Time-lapse algorithms and morphological selection of day-5 embryos for transfer: a preclinical validation study.

Area of Science:

• Reproductive medicine
• Embryology
• In vitro fertilization (IVF)

Background:

• Time-lapse imaging is increasingly used in IVF to monitor embryo development.
• Automated algorithms aim to objectively select the best embryo for transfer, but their reliability is debated.

Purpose of the Study:

• To assess the agreement between different published time-lapse algorithms for selecting day-5 embryos.
• To evaluate the concordance of these algorithms with embryologist assessments and actual embryo transfer choices.

Main Methods:

• Prospective study involving 100 IVF cycles and 428 embryos cultured in an EmbryoScope.
• Analysis of inter-algorithm agreement using Fleiss kappa coefficient.
• Comparison of algorithm-selected embryos with those chosen by embryologists.

Main Results:

• Most algorithm pairs demonstrated poor to fair agreement (kappa < 0.725).
• Agreement between algorithms and embryologists was generally poor to moderate (kappa < 0.802).
• Significant variability exists in embryo selection criteria among algorithms and human experts.

Conclusions:

• Current time-lapse algorithms show limited agreement, questioning their universal clinical applicability.
• External validation of these algorithms across diverse clinical settings is essential.
• Further research is needed to improve the accuracy and consistency of automated embryo selection.