Embryo Culture Systems and Morphokinetics in IVF: The Role of Time-Lapse Monitoring and Artificial Intelligence in Embryo Selection

Abstract

Background: Embryo selection and optimal culture conditions are critical determinants of success in assisted reproductive technology (ART). Despite ongoing advances, implantation and live birth rates in in vitro fertilization (IVF) remain suboptimal. This systematic review evaluates the roles of embryo culture media (sequential versus single-step), time-lapse monitoring (TLM), and artificial intelligence (AI)-based embryo selection in improving IVF outcomes.

Methods: This systematic review was conducted in accordance with PRISMA guidelines. PubMed, Scopus, and Web of Science were searched up to June 2025. Twenty-nine studies encompassing more than 18,500 IVF/intracytoplasmic sperm injection (ICSI) cycles were included. Methodological quality was assessed using RoB2, Newcastle-Ottawa Scale, and CLAIM guidelines based on the study design. Outcomes related to embryo culture media, morphokinetic assessment using TLM, and performance of AI scoring systems were analyzed via narrative synthesis using descriptive quantitative pooling.

Results: Sequential and single-step media yielded largely comparable clinical outcomes, with single-step media showing a slight advantage in blastocyst formation rates (57.1% vs. 53.2%; p=0.03) but no significant differences in clinical pregnancy or live birth rates. TLM improved ongoing pregnancy rates in descriptive analyses (45.1% vs. 34.8%; p<0.01) and reduced early pregnancy loss. AI-based systems demonstrated superior predictive performance (AUC 0.76–0.91); however, clinical validation findings varied significantly across systems, highlighting distinct predictive performance rather than definitive, universal clinical outcome improvement.

Conclusion: While selection of culture media appears to have limited impact on clinical outcomes, TLM and AI-assisted embryo selection represent meaningful advances in IVF practice. Integration of single-step culture systems with validated AI-driven TLM tools may enhance embryo selection efficiency. Further large, prospective studies are needed to establish standardized clinical protocols and minimize underlying study heterogeneity.