An Analytical Review of AI-Driven Techniques for Polycystic Ovary Syndrome Prediction

Fatima Khan Sarguroh, Srivaramangai R

Abstract: Polycystic Ovary Syndrome (PCOS) is a common hormonal disorder affecting women of reproductive age and is often linked to irregular periods, hormonal imbalance, weight gain, and infertility. Early diagnosis is important to prevent long-term health risks such as diabetes, cardiovascular disease, and reproductive complications. However, traditional diagnostic methods like ultrasound scans and hormone tests often fail to give accurate results because symptoms vary among individuals. In recent years, Artificial Intelligence (AI) has become an effective approach to improve PCOS prediction. This study reviews various AI-based methods, focusing on Machine Learning (ML), Deep Learning (DL), and Hybrid or Ensemble models used in PCOS diagnosis. The review shows that Machine Learning models such as Random Forest and SVM work well for structured clinical and hormonal data, while Deep Learning models like CNN and LSTM are better suited for image-based and complex data. Ensemble and Hybrid models, which combine both ML and DL techniques, provide the most reliable and accurate outcomes by integrating multiple data types. The study also highlights common challenges such as limited datasets, lack of data integration, and the need for explainable and lightweight AI systems. Overall, the findings suggest that AI-driven approaches can significantly improve early detection and management of PCOS, helping healthcare professionals provide faster and more accurate diagnosis for women?s health.

Keywords: Polycystic Ovary Syndrome (PCOS), Artificial Intelligence (AI), Machine Learning (ML), Deep Learning (DL), Ensemble Models, Medical Diagnosis, Prediction Model

How to Cite?: Fatima Khan Sarguroh, Srivaramangai R, "An Analytical Review of AI-Driven Techniques for Polycystic Ovary Syndrome Prediction", Volume 14 Issue 12, December 2025, International Journal of Science and Research (IJSR), Pages: 1347-1352, https://www.ijsr.net/getabstract.php?paperid=SR251217131307, DOI: https://dx.doi.org/10.21275/SR251217131307