A Comparative Study of Machine Learning Algorithms for Embryo Selection in In Vitro Fertilization

Embryo selection plays a critical role in the success of In Vitro Fertilization (IVF), yet conventional methods relying on morphological assessment and expert judgment are often subjective and inconsistent. To address these limitations, this study investigates the application of machine learning (ML) algorithms for automated, objective, and accurate embryo classification. A comparative analysis was conducted on six prominent ML models: Support Vector Machine (SVM), Random Forest (RF), k-Nearest Neighbors (k-NN), Convolutional Neural Network (CNN), ResNet, and ConvNeXT. Publicly available datasets, including IVF time-lapse sequences and the Early Embryo Viability Assessment (Eeva) dataset, were utilized for training and validation. Evaluation metrics such as accuracy, precision, recall, F1-score, and Area Under the Curve (AUC) were employed to assess model performance. The results show that deep learning models, especially ConvNeXT and ResNet, performed better than traditional algorithms, reaching accuracy scores over 93% and high AUC values.