slide visualization. BMC Bioinformatics. 2024; 25(1):134. doi: 10.1186/s12859-024-05758-x.

Cai L, Gao J, Zhao D. A review of the application of deep learning in medical image classification and segmentation. Ann Transl Med. 2020; 8(11):713. doi: 10.21037/atm.2020.02.44.

Taye MM. Theoretical understanding of convolutional neural network: concepts, architectures, applications, future directions. Computation. 2023; 11(3):52. doi: 10.3390/computation11030052.

Muhammed T. Automated classification of histopathology images using transfer learning. Artif Intell Med. 2019; 101:101743. doi: 10.1016/j.artmed.2019.101743.

Wu M, Yan C, Liu H, Liu Q. Automatic classification of ovarian cancer types from cytological images using deep convolutional neural networks. Biosci Rep. 2018; 38(3): BSR20171150. doi: 10.1042/BSR20180289.

Farahani H, Boschman J, Farnell D, et al. Deep learning-based histotype diagnosis of ovarian carcinoma whole-slide pathology images. Mod Pathol. 2022; 35(12):1983-90. doi: 10.1038/s41379-022-01146-z.

Jiang J, Tekin B, Guo R, Liu H, Huang Y, Wang C. Digital pathology-based study of cell- and tissue-level morphologic features in serous borderline ovarian tumor and high-grade serous ovarian cancer. J Pathol Inform. 2021; 12:24. doi: 10.4103/jpi.jpi_76_20.

Liu Y, Lawson BC, Huang X, Broom BM, Weinstein JN. Prediction of ovarian cancer response to therapy based on deep learning analysis of histopathology images. Cancers (Basel). 2023; 15(16):4085. doi: 10.3390/cancers15164044.

Laury AR, Blom S, Ropponen T, Virtanen A, Carpén OM. Artificial intelligence-based image analysis can predict outcome in high-grade serous carcinoma via histology alone. Sci Rep. 2021; 11(1):19165. doi: 10.1038/s41598-021-98480-0.

Tanabe K, Ikeda M, Hayashi M, et al. Comprehensive serum glycopeptide spectra analysis combined with artificial intelligence (CSGSA-AI) to diagnose early-stage ovarian cancer. Cancers (Basel). 2020; 12(9):2551. doi: 10.3390/cancers12092373.

Nero C, Boldrini L, Lenkowicz J, et al. Deep-learning to predict BRCA mutation and survival from digital H&E slides of epithelial ovarian cancer. Int J Mol Sci. 2022; 23(19):11539. doi: 10.3390/ijms231911326.

Hu F, Schutt L, Kozlowski C, Regan K, Dybdal N, Schutten MM. Ovarian toxicity assessment in histopathological images using deep learning. Toxicol Pathol. 2020; 48(2):350-61. doi: 10.1177/0192623319877871.

Breen J, Allen K, Zucker K, et al. Artificial intelligence in ovarian cancer histopathology: a systematic review. NPJ Precis Oncol. 2023; 7(1):83. doi: 10.1038/s41698-023-00432-6.

Aljerian N, Arafat M, Aldhubib A, et al. Artificial intelligence in healthcare and its application in Saudi Arabia. Int J Innov Res Med Sci. 2022; 7:666-70. doi: 10.23958/ijirms/vol07-i11/1558.

Abdulrahman A. Towards ovarian cancer diagnostics: a vision transformer-based computer-aided diagnosis framework with enhanced interpretability. Results Eng. 2024; 23:102651. doi: 10.1016/j.rineng.2024.102651.

Wang C-W, Chang CC, Lo SC, et al. A dataset of histopathological whole slide images for classification of treatment effectiveness to ovarian cancer (Ovarian Bevacizumab Response) (Version 2) [Data set]. Cancer Imaging Arch. 2021. doi: 10.7937/TCIA.985G-EY35.

Bankhead P, Loughrey MB, Fernández JA, et al. QuPath: open-source software for digital pathology image analysis. Sci Rep. 2017; 7:168-78. doi: 10.1038/s41598-017-17204-5.

Lim D, Murali R, Murray MP, Veras E, Park KJ, Soslow RA. Morphological and immunohistochemical reevaluation of tumors initially diagnosed as ovarian endometrioid carcinoma with emphasis on high-grade tumors. Am J Surg Pathol 2016; 40:302-12. doi: 10.1097/pas.0000000000000550