Subjects: Six embryologists on staff at the laboratory, who ranged from junior to senior skillset levels.

Main Outcome Measures: Embryologists’ confidence levels in morphology assessments, intra-observer agreement for embryo grading, and assessment duration when using time-lapse images compared to direct observation under microscopes.

Results: From 6,435 morphology assessments across 714 embryos (62 cohorts), high confidence levels were observed on Day 5 for gradings using either direct observation (97.7%) or timelapse images (94.6%), with no statistically significant difference (p = 0.2333). Confidence was consistently high (>85%) across Days 1, 3, 6, and 7, with direct observation having a higher confidence rate than time-lapse-based assessment on Days 1 and 3 (p < 0.0001 and p = 0.0224, respectively). Intra-observer agreement between direct observation vs. time-lapse-based assessment showed good-to-excellent concordance (κ ≥ 0.60) except for blastocyst morphology grade (κ = 0.58). The highest agreement was observed for Day 5–7 embryo disposition decisions, with 94.1% concordance (κ = 0.89). Assessment durations were comparable between the two grading methods.

Conclusion: Embryologists maintained high confidence in morphology assessments using timelapse systems, even under time constraints. Intra-observer agreement for embryo grading, assessed by comparing the same embryologist’s gradings using time-lapse images vs. direct observation, was comparable to previously reported levels of agreement for traditional morphology gradings. These findings highlight the feasibility of flexible or remote assessments without disrupting embryo culture, enhancing workflow adaptability and paving the way for advancements with AI integration.

Read the full research paper here.