Predictive Analysis of IVF Success in Nigeria Using Deep Learning: Male vs Female Fertility Factors

Abstract

Infertility affects millions of couples globally, with in vitro fertilization (IVF) emerging as a common assisted reproductive technology (ART). Despite its success, predicting IVF outcomes remains complex due to the multifactorial nature of fertility. This study presents a deep learning-based approach to predict IVF success in Nigeria by analyzing and comparing the predictive power of male and female fertility factors. A comprehensive dataset comprising clinical and laboratory data from both partners was collected and preprocessed. Convolutional Neural Networks (CNNs) and Deep Neural Networks (DNNs) were employed to develop models trained on male-only, female-only, and combined datasets. Evaluation metrics such as accuracy, precision, recall, F1-score, and AUC-ROC were used to assess performance. The results reveal that models trained on combined male and female factors significantly outperformed those trained on individual datasets, with an overall accuracy of 87.3% and an AUC of 0.91. Female age, oocyte quality, and endometrial thickness were identified as strong predictors, while sperm morphology and motility also showed substantial influence. These findings highlight the importance of integrated data analysis for improving IVF prognostication. This research underscores the potential of AI-driven decision support systems in enhancing clinical strategies and personalized treatment planning for infertile couples.