Background Recently, the combination of deep learning and time-lapse imaging provides an objective, standard and scientific solution for embryo selection. However, the reported studies were based on blastocyst formation or clinical pregnancy as the end point. To the best of our knowledge, there is no predictive model that uses the outcome of live birth as the predictive end point. Can a deep learning model predict the probability of live birth from time-lapse system? Methods This study retrospectively analyzed the time-lapse data and live birth outcomes of embryos samples from January 2018 to November 2019. We used the SGD optimizer with an initial learning rate of 0.025 and cosine learning rate reduction strategy. The network is randomly initialized and trained for 200 epochs from scratch. The model is quantitively evaluated over a hold-out test and a 5-fold cross-validation by the average area under the curve (AUC) of the receiver operating characteristic (ROC) curve. Results The deep learning model was able to predict live birth outcomes from time-lapse images with an AUC of 0.968 in 5-fold stratified cross-validation. Conclusions This research reported a deep learning model that predicts the live birth outcome of a single blastocyst transfer. This efficient model for predicting the outcome of live births can automatically analyze the time-lapse images of the patient’s embryos without the need for manual embryo annotation and evaluation, and then give a live birth prediction score for each embryo, and sort the embryos by the predicted value.

The aim of this study was to assess the relationship between early developmental kinetics and the competence to result in a live birth as well as the impact of maternal age and the number of retrieved oocytes. This retrospective cohort study included 3,021 single-embryo transfer cycles and assessed live birth outcomes paired with morphokinetic data; 1,412 transfers resulted in live births (LB), and 1,609 did not (NLB). Early morphokinetic parameters between LB and NLB embryos were compared from patients stratified into four age groups (20-25, 26-30, 31-36, and ≥37 years) and between embryos in the same competence groups within the age groups. Early morphokinetic parameters were also compared between LB and NLB embryos from patients stratified into four groups based on the number of oocytes harvested (≤7, 8-14, 15-21, and ≥22). The association between morphokinetic parameters and LB was tested using univariate and multivariate analyses. This study indicated that embryos resulting in LB generally exhibit faster developmental dynamic parameters than embryos that do not. However, this difference decreased in the younger (20-25 years) and older (≥37 years) age groups. In addition, when the number of harvested oocytes was low (≤7) or high (≥22), this difference was less obvious. The morphokinetic parameters of embryonic cleavage are an effective reference value for embryo selection strategies aimed at increasing live birth rates, especially for patients aged 26–36 years, with 8–21 harvested oocytes.

Polar bodies are tiny cells that are extruded during oocyte meiosis and are generally considered not essential for embryonic development. Therefore, polar bodies have been widely used as important materials for the preimplantation genetic diagnosis of human embryos. Recent studies have shown that polar bodies mediate embryonic development and that their morphology is related to embryo quality and developmental potential. However, the relationship between the emission of the polar body and embryonic euploidy remains unclear. In this study, a total of 1,360 blastocyst trophectoderm (TE) biopsies were performed, and blastocyst ploidy results were correlated with the state of polar bodies. The results showed that polar body angle size and polar body status are not directly related to whether the blastocysts are euploid, aneuploid, or mosaic (p > 0.05). Therefore, in the process of clinical embryo selection, embryologists should not predict the euploidy of blastocysts based on the state of polar bodies, thus affecting embryo selection.