Abstract Study question Does embryo diameter and surface area at tsB and tB as determined by AI predict live birth (LB) potential amongst euploid blastocysts? Summary answer Euploid blastocysts with optimal diameter (tsB:120-129; tB:136-140μm) and surface area (tsB:11771-13577; tB:15000-15643μm2) are more likely to lead to LB than larger or smaller blastocysts. What is known already Current embryology consensus favours larger blastocysts based on Gardner grading, associating expansion with higher quality. Human assessment of embryos is subjective, and it is unknown whether this dogma holds when assessing pre-expansion blastocyst stages. The impact of blastocyst surface area and diameter at tSB and tB; and pace of development on LB remains unclear. This may help clarify whether large blastocysts are associated with accelerated embryo growth and abnormal metabolic processes. The study’s objective was to use AI to quantitatively determine the optimal embryo surface area and diameter of euploid blastocysts that lead to live birth. Study design, size, duration Retrospective study including 97 time-lapse videos of euploid embryos with known live birth outcomes from January to December 2021 Participants/materials, setting, methods Embryo diameter and surface area were measured at start of blastulation (tSB) and at time of blastulation (tB) by an AI model (CHLOE-EQ, Fairtility Ltd). Live birth (LB) outcomes were manually annotated by embryologists. An Optimal embryo surface area and diameter were determined by the highest live birth rate and the suboptimal range was determine by a decreased Live birth rate. A comparison between the optimal and suboptimal live birth rate was measured using chi-square. Main results and the role of chance Moderately sized embryos were found to have higher LB outcomes. At tSB, optimal surface area was 11771-13577 um2, while optimal embryo diameter was 120-129μm. LB rate for optimal embryo surface area was higher than for suboptimal [Optimal 80% (20/25) vs Suboptimal 55% (40/72), p < 0.001]. In addition, optimal embryo diameter had higher LB rates than embryos with suboptimal diameter [80% (16/20) vs 52% (31/60), p < 0.05). At tB, optimal embryo area was 15000-15643μm2 and optimal embryo diameter 136-140μm. LB rate for optimal embryo surface area was higher than for suboptimal 82% (14/17) vs 49% (36/73), p < 0.05]. Optimal embryo diameter had higher LB rates than embryos with suboptimal diameter (77% 20/26 vs 47% 30/64, p < 0.05). Embryos with an optimal surface area and diameter were slower than bigger suboptimal embryos: t2 (26±2 vs 25±2), t8 (59±8 vs 53±6), t9 + (71±6 vs 67±6), tM (82±7 vs 77±8), tSB (99±6 vs 93±6), tB (107±7 vs 100±6), tEB (137±2 vs 122±12); t8-t5 (12±10 vs 6±5), t8-t2 (33±8 vs 28±5), tEB-tSB (34±2 vs 23±9), p < 0.05. Optimal vs small suboptimal embryos were faster in t9 + (71±6 vs 76±4), tEB (121±11 vs 137±2, tEB-tSB (20±10 vs 34±1). But were slower in t8-t5 (12±10 vs 5±6), t8-t2 (33±8 vs 28±5), p < 0.05. Limitations, reasons for caution This is the first time a quantitative embryo surface area and diameter has been studied in its correlation with live birth. Generalisation requires the replication across diverse centers with varying demographic profiles to validate findings and establish broader applicability within the context of assisted reproductive technologies. Wider implications of the findings This data-driven approach can support the embryo selection process, correlating quantitative parameters with live birth outcomes, thereby refining the precision of the decision-making in IVF. This level of quantitative data of embryo surface area and diameter is only accessible through AI tools. Trial registration number IRB-007CITY-FL21.08.2023
Read full abstract