MACHINE LEARNING AND ARTIFICIAL INTELLIGENCE FOR THE DIAGNOSIS OF POLYCYSTIC OVARY SYNDROME (PCOS): A NARRATIVE REVIEW

Polycystic Ovary Syndrome (PCOS), now broadly termed as Polyendocrine Metabolic Syndrome, is one of the most common endocrine disorders among women of reproductive age, with high global prevalence ranging from approximately 6% to 21% depending on the diagnostic criteria applied, and the majority of cases remaining clinically undiagnosed. Conventional diagnosis relies on the Rotterdam criteria, which require evaluating menstrual irregularity, hyperandrogenism (clinical or biochemical), and polycystic ovarian morphology on USG — a time-consuming, resource-intensive, and subject to inter-observer variability. In recent years, Artificial Intelligence (AI) and Machine Learning (ML) techniques have been widely used to automate and improve the accuracy of PCOS diagnosis. This paper presents a narrative review of recent ML-based approaches for PCOS diagnosis, covering multi-model approaches, deep neural learning architectures, feature selection strategies, and explainable AI techniques. Twelve representative studies published between 2022 and 2025 are examined, with reported diagnostic accuracies ranging from approximately 82.5% to 99.8%. The most frequently used algorithms include Random Forest, Support Vector Machine, XGBoost, and stacked ensemble classifiers, with a Kaggle-hosted dataset of 541 patients from Kerala, India, being the most widely reused public resource. Despite promising results, this review identifies significant gaps, including over-reliance on a small number of public datasets, inconsistent handling of class imbalance, constrained external validation across diverse populations, and limited adoption of standardized Rotterdam-based diagnostic criteria in model design. Recommendations for future perspectives, including region-specific data collection and explainable, clinically deployable models, are discussed.

• Barrera, F. J., Brown, E. D. L., Rojo, A., Obeso, J., Plata, H., Lincango, E. P., Terry, N., Rodríguez-Gutiérrez, R., Hall, J. E., & Shekhar, S. (2023). Application of machine learning and artificial intelligence in the diagnosis and classification of polycystic ovarian syndrome: A systematic review. Frontiers in Endocrinology, 14: 1106625. https://doi.org/10.3389/fendo.2023.1106625
• Mohi Uddin, K. M., Bhuiyan, M. T. A., Rahman, M. M., Islam, M. M., & Uddin, M. A. (2025). Early PCOS detection: A comparative analysis of traditional and ensemble machine learning models with advanced feature selection. Engineering Reports, 7: e70008. https://doi.org/10.1002/eng2.70008
• Neven, A. C. H., Forslund, M., Ranasinha, S., Sethi, P., Dhungana, R. R., Mousa, A., Tay, C. T., Teede, H., & Boyle, J. A. (2026). Prevalence of polycystic ovary syndrome: A global and regional systematic review and meta-analysis. Human Reproduction Update, 32(3): 277–312. https://doi.org/10.1093/humupd/dmaf030
• Panjwani, B., Yadav, J., Mohan, V., Agarwal, N., & Agarwal, S. (2025). Optimized machine learning for the early detection of polycystic ovary syndrome in women. Sensors, 25(4): 1166. https://doi.org/10.3390/s25041166
• Rahman, M. M., et al. (2024). Empowering early detection: A web-based machine learning approach for PCOS prediction. Informatics in Medicine Unlocked, 47: 101500. https://doi.org/10.1016/j.imu.2024.101500
• Scientific Reports (Nature Portfolio). (2025). A machine learning approach for non-invasive PCOS diagnosis from ultrasound and clinical features. Scientific Reports. https://doi.org/10.1038/s41598-025-10453-9
• Silva, I. S., Ferreira, C. N., Costa, L. B. X., et al. (2022). Polycystic ovary syndrome: Clinical and laboratory variables related to new phenotypes using machine-learning models. Journal of Endocrinological Investigation, 45: 1–9.
• World Health Organization. (2021). Breast cancer and related endocrine disorders — WHO fact sheets on reproductive health. https://www.who.int
• Zad, Z., Jiang, V. S., Wolf, A. T., Wang, T., Cheng, J. J., Paschalidis, I. Ch., & Mahalingaiah, S. (2024). Predicting polycystic ovary syndrome with machine learning algorithms from electronic health records. Frontiers in Endocrinology, 15: 1298628. https://doi.org/10.3389/fendo.2024.1298628