AI-Driven Prediction of IVF Success: A Hybrid Machine Learning Approach for Personalized Fertility Treatment

In vitro fertilization (IVF) remains a crucial assisted reproductive technology, yet its success rates continue to be unpredictable due to various patient-specific and clinical factors. Traditional statistical models used for predicting IVF outcomes are limited in accuracy, leading to uncertainties in treatment planning. Recent advancements in artificial intelligence (AI) and machine learning (ML) have demonstrated significant potential in improving IVF prognostics by analyzing large-scale clinical datasets and identifying complex, non-linear relationships. This study explores the application of ensemble learning, deep learning, and hybrid models to predict IVF success, utilizing a multi-factorial dataset that integrates patient demographics, ovarian reserve markers, stimulation protocols, embryological assessments, and real-time monitoring data. The study evaluates Random Forest + XGBoost, Artificial Neural Networks (ANN), and a Hybrid ANN + XGBoost model, demonstrating that the hybrid approach achieves the highest accuracy (F1-score: 0.86, ROC-AUC: 0.93). Further, the study highlights the importance of explainable AI (XAI), real-time data integration, and multi-center validation for improving model generalizability and clinical adoption. Future research should focus on personalized treatment recommendations, cost-effective AI solutions, and ethical considerations in AI-driven reproductive medicine. By leveraging AI, this study aims to optimize IVF success prediction, ultimately reducing financial and emotional burdens on patients while improving fertility treatment outcomes.