PREDICTIVE MODELING OF PCOS: A MULTI-LINEAR REGRESSION ANALYSIS INCORPORATING HORMONAL, CLINICAL, AND LIFESTYLE PARAMETERS

Abstract

Abstract

Polycystic Ovary Syndrome (PCOS) is a complex endocrine disorder affecting reproductive-aged women, characterized by hormonal imbalances and diverse clinical manifestations. This study aims to develop a multi-linear regression model to predict PCOS based on various parameters. Comprehensive assessments including menstrual cycle regularity, hair growth weight gain, fast food consumption, skin darkening, follicle counts (left and right ovaries), insulin levels, Anti-Mullerian Hormone (AMH) concentrations, and the presence of pimples were performed on a group of individuals presenting with PCOS symptoms. Statistical analysis involved correlation studies and the development of a multi-linear regression model to elucidate relationships between these parameters and the diagnosis of PCOS. Preliminary findings suggest significant associations between fast food intake, irregular cycles, weight gain, skin darkening, excess hair growth, follicle counts, insulin levels, AMH concentrations, and the presence of pimples with the manifestation of PCOS. The multi-linear regression model exhibited predictive capability, offering insights into the combined influence of these parameters on the likelihood of PCOS development. This Model will be able to approximately predict PCOS with the help of the symptoms.