Machine Learning-Based Detection and Diagnosis of Polycystic Ovary Syndrome (PCOS)

Abstract: Polycystic Ovary Syndrome (PCOS) is a condition that affects women during their reproductive years. This project aims to reduce the risk of serious health complications by enabling early detection of PCOS through advanced machine learning techniques. Using a dataset from Kaggle that includes both clinical and physical attributes of women, the project focuses on predicting PCOS effectively. Additional features integrated into the system include a menstrual cycle tracker, customized diet and yoga plans, PCOS detection via ultrasound imagery, and access to virtual doctor consultations. To support this, three distinct machine learning models have been developed: PCOS Model 1, which achieved 97% accuracy using the XGBoost algorithm; PCOS Model 2, with 92% accuracy using Random Forest; and the Image PCOS Model, which attained 96% accuracy using a Convolutional Neural Network (CNN). These models significantly enhance early diagnosis efforts and contribute to creating a holistic, user-friendly platform for managing women's health.

Keywords: Consultation, detection, hormonal, PCOS, XGBoost, ovary, menstrual etc.