Predicting Polycystic Ovary Syndrome Among Reproductive-Aged Women in Bangladesh Using Machine Learning Algorithms: Development of a Hospital-Based Predictive Model
November 2025
TLDR Machine learning can accurately predict Polycystic Ovary Syndrome in women using clinical features.
The study developed a predictive model for Polycystic Ovary Syndrome (PCOS) among 212 reproductive-aged women in Bangladesh using machine learning (ML) algorithms. The Extreme Gradient Boosting (XGBoost) model showed the highest predictive performance with an accuracy of 99.63%, sensitivity of 99.45%, and specificity of 99.81%. The study found that clinical features were more predictive than psychological aspects. The results suggest that ML frameworks can significantly improve PCOS prediction in resource-limited settings, and future research should incorporate biochemical indicators for broader application in women's reproductive health.
