Predictive Modeling of Success Rate for Embryo Transfer in Assisted Reproductive Technologies

Abstract

Abstract Background: Assisted reproductive technologies (ART) are recent improvements in infertility treatment. However, there is no significant increase in pregnancy rates and the costly and complex processes of these approaches are remaining as challenging issues. The aim of this study is to development a computational prediction model for implantation outcome after an embryo transfer cycle. Methods: In this study, information of 500 patients and 1360 transferred embryos, including cleavage and blastocyst stages and fresh or frozen embryos, from April 2016 to February 2018 were collected. Dataset containing 82 attributes and a target label (indicating positive and negative implantation outcomes) was constructed. Variety of dominant machine learning approaches was examined based on their performance to predict embryo transfer outcome. In addition, feature selection (FS) procedures were used to identify effective predictive factors and provide the best performance of each classifier according to the optimum number of features. Results: The results revealed that random forest (RF) was the best classifier, with 90.40% accuracy and 93.74% AUC, with optimum features based on a 10-fold cross-validation test. Among all the algorithms applied, the ideal number of most-relevant features was 78, according to the Support Vector Machine-Feature Selection (SVM-FS) algorithm, and FSH/HMG dosage for ovarian stimulation was the most important predictive factor across all examined embryo transfer features. Conclusion(s): The proposed machine learning-based prediction model could predict ET outcome and implantation of embryos with high accuracy, before the start of an ET cycle. Presented computational prediction model provides a clinical decision support tool for selecting the best embryos that lead to improvement in ART success rates.