L26/P-254 Comparing Ongoing Pregnancy rate After Single Vitrified-Warmed Blastocyst Transfer Using a Deep-Learning Model vs Gardner blastocyst assessment
Human Reproduction · 01.07.2026 · 41 (Supplement_1)
DOI: 10.1093/humrep/deag083.590 ↗Kurz gefasst
Fachartikel in der Zeitschrift Human Reproduction, online erschienen am 08.07.2026. Verfasst von S. Avivi und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Study question Does the implementation of a deep-learning model, such as iDA Score, improve the ability to predict ongoing pregnancy after a single vitrified-warmed blastocyst transfer (SVBT)? Summary answer The iDA-V1 model performs similarly to expert embryologists in predicting pregnancy outcomes, with significant differences in clinical pregnancy rates between highest and lowest iDA-V1 scores. What is known already Selecting the most viable embryos for transfer is a critical step in in-vitro fertilization (IVF). Traditional blastocyst assessment relies on subjective microscopic visualization based on Gardner’s standardized morphology. In recent years there has been a shift towards an automated deep-learning system for blastocysts scoring and selection such as the iDA Score. Study design, size, duration A retrospective analysis was conducted on 751 SVBT cycles using either Gardner grading or iDA Score v1.0 model (Vitrolife, Gothenburg, Sweden) according to the period the embryos were frozen. During 2022 all embryos were frozen according to Gardner’s grading and during 2023-2024 based on iDa score. The discriminative performance of each embryo scoring model in predicting pregnancy outcomes was compared. Participants/materials, setting, methods All women treated at our center who underwent single vitrified-warmed blastocyst transfer during the study period were included in the study. Clinical pregnancy rate (CPR), defined as the presence of a heartbeat. Main results and the role of chance The CPR was 26.2% for the iDA-V1 SVBT group (n = 458) and 30.3% for the Gardner grading group (n = 293). No significant differences were observed between the two groups. Among the SVBT cycles with 292 embryos scoring excellent on the iDA-V1 model (Group A, 8.5–10), the CPR was 28.4%. For 122 embryos in Group B with scores of 7.5–8.4, the CPR was 23.8%, while the remaining 7 embryos with scores below 7.5 showed a CPR of 18.2%. For SVBT embryos assessed using Gardner grading, those graded Z 2-6 AA, AB, and BA had a CPR of 30.9%, 86 embryos graded Z 2-6 BB had a CPR of 31.4%, and the remaining 14 embryos showed a CPR of 14.1%. Patient’s mean age on the day of blastocyst freezing was similar in both the iDA-V1 (32.3 ± 5.4) and Gardner grading (33.7 ± 5.6) groups. Limitations, reasons for caution Embryos in the iDA-V1 group were cultured in an EmbryoScope time-lapse incubator, whereas Gardner-group embryos were cultured in a K-System benchtop incubator. A potential selection bias due to different treatment periods cannot be excluded. Mean patient age was higher in the Gardner group (p = 0.0005) but not clinically significant. Wider implications of the findings The findings suggest that AI-based embryo assessment may serve as a reliable decision-support tool in embryo selection, performing at a level comparable to expert embryologists. Such systems could improve standardization, reduce inter-observer variability, and support clinical decision-making. Trial registration number No
Zitieren
APA
Avivi, S., Hourvitz, A., & Strassburger, D. (2026). L26/P-254 Comparing Ongoing Pregnancy rate After Single Vitrified-Warmed Blastocyst Transfer Using a Deep-Learning Model vs Gardner blastocyst assessment. Human Reproduction, 41(Supplement_1). https://doi.org/10.1093/humrep/deag083.590
Vancouver
Avivi S, Hourvitz A, Strassburger D. L26/P-254 Comparing Ongoing Pregnancy rate After Single Vitrified-Warmed Blastocyst Transfer Using a Deep-Learning Model vs Gardner blastocyst assessment. Human Reproduction. 2026;41(Supplement_1). doi:10.1093/humrep/deag083.590
BibTeX
@article{Avivi2026,
author = {Avivi, S. and Hourvitz, A. and Strassburger, D.},
title = {L26/P-254 Comparing Ongoing Pregnancy rate After Single Vitrified-Warmed Blastocyst Transfer Using a Deep-Learning Model vs Gardner blastocyst assessment},
journal = {Human Reproduction},
year = {2026},
volume = {41},
number = {Supplement_1},
doi = {10.1093/humrep/deag083.590}
}
EUVIMED