Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{171428,
        author = {Prof. J. I. Nandalwar and Dr. P. M. Jawandhiya},
        title = {Leveraging Machine Learning Techniques for Early Detection of Polycystic Ovary Syndrome (PCOS) Using Clinical and Physical Parameters: A Comprehensive Analysis},
        journal = {International Journal of Innovative Research in Technology},
        year = {2024},
        volume = {11},
        number = {7},
        pages = {3194-3207},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=171428},
        abstract = {Polycystic Ovary Syndrome (PCOS) is a prevalent hormonal, complex and multifaceted endocrine disorder affecting women of reproductive age, leading to complications such as infertility, irregular menstrual cycles, and metabolic disturbances. Timely and accurate diagnosis is crucial for effective management and treatment of the condition. Its diagnosis often relies on subjective clinical assessments and invasive tests, leading to delays in detection and management. This study explores the potential of machine learning (ML) techniques to facilitate the early detection of PCOS by leveraging a dataset containing clinical and physical parameters. A dataset consisting of 541 women’s records, collected from multiple hospitals was employed in the study. The dataset includes 44 features such as age, BMI, follicle count, hormonal levels (FSH, LH, TSH, etc.), and menstrual cycle characteristics.
The data underwent comprehensive preprocessing, including handling missing values, encoding categorical variables, and feature selection to identify the most relevant predictors of PCOS. Various machine learning models were implemented, including Random Forest, Logistic Regression, Support Vector Machine (SVM), and K-Nearest Neighbors (KNN). Model performance was evaluated using accuracy, precision, recall, and F1-score and the area under the ROC curve (AUC) metrics. Among these models, Random Forest demonstrated superior performance, achieving an accuracy of 91%, with BMI, follicle count, and LH levels emerging as the most significant predictors of PCOS.
The results suggest that machine learning can serve as a valuable tool for early PCOS diagnosis, offering a non-invasive, data-driven approach that can be integrated into clinical workflows. This study not only provides a reliable predictive model for PCOS but also highlights the critical features that influence the condition, supporting clinical decision-making. Future research will explore expanding the feature set to include genetic factors and larger datasets to improve model generalization. By offering an efficient and cost-effective alternative to conventional diagnostic methods, this work contributes to the growing intersection of artificial intelligence and healthcare, advancing personalized treatment strategies for women with PCOS.},
        keywords = {},
        month = {December},
        }