EUVIMEDEUVIMED entdecken
Crossref

L26/P-736 Integrating Artificial Intelligence and Multivariable Modeling Uncovers Effects of LH Source on Oocyte Competence and Embryo Viability

T. Carrión, L. Carrión-Sisternas, L. Murria, L. Bori, E. Bosch, J. Bellver, M. Meseguer

Human Reproduction · 01.07.2026 · 41 (Supplement_1)

DOI: 10.1093/humrep/deag083.1063 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Human Reproduction, online erschienen am 08.07.2026. Verfasst von T. Carrión und 6 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Study question Does LH source (r-hFSH/r-hLH vs HP-hMG) affect oocyte competence, embryo quality, euploidy and clinical outcomes when evaluated using artificial intelligence (AI) and robust statistical models? Summary answer LH source impacts oocyte competence, fertilization, euploidy and clinical outcomes with effects revealed only after AI-based embryo quality stratification and remaining undetected in conventional analyses. What is known already The clinical benefit of LH supplementation during ovarian stimulation remains controversial. Conflicting results are mainly attributed to heterogeneity in study populations, stimulation protocols and analytical approaches relying on global outcomes such as pregnancy or live birth rates. These analyses often fail to account for the intrinsic biological quality of oocytes and embryos, potentially masking subtle but clinically relevant effects. Artificial intelligence–based tools enable objective and standardized assessment of oocyte competence and embryo implantation potential. Integrating AI-derived metrics with robust statistical approaches may therefore improve the evaluation of ovarian stimulation strategies and clarify the role of LH supplementation. Study design, size, duration A retrospective paired-cohort study included 287 autologous ICSI cycles performed at IVI Valencia between 2018 and 2023. Cycles were stimulated with r-hFSH/r-hLH (2:1; n = 141) or HP-hMG (n = 146). Mature (MII) oocyte images and videos of embryo development were retrospectively extracted from time-lapse incubators. A total of 481 embryos underwent PGT-A and 280 single embryo transfers had known clinical outcomes, allowing integrated analysis of oocyte competence, embryo quality, and clinical results. Participants/materials, setting, methods Oocyte competence was evaluated using MAGENTA™ IVF, generating a continuous score from 0 to 10. Embryos were classified into four AI-defined implantation potential categories (very low, low, moderate and high) using EMBRYOLY™. Outcomes included oocyte yield, fertilization, usable blastocysts, euploidy and live birth. Associations with LH source were analyzed using multivariable regression and Generalized Estimating Equations (GEE) models, adjusting for intrapatient clustering, stimulation covariates and protocol type, with sensitivity analyses restricted to GnRH antagonist cycles. Main results and the role of chance Cycles stimulated with r-hFSH/r-hLH yielded more MII oocytes than HP-hMG (8.8 ± 5.2 vs. 6.7 ± 4.8, respectively; p<0.01), an association confirmed in multivariate models (B = 2.13; p=0.015), together with AMH and total LH dose. A total of 2,190 high-quality MII oocyte images (1,251 from r-hFSH/r-hLH and 939 from HP-hMG cycles) were analyzed using MAGENTA. Oocyte quality was significantly higher in the r-hFSH/r-hLH group (5.16 ± 2.72 vs. 4.65 ± 2.48; p<0.001), confirmed by GEE in both the overall analysis (B = 2.56; p=0.004) and antagonist-only cycles (B = 2.82; p=0.003). Fertilization rates were also higher in r-hFSH/r-hLH group (77.3% vs. 71.6%; p=0.002), with increased odds of correct fertilization in GEE models (overall OR = 2.18; p<0.001; antagonists OR = 2.28; p=0.001). No significant differences were observed in overall usable blastocyst, euploidy or live birth rates. However, after stratification by AI-predicted implantation potential, low-potential embryos from r-hFSH/r-hLH cycles showed higher usable blastocyst rates (55.2% vs. 44.1%; p=0.022) and higher euploidy rates across low, moderate and high categories (45.2% vs. 30.9%, 77.8% vs 35.5% and 100% vs 52.4%; p<0.05). GEE confirmed increased odds of euploidy with r-hFSH/r-hLH (OR = 2.35; p=0.048), with a similar trend in antagonist cycles (OR = 2.95; p=0.052). Live birth rate was also higher among low-potential embryos (48.3% vs. 28.9%; p=0.024). Limitations, reasons for caution The retrospective design limits causal inference. Sample size was reduced in subgroup analyses, particularly among embryos classified in the highest AI-defined implantation potential category, potentially limiting statistical power. Therefore, larger sample size and highly controlled cohort studies may be needed to confirm our results. Wider implications of the findings AI-based embryo stratification and oocyte-level assessment reveal, for the first time, direct effects of LH source on oocyte competence and embryo viability. This biology-driven approach links ovarian stimulation protocols to intrinsic oocyte quality, uncovering effects masked by conventional analyses and highlighting the value of artificial intelligence in assisted reproduction research. Trial registration number No

Zitieren

APA

Carrión, T., Carrión-Sisternas, L., Murria, L., Bori, L., Bosch, E., Bellver, J., & Meseguer, M. (2026). L26/P-736 Integrating Artificial Intelligence and Multivariable Modeling Uncovers Effects of LH Source on Oocyte Competence and Embryo Viability. Human Reproduction, 41(Supplement_1). https://doi.org/10.1093/humrep/deag083.1063

Vancouver

Carrión T, Carrión-Sisternas L, Murria L, Bori L, Bosch E, Bellver J, et al. L26/P-736 Integrating Artificial Intelligence and Multivariable Modeling Uncovers Effects of LH Source on Oocyte Competence and Embryo Viability. Human Reproduction. 2026;41(Supplement_1). doi:10.1093/humrep/deag083.1063

BibTeX
@article{Carrin2026,
  author = {Carrión, T. and Carrión-Sisternas, L. and Murria, L. and Bori, L. and Bosch, E. and Bellver, J. and Meseguer, M.},
  title = {L26/P-736 Integrating Artificial Intelligence and Multivariable Modeling Uncovers Effects of LH Source on Oocyte Competence and Embryo Viability},
  journal = {Human Reproduction},
  year = {2026},
  volume = {41},
  number = {Supplement_1},
  doi = {10.1093/humrep/deag083.1063}
}
Kostenlos starten →