P-741 Morphokinetic Goldilocks: assessing the morphokinetic range to identify embryos with the optimal chance of being euploid

Abstract Study question Can CHLOE-EQ, an AI embryo assessment support tool, automatically identify the optimal time-range of morphokinetic events where chance euploidy is maximized? Summary answer Embryos that are within normal range have increased chances being euploid compared to embryos developing at a pace outside range. What known already The introduction time-lapse technologies in IVF has revealed quantitative and qualitative parameters predict viability (ESHRE Workshop group, 2020), but their time-consuming subjective. Artificial Intelligence (AI) based tools, such as CHLOE-EQ (Fairtility), ideally suited annotate morphokinetics part tools quantify quality detect abnormalities. There been several attempts literature ploidy with morphokinetics. We postulated develop (not too fast not slow) would be more likely euploid. design, size, duration Retrospective case-controlled study 1328 videos collected 2022 from ICSI private single fertility clinic. 142 those were biopsied genetically tested by NGS. assessed embryologist tool. Participants/materials, setting, methods Time-lapse annotated using CHLOE-EQ(Fairtility) for morphokinetics, number pronucleates anomalies. frequency distribution each parameter was between aneuploid establish ranges rate. (maximum rate) sub-optimal (outside range) (t-test). Efficacy blastocyst, utilisation prediction CHLOE blast score 68hpi area under curve (AUC). Main results role For event, identification euploids identified (tPNf:21.37-25.78; t2:24.01-28.6; t3:34.07-39.20; t4:35.5-40.64; t5:46.12-53.92; t6:48.77-55.63; t7:50.22-57.45; t8:52-60.21; t9:67.35-75.55; tM:78.49-89.08; tsB:92.20-102.39; tB:99.54-109.83; tEB:106.42-120.38). Optimal smaller than total all (p < 0.001): tPNf (0.27 vs 152.36), t2(5.52 158.96), t3(22.7 159.29), t4(30.38 167.96), t5(32.02 168.29), t6(35.58 155.44), t7(41.04 157.65), t8(41.37 158.06), t9(48.85 158.39), tM(56.4 163.89), tSB(84.74 173.26), tB(93.01 168.62); tEB(95.96 164). across had higher rate suboptimal [50% (11/20), 35.35% (35/99), NS]. CHLOE-BLAST Score predictive blastulation (AUC=0.86), whilst (AUC 0.88) (AUC=0.64) Ranking (AUC=0.91) selection transfer (AUC=0.80). Limitations, reasons caution This single-center, retrospective study, only blastocysts deemed suitable biopsy ploidy. Therefore, non-blastocysts or inferior unknown, creating potential bias regarding lower cutoff threshold ranges. Wider implications findings can time maximise euploid, potentially valuable biomarker assessment, selection, managing patients expectations down individual embryos, helping reduce viable discarded. Trial registration Not applicable