Building Personalized Risk Prediction Models fBuilding Personalized Risk Prediction Models for Polycystic Ovary Syndrome Using Machine Learning Techniques: A Retrospectivor Polycystic Ovary Syndrome Using Machine Learning Techniques: A Retrospective Study

Abstract

Due to the highly diverse pathological features of Polycystic Ovary Syndrome (PCOS), traditional diagnostic methods often struggle to accurately predict the risk of the disease. With the increasing application of artificial intelligence technology in the medical field, machine learning offers a new approach to enhance the accuracy and efficiency of PCOS prediction.In this study, we employed ten machine learning algorithms, including Random Forest (RF), Linear Discriminant Analysis (LDA), and Support Vector Machine (SVM), to analyze clinical retrospective data from multiple centers to identify and evaluate key risk factors associated with PCOS. First, rigorous data preprocessing was conducted, followed by feature selection using the LASSO regression method. Cross-validation techniques were then used to assess the performance of each prediction model on the training dataset. Finally, the models were validated on an independent test set to evaluate their generalization ability.In cross-validation, the RF model excelled in all performance metrics, particularly achieving an average accuracy of 92.56% and an F1 score of 92.63%. Evaluation results on the test set also confirmed the superior performance of the RF model, with LDA showing outstanding performance in specific metrics. Furthermore, the selected 12 key risk factors demonstrated significant clinical relevance for PCOS prediction.This study demonstrates the potential application of machine learning methods in PCOS risk prediction. Our model accurately identifies high-risk PCOS patients, providing robust data support for early diagnosis and personalized treatment. This work not only improves the efficiency and effectiveness of PCOS management but also lays the foundation for future applications of artificial intelligence technology in other complex diseases.