Early Detection of Polycystic Ovary Syndrome Using a Hybrid Machine Learning Model

Polycystic Ovary Syndrome (PCOS) is a prevalent endocrine disorder affecting 6-12% of women of reproductive age, leading to infertility, metabolic complications, and increased risk of type 2 diabetes. Early detection is crucial for timely intervention, yet traditional diagnostic methods like ultrasound and hormonal assays are invasive and resource-intensive. This study introduces a novel hybrid machine learning model integrating Linear Regression (LR) and Random Forest (RF) to predict PCOS using clinical and bio-chemical features from a dataset of 500 patients. The model achieves an accuracy of 87.0%, with precision, recall, and F1-score of 86.5%, 87.2%, and 86.8%, respectively. Feature importance analysis highlights insulin levels, body mass index (BMI), and the LH/FSH ratio as critical predictors. Visualizations, including a confusion matrix and ROC curve, enhance interpretability, offering a scalable, non-invasive diagnostic tool. Future enhancements include multi-center validation and wearable technology intearation.